Argel Bandala
Sex: Male
Education:
Doctor of Philosophy in Electronics and Communication Engineering, De La Salle University, 2015
Master of Science in Electronics and Communication Engineering, De La Salle University, 2011
Bachelor of Science in Electronics and Communications Engineering, Polytechnic University of the Philippines, 2008
Field of Specialization
Pattern Recognition
Image Recognition
Computational Intelligence
Mobile Robotics
Electronic Engineering
Artificial Neural Networks
Artificial Intelligence
Researches:
Article title: Swarming Algorithm for Unmanned Aerial Vehicle (UAV) Quadrotors: Swarm Behavior for Aggregation, Foraging, Formation, and Tracking
Authors: Argel Bandala, Elmer P. Dadios, Ryan Rhay P. Vicerra, Laurence A. Gan Lim
Publication title: Journal of Advanced Computational Intelligence and Intelligent Informatics 18(5):745-751, September 2014
Abstract:
This paper presents the fusion of swarm behavior in multi robotic system specifically the quadrotors unmanned aerial vehicle (QUAV) operations. This study directed on using robot swarms because of its key feature of decentralized processing amongst its member. This characteristic leads to advantages of robot operations because an individual robot failure will not affect the group performance. The algorithm emulating the animal or insect swarm behaviors is presented in this paper and implemented into an artificial robotic agent (QUAV) in computer simulations. The simulation results concluded that for increasing number of QUAV the aggregation accuracy increases with an accuracy of 90.62%. The experiment for foraging revealed that the number of QUAV does not affect the accuracy of the swarm instead the iterations needed are greatly improved with an average of 160.53 iterations from 50 to 500 QUAV. For swarm tracking, the average accuracy is 89.23%. The accuracy of the swarm formation is 84.65%. These results clearly defined that the swarm system is accurate enough to perform the tasks and robust in any QUAV number.
Article title: Lettuce growth stage identification based on phytomorphological variations using coupled color superpixels and multifold watershed transformation
Authors: Ronnie Concepcion II, Jonnel Alejandrino, Sandy C. Lauguico, Rogelio Ruzcko Tobias, et al.
Publication title: International Journal of Advances in Intelligent Informatics 6(3): 261-277, November 2020
Abstract:
Identifying the plant's developmental growth stages from seed leaf is crucial to understand plant science and cultivation management deeply. An efficient vision-based system for plant growth monitoring entails optimum segmentation and classification algorithms. This study presents coupled color-based superpixels and multifold watershed transformation in segmenting lettuce plant from complicated background taken from smart farm aquaponic system, and machine learning models used to classify lettuce plant growth as vegetative, head development and for harvest based on phytomorphological profile. Morphological computations were employed by feature extraction of the number of leaves, biomass area and perimeter, convex area, convex hull area and perimeter, major and minor axis lengths of the major axis length the dominant leaf, and length of plant skeleton. Phytomorphological variations of biomass compactness, convexity, solidity, plant skeleton, and perimeter ratio were included as inputs of the classification network. The extracted Lab color space information from the training image set undergoes superpixels overlaying with 1,000 superpixel regions employing K-means clustering on each pixel class. Six-level watershed transformation with distance transformation and minima imposition was employed to segment the lettuce plant from other pixel objects. The accuracy of correctly classifying the vegetative, head development, and harvest growth stages are 88.89%, 86.67%, and 79.63%, respectively. The experiment shows that the test accuracy rates of machine learning models were recorded as 60% for LDA, 85% for ANN, and 88.33% for QSVM. Comparative analysis showed that QSVM bested the performance of optimized LDA and ANN in classifying lettuce growth stages. This research developed a seamless model in segmenting vegetation pixels, and predicting lettuce growth stage is essential for plant computational phenotyping and agricultural practice optimization.
Article title: Gabor-enhanced histogram of oriented gradients for human presence detection applied in aerial monitoring
Authors: Anton Louise De Ocampo, Argel Bandala, Elmer P. Dadios
Publication title: International Journal of Advances in Intelligent Informatics 6(3):223, November 2020
Abstract:
In UAV-based human detection, the extraction and selection of the feature vector are one of the critical tasks to ensure the optimal performance of the detection system. Although UAV cameras capture high-resolution images, human figures' relative size renders persons at very low resolution and contrast. Feature descriptors that can adequately discriminate between local symmetrical patterns in a low-contrast image may improve a human figures' detection in vegetative environments. Such a descriptor is proposed and presented in this paper. Initially, the acquired images are fed to a digital processor in a ground station where the human detection algorithm is performed. Part of the human detection algorithm is the GeHOG feature extraction, where a bank of Gabor filters is used to generate textured images from the original. The local energy for each cell of the Gabor images is calculated to identify the dominant orientations. The bins of conventional HOG are enhanced based on the dominant orientation index and the accumulated local energy in Gabor images. To measure the performance of the proposed features, Gabor-enhanced HOG (GeHOG) and other two recent improvements to HOG, Histogram of Edge Oriented Gradients (HEOG) and Improved HOG (ImHOG), are used for human detection on INRIA dataset and a custom dataset of farmers working in fields captured via unmanned aerial vehicle. The proposed feature descriptor significantly improved human detection and performed better than recent improvements in conventional HOG. Using GeHOG improved the precision of human detection to 98.23% in the INRIA dataset. The proposed feature can significantly improve human detection applied in surveillance systems, especially in vegetative environments.
Article title: Metaheuristic Optimization of Ammonia Factor as a Eutrophication Pollution Emission Descriptor for Trophic State Stability
Authors: Ronnie Concepcion II, Sandy C. Lauguico, Argel Bandala, Jonnel Alejandrino, et al.
Publication title: International Journal of Environmental Science and Development 11(10):460-470, October 2020
Abstract:
Aquaponic toxicity relies on the combinations of its pollution parameters that are dissolved in water and emitted in air. Ammonia is considered as an important indicator affecting aquaculture species, water nutrient imbalance and air pollution. Trophic state of aquatic body is measured by ammonia. In this study, the suitability of metaheuristic models, namely, genetic algorithm, simulated annealing, water cycle algorithm, enhanced vibrating particles system and particle swarm optimization, in determining the optimum condition of ammonia factor for providing minimal toxicity and oligotrophication was determined by varying its corresponding hyperparameters. The parameters that were optimized are water temperature and pH level. These parameters significantly affect ammonia factor that is an essential contributor to eutrophication. The optimized genetic algorithm yielded the practical-ideal fitness function value for ammonia factor as to compare with other optimized metaheuristics based on optimizing time. It selected the 50 fittest individuals based on their fitness score with the rate of 0.2 and proceeds to recombination process to extract characteristics from parent chromosomes with crossover rate of 0.8. The mutation rate of 0.01 was injected to form diversity and to test if the global solution was attained. The tournament size is 4 and the reproduction elite count is 2.5. The best condition of the ammonia factor was extracted when the number of generations has been reached. The GA results showed that the optimum condition for ammonia factor that will prevent eutrophication and provide ecological balance in aquaponic system needs a temperature of 29.254 °C and pH of 7.614.
Full text available upon request to the author
Article title: A Machine Learning Approach of Lattice Infill Pattern for Increasing Material Efficiency in Additive Manufacturing Processes
Authors: Argel Bandala, Jonnel Alejandrino, Ronnie Concepcion II, Sandy C. Lauguico, Rogelio Ruzcko Tobias, et al.
Publication title: International Journal of Mechanical Engineering and Robotics Research 9(9), September 2020
Abstract:
Additive Manufacturing (AM) has become ubiquitous in manufacturing three-dimensional objects through 3D printing. Traditional analytical models are still widely utilized for low – cost 3D Printing, which is deficient in terms of process, structure, property and performance relationship for AM. This paper focuses on the introduction of a new infill pattern – the lattice infill to increase material efficiency of 3D prints, coupled with Machine Learning (ML) technique to address geometric corrections in modelling the shape deviations of AM. Encompassed by ML algorithms, the neural network (NN) is used to handle the large dataset of the system. The 3D coordinates of the proposed infill pattern are extracted as the input of the NN model. The optimization technique of scaled conjugate gradient (SCG) is the algorithm used to train the feedforward ANN, and sigmoidal function was used as the activation type for output neurons. There is 0.00776625 cross-entropy (CE) performance and 98.8% accuracy during network training. The trained network is implemented to STL file for geometric corrections of the lattice infill pattern then made in a 3D printer slicing software. Conventional designs such as the cubic and grid infill pattern were also made for comparison. Engineering simulation software were used to simulate all three infill patterns, to measure approximate product weight, stress performance and displacement, given that there is an external force applied. Comparisons showed that the new infill pattern is more efficient than conventional infill patterns saving material up to 61.3%. Essentially increasing the amount of prints produced per spool by 2.5 times. The structure of the proposed design can also resist up to 1.6kN of compressive load prior to breaking.
Article title: Prediction of Cultivation Period and Canopy Area in Lettuce Using Multi- Temporal Visible RGB-Based Vegetation Indices and Computational Intelligence
Authors: Ronnie Concepcion II, Elmer P. Dadios, Argel Bandala, Edwin Sybingco
Publication title: International Journal of Advanced Science and Technology 29(7):12600-12625, July 2020
Abstract:
Crop growth monitoring is a manifestation of precision cultivation that demands efficient nondestructive computational phenotyping. Vegetation index (VI) plays important role in addressing the issue of vision-based crop phenotyping as the color transformation it exhibits correspond to pattern of phytomorphological descriptors of crops by enhancing vegetation properties. This study deals with predicting the cultivation period in terms of weeks after germination (WAG) and photosynthetic canopy area in mm2 based on extracted RGB-based vegetation indices from digital imagery. In this paper, computational phenotyping was employed through combined machine learning and deep learning models. The morphotype used in this study is loose-leaf lettuce and the employed complete crop life cycle is ten weeks from germination to harvesting that all happens inside a close environment microclimatic chamber with aquaponics as the cultivation technology. Multi-temporal approach of image collection was performed by capturing 30 sample lettuces every week for ten consecutive weeks using digital camera that is oriented directly downward over 12 inches stand. 15 RGB-based VIs were extracted from each image and subjected to multidimensional reduction to avoid overfitting using hybrid neighborhood component analysis (NCA), principal component analysis (PCA) and classification tree (CT). ResNet101, InceptionV3 and MobileNetV2 deep learning models, and Naïve Bayes (NB), linear discriminant analysis (LDA) and K-nearest neighbors (KNN) machine learning models were experimented to predict cultivation period. Bayesian regression neural network, regression tree and ensemble regression machine learning models were used to predict canopy area using selected RGB-based indices, namely normalized difference index (NDI), color index of vegetation extraction (CIVE), excess green minus excess red index (ExGR), vegetative index (VEG), combined indices 1 (COM1) and green minus blue index (GBI). The optimized models showed that ResNet101 with RGB color space yield the best results in cultivation period prediction with accuracy of 86.04% and coefficient of determination of 0.9211. The regression tree model with combination of NDI, CIVE, ExGR, VEG, COM1 and GBI vegetation indices in predicting canopy area resulted to a percentage difference of 0.48% and coefficient of determination of 0.8178. Thus, the developed model is highly practical and efficient for visible RGB-based imagery in crop phenotyping
Full text available upon request to the author
Article title: Environmental impact prediction of microalgae to biofuels chains using artificial intelligence: A life cycle perspective
Authors: Andres Philip Mayol, Argel Bandala, Jayne Lois San Juan, Edwin Sybingco, et al.
Publication title: IOP Conference Series Earth and Environmental Science 463:012011, April 2020
Abstract:
Biofuels derived from microalgae is an emerging technology that can supply fuel demand and alleviate greenhouse gas emissions. However, exclusively producing biofuels from microalgae remains to be commercially unsustainable because of its high investment and operating costs. A promising opportunity to address this are algal bio-refineries. Nonetheless, there is still a need to verify the environmental sustainability of this system along its entire process chain, from raw material acquisition to end-of-life. This study utilizes a life-cycle perspective approach to assess the sustainability of the algal bio-refinery and developed environmental impact prediction model using artificial intelligence, particularly adaptive neuro fuzzy inference system. Results will indicate the environmental impacts of a bio-refinery system identifying its major hotspots on different environmental impact categories. Results show that in the investigated proposed algal bio-refinery, the transesterification process had a huge contribution on the overall environmental impact having over 51.5 % of the total weight. In addition, ANFIS results showed the correlation of input parameters with respect to the environmental impact of the system. The model also indicated that there is a perfect correlation between the two parameters. The model and its accuracy should be further validated with the use of real data.
Article title: Tomato Growth Stage Monitoring for Smart Farm Using Deep Transfer Learning with Machine Learning-based Maturity Grading
Authors: Robert de Luna, Elmer P. Dadios, Argel Bandala, Ryan Rhay P. Vicerra
Publication title: Agrivita 42(1), February 2020
Abstract:
The tomato farming industry needs to adopt new ideas in applying the technology for its growth monitoring and main. Machine vision and image processing techniques have become useful in the increasing need for quality inspection of fruits, particularly, tomatoes. This paper deals with the design and development of a computer-vision monitoring system to assess the growth of tomato plants in a chamber by detecting the presence of flowers and fruits. The system also provides maturity grading for the tomato fruit. Two pre-trained deep transfer learning models were used in the study for the detection of flowers and fruits, namely, the Regional-based Convolutional Neural Network (R-CNN) and the Single Shot Detector (SDD). Maturity classification of tomato fruits are implemented using the Artificial Neural Network (ANN), K-Nearest Neighbors (KNN), and the Support Vector Machine (SVM). Evaluation results show that for the detection of flowers and fruits, the over-all accuracy of the R-CNN is 1.67% for flower detection and 19.48% for the fruit detection while SSD registered 100% and 95.99% for flower and fruit detection respectively. In the machine learning for maturity grading, SVM produced the training-testing accuracy rate of 97.78%-99.81%, KNN with 93.78%-99.32%, and ANN with 91.33%-99.32%.
Article title: Object Detection in X-ray Images Using Transfer Learning with Data Augmentation
Authors: Reagan L. Galvez, Elmer P. Dadios, Argel Bandala, Ryan Rhay P. Vicerra
Publication title: International Journal on Advanced Science and Engineering and Information Technology9(6):2147-2153, December 2019
Abstract:
Object detection in X-ray images is an interesting problem in the field of machine vision. The reason is that images from an X-ray machine are usually obstructed with other objects and to itself, therefore object classification and localization is a challenging task. Furthermore, obtaining X-ray data is difficult due to an insufficient dataset available compared with photographic images from a digital camera. It is vital to easily detect objects in an X-ray image because it can be used as decision support in the detection of threat items such as improvised explosive devices (IED's) in airports, train stations, and public places. Detection of IED components accurately requires an expert and can be achieved through extensive training. Also, manual inspection is tedious, and the probability of missed detection increases due to several pieces of baggage are scanned in a short period of time. As a solution, this paper used different object detection techniques (Faster R-CNN, SSD, R-FCN) and feature extractors (ResNet, MobileNet, Inception, Inception-ResNet) based on convolutional neural networks (CNN) in a novel IEDXray dataset in the detection of IED components. The IEDXray dataset is an X-ray image of IED replicas without the explosive material. Transfer learning with data augmentation was performed due to limited X-ray data available to train the whole network from scratch. Evaluation results showed that individual detection achieved 99.08% average precision (AP) in mortar detection and 77.29% mAP in three IED components.
Article title: Size Classification of Tomato Fruit Using Thresholding, Machine Learning, and Deep Learning Techniques
Authors: Robert de Luna, Elmer P. Dadios, Argel Bandala, Ryan Rhay P. Vicerra
Publication title: Agrivita 41(3), October 2019
Abstract:
The size of tomato fruits is closely related to the market segment and price. Manual sorting in tomato is very dependent on human interpretation and thus, very prone to error. The study presents thresholding, machine learning, and deep learning techniques in classifying the tomato as small, medium, and large based from a single tomato fruit image implemented using Open CV libraries and Python programming. Tomato images with different sizes are gathered where features like area, perimeter, and enclosed circle radius are extracted. The experiment shows that using thresholding, a classification accuracy of 85.83%, 65.83%, and 80% was achieved for area, perimeter, and enclosed circle radius, respectively. For machine learning, the training accuracy rates were recorded as 94.00%-95.00% for SVM, 97.50-92.50% for KNN and 90.33-92.50% for ANN. Comparison of models revealed that SVM is the most model without over fitting. The deep learning approach, regardless of the algorithm, produced low performances with 82.31%-78.21%-55.97% training-validation-testing accuracy for VGG16, 48.17%-41.44%-37.64% for InceptionV3, and 56.05%-44.96%-22.78% for ResNet50 models. Comparative analysis showed that machine learning technique bested the performance of the thresholding and deep learning techniques in classifying the tomato fruit size in terms of accuracy performance.
Article title: Computer vision performance metrics evaluation of object detection based on Haar-like, HOG and LBP features for scale-invariant lettuce leaf area calculation
Authors: Pocholo James Loresco, Argel Bandala, Alvin B. Culaba, Elmer P. Dadios
Publication title: International Journal of Engineering and Technology 7(4):4866-4872, February 2019
Abstract:
Leaf area can be used as a growth parameter as such it increases as the stage of lettuce progresses. Consideration of scale invariance in estimating the area poses challenging machine vision problems in a smart farm setup. To address this, a marker with a known area is utilized for the system for normalizing area measurements. This study proposes an automated object detection (marker) using Viola-Jones algorithm that uses Haar-like, HOG and LBP features. Performances of the system using each feature at varying illuminations and distances are then compared. Based on the result of this study, the highest performance in general, based on accuracy, precision, and false positive rate is achieved using HOG features.
Article title: Utilization of the Physicomimetics Framework for Achieving Local, Decentralized, and Emergent Behavior in a Swarm of Quadrotor Unmanned Aerial Vehicles (QUAV)
Authors: Argel Bandala, Reiichiro Christian S. Nakano, Ryan Rhay P. Vicerra, Laurence A. Gan Lim, et al.
Publication title: Journal of Advanced Computational Intelligence and Intelligent Informatics 21(2): 181-188, March 2017
Abstract:
This paper presents the implementation of the physicomimetics framework in governing the behavior of a swarm of quadrotors. Each quadrotor uses only local information about itself and the neighboring quadrotors to determine its own movement by applying the principles of physicomimetics. Through these localized and relatively simple interactions, the swarm of quadrotors was able to organize itself into various structures and exhibit different swarm behaviors such as aggregation, obstacle avoidance, lattice formation, and dispersion.
Article title: Smoothed Particle Hydrodynamics Approach to Aggregation of Quadrotor Unmanned Aerial Vehicle Swarm
Authors: Argel Bandala, Jose Martin Maningo, Ryan Rhay P. Vicera, Laurence A. Gan Lim, et al.
Publication title: Journal of Advanced Computational Intelligence and Intelligent Informatics 21(2):181-188, March 2017
Abstract:
This paper uses a fluid mechanics approach to perform swarming aggregation on a quadrotor unmanned aerial vehicle (QUAV) swarm platform. This is done by adapting the Smoothed Particle Hydrodynamics (SPH) technique. An algorithm benchmarking is conducted to see how well SPH performs. Simulations of varying set-ups are experimented to compare different algorithms with SPH. The position error of SPH is 30% less than the benchmark algorithm when a target enclosure is introduced. SPH is implemented using Crazyflie quadrotor swarm. The aggregation behavior exhibited successfully in the said platform.
Article title: Implementation of Swarm Social Foraging Behavior in Unmanned Aerial Vehicle (UAV) Quadrotor Swarm
Authors: Argel Bandala, Gerard Ely Ucab Faelden, Ryan Rhay P. Vicerra, Laurence A. Gan Lim, et al.
Publication title: Journal of Advanced Computational Intelligence and Intelligent Informatics 21(2): 197-204, March 2017
Abstract:
One of the novel approaches in multiple quadrotor control is swarm robotics. It aims to mimic social behaviors of animals and insects. This paper presents the physical implementation of the swarm social foraging behavior in unmanned aerial vehicle quadrotors. To achieve this, it first explores the basic behavior of aggregation. It is implemented over a quadrotor swarm test-bed that makes use of external motion capture cameras. The completed algorithm makes use of the artificial potential function model combined with the environment resource profile model. Results show successful demonstration of the social foraging algorithm with minimal error in position. Also, the proposed algorithm’s performance presents an increase in aggregation speed and time as the number of swarm member increases.
Full text available upon request to the author
Article title: Development of a flexible serpentine robot for disaster surveillance operations
Authors: Argel Bandala, John William F. Orillo
Publication title: Jurnal Teknologi 78(5-9), May 2016
Abstract:
This paper presents the development of a snake robot with vision system. This can be used for disaster aid and lessen the danger that the rescuers may encounter. The design of the snake robot considers the use of its own body segments for motion using rectilinear motion rather than using wheels. The use of segments enables the snake to move on flat and uneven terrain. Servo motors will be used for the movement of each joint and it will be powered by a lithium-polymer battery. Accelerometers and gyroscopes will serve as the input and orientation sensors, a head-mounted camera will be used to detect its location and where it is moving. An Arduino Pro Mini will be used for the controller and will be configured to receive commands from an XBee wireless transmission transceiver from the base computer. A graphical user interface in a base computer will serve as the interface of the robot’s operator and the robot. Its main movement will be based on a biological snake’s rectilinear motion which is embedded in the robot’s control system.
Full text available upon request to the author
Article title: Dynamic Aggregation Method for Target Enclosure Using Smoothed Particle Hydrodynamics Technique – An Implementation in Quadrotor Unmanned Aerial Vehicles (QUAV) Swarm –
Authors: Argel Bandala and Elmer P. Dadios
Publication title: Journal of Advanced Computational Intelligence and Intelligent Informatics XX(1):84-91, January 2016
Abstract:
This paper presents an aggregation behavior derived from fluid characteristics by adapting Smoothed Particle Hydrodynamics (SPH) Technique. The most basic behavior in a swarm-like system is aggregation. The essential requirement of a swarm is to aggregate or collect itself in proximity to a singular point in order to execute higher level swarm behaviors. The aggregation behavior is further put into use by initiating a near convergence status in a single target enclosing it by the swarm with a given specific distance by using different fluid containers. In this paper, there are three fluid containers each is introduced with different characteristics. These containers are plane, spherical and toroidal containers. Using computer simulations with different trials, the proponents were able to determine the accuracy of containing the swarm elements in a desirable area. Furthermore, the ability of the swarm to maintain collectiveness is tested. The experiment results showed that the plane fluid container yielded an accuracy of 84.88%. A spherical fluid container displayed an accuracy of 95.23%. And using toroidal particle container showed an accuracy of 92.44%.
Full text available upon request to the author
Article title: Dynamic Rate Allocation Algorithm Using Adaptive LMS End-to-End Distortion Estimation for Video Transmission over Error Prone Network
Authors: Argel Bandala, Angelo Rejaba dela Cruz, Ryan Rhay P. Vicerra, Elmer P. Dadios
Publication title: Journal of Advanced Computational Intelligence and Intelligent Informatics XX(1):106-110
Abstract:
Because of the inherent trade-off between source distortion and channel distortion in video transmission systems, joint optimization between bit-rate and distortion is still a challenging task. In this paper, we propose a method where the bit-rate allocation between source and channel encoder is controlled by the estimated end-to-end distortion at the encoder. The distortion estimation scheme is based on the adaptive forward linear predictor using least-mean square (LMS) algorithm. The forward predictor used the past values of actual end-to-end distortion to estimate the current distortion. The results show good estimate of end-to-end distortion and the proposed scheme improves video quality as compared to a standard rate control of H.264/AVC. The proposed scheme dynamically allocates the source encoder bits based on the estimated distortion.
Article title: Swarm Robot System for Underwater Communication Network
Authors: Argel Bandala, Ryan Rhay P. Vicerra, Elmer P. Dadios, Laurence A. Gan Lim
Publication title: Journal of Advanced Computational Intelligence and Intelligent Informatics 18(5):769-775, September 2014
Abstract:
This paper presents a swarm robot simulator for implementing underwater wireless communication network. Swarm intelligence is based on the collective behavior of social insects and animals such as ants, bees and others. In this paper, swarm was applied to overcome the challenges of transmitting data in a large underwater environment. A robot considered to be a member of the swarm acts as a simple “physical” carrier of the data, it moves until they converge and manage to form a link connecting the data transmitter and receiver. The system is developed, simulated and tested using a coded simulator.
Article title: Synchronized Dual Camera Vision System for Locating and Identify Highly Dynamic Objects
Authors: Noel Gunay, Argel Bandala, Elmer P. Dadios, Ryan Rhay P. Vicerra, et al.
Publication title: Journal of Advanced Computational Intelligence and Intelligent Informatics 18(5): 776-783, September 2014
Abstract:
This paper presents machine vision for locating and identifying 23 highly dynamic objects on 4.4 meters by 2.8 meters micro robot soccer playing field. The approach is based from the idea that the two camera vision subsystems should be synchronized and well informed in real time of the combined vision data and a selection of objects to track under each other’s camera view. A measure of effectiveness on using incremental tracking for two-camera operation is developed and is used to evaluate the introduced approach through experimentation. A real-time visualization of the whole playfield containing the 22 micro robots and a golf ball is also provided for the system operator to validate the objects’ actual poses with the vision system’s measurements. Results show that the proposed technique is very fast, accurate, reliable, and robust to external disturbances
Papers Presented:
Article title: Tomato Septoria Leaf Spot Necrotic and Chlorotic Regions Computational Assessment Using Artificial Bee Colony-Optimized Leaf Disease Index
Authors: Ronnie Concepcion II, Sandy C. Lauguico, Elmber P. Dadios, Argel Bandala, et al.
Conference title: 2020 IEEE Region 10 Conference (TENCON), November 2020
Abstract:
Visual inspection of plant health status and disease severity may yield subjective assessments due to error-prone sphere of colors and textures as affected by angular photosynthetic light source and the complexity of chlorosis. Quantification of damages on leaves due to destructive diseases is paramount for plant and pathogen interactions. To address this challenge, the proposed solution is the integration of computer vision and computational intelligence for tomato Septoria leaf spot necrotic and chlorotic region computational assessment. Dataset contains healthy and diseased tomato leaves that were captured individually. Non-vegetation pixels removal was done using CIELab color space. RGB color components and five Haralick texture features were extracted from the segmented leaf. Hybrid neighborhood component analysis and ReliefF algorithm were employed to select the important predictors resulting to RGB-entropy vector. A new tomato leaf disease index (tomLDI) optimized using artificial bee colony (ABC) was developed by normalizing visible red reflectance, and introducing red-green and red-blue reflectance ratios to enhance Septoria leaf spots pixels and reducing sensitivity to healthy green pixels. KNN bested classification tree, linear discriminant analysis and Naïve Bayes in detecting Septoria leaf disease with accuracy of 97.46%. Deep transfer image regression was tested using raw infected leaf images and the tomLDI transformed colored channels through MobileNetV2, ResNet101 and InceptionV3. Using tomLDI channel, MobileNetV2 and ResNet101 bested other networks in estimating leaf diseased region percentage and number of Septoria spots with R 2 values of 0.9930 and 0.9484 respectively. tomLDI channel proved to be more accurate than using raw images for regression.
Full text available upon request to the author
Article title: Genetic Algorithm-Based Visible Band Tetrahedron Greenness Index Modeling for Lettuce Biophysical Signature Estimation
Authors: Ronnie Concepcion II, Sandy C. Lauguico, Rogelio Ruzcko Tobias, Argel Bandala, Elmer P. Dadios, et al.
Conference title: 2020 IEEE Region 10 Conference (TENCON), November 2020
Abstract:
Lightness signal and color reflectance constitute the reflected luminance spectra from camera captured image to camera lenses. The intensity of lightness and visible RGB signals deviates as the camera distance to object varies. The presence of uneven distribution of photosynthetic light causes angular light effect of shadowing on the focal object and light emitting objects placed on the visually noisy background added a challenge in materializing an efficient greenness index for crop phenotyping. The proposed method in this study compensates excessive relative brightness on the image by introducing lightness rectification coefficient and employing genetic algorithm to derive a novel visible tetrahedron greenness index (gvTeGI) based on normalized green waveband. Hybrid neighborhood component analysis and Pearson's correlation coefficient approach for feature selection resulted to retaining photosynthetic canopy area, and correlation and homogeneity texture features as highly important descriptors for biophysical signatures considered in this study which are lettuce fresh weight, height and number of spanning leaves. The selection, crossover and mutation rates used to optimize the genetic algorithm model are 0.2, 0.8 and 0.01 respectively. Indoor and outdoor aquaponic system was deployed for 6-week full crop life cycle cultivation. Regression machine learning models were used to estimate biophysical signatures from extracted gvTeGI channels. Optimized Gaussian processing regression model bested regression support vector machine and regression tree in estimating fresh weight, height and number of spanning leaves with R 2 values of 0.7939, 0.7662 and 0.7446. The proposed gvTeGI proved to be more accurate than previously published greenness index for the estimation of biophysical signatures of lettuce using consumer-grade RGB camera.
Full text available upon request to the author
Article title: Adaptive Compensator of Magnetic Levitation System using Symbolic Regression
Authors: Maria Gemel B. Palconit, Rizaldo B. Fuentes, Wilen Melsedec Oficiar Narvios, Argel Bandala, et al.
Conference title: 2020 IEEE Region 10 Conference (TENCON), November 2020
Abstract:
The tuning process for a magnetic levitation to control the object's gap from the electromagnet is laborious and demands immense effort to obtain an adaptive PID compensator. Hence, this study has schemed an unexplored adaptive feedforward compensator for a 1-DOF maglev system using equation search based on a symbolic regression through an evolutionary algorithm. Results have shown an exceptional accuracy with an r 2 of 0.9997, almost zero root mean square error (RMSE) and mean absolute error (MAE). The approach has paved the way for an adaptive nonlinear system requiring a highly accurate model with a baseline dataset containing few modifiable parameters.
Article title: Towards Tracking: Investigation of Genetic Algorithm and LSTM as Fish Trajectory Predictors in Turbid Water
Authors: Maria Gemel B. Palconit, Vincent Jan Dela Cruz Almero, Marife Rosales, Edwin Sybingco, Argel Bandala, et al.
Conference title: 2020 IEEE Region 10 Conference (TENCON), November 2020
Abstract:
Monitoring the dynamics of fish behavior is impactful both in the research for fisheries and aquaculture production. One of the most explored approaches to monitor the fish is tracking-by-detection along with computer vision. Presently, there are several challenges in this field, including underwater environment conditions and fish movement complexity. This study presents an initial investigation towards tracking the fish by predicting the trajectory 2D coordinates of fish from the sequential sampled frames in underwater videos. Here, the authors explored the Genetic Algorithm based on natural evolution selection and the Long Short-Term Memory (LSTM) algorithm. Results have shown tolerable trajectory prediction inaccuracies using the GA and LSTM. Specifically, it obtained the Mean Absolute Percentage Error at 2.8% to 30.5% and 3.33% to 17.74% for GA and LSTM, respectively. These results have allowed the authors and researchers to extend its study towards tracking the fish using these approaches.
Full text available upon request to the author
Article title: Hybrid Tree-Fuzzy Logic for Aquaponic Lettuce Growth Stage Classification Based on Canopy Texture Descriptors
Authors: Rogelio Ruzcko Tobias, Matt Ervin Gatchalian Mital, Ronnie Concepcion II, Argel Bandala, et al.
Conference title: 2020 IEEE Region 10 Conference (TENCON), November 2020
Abstract:
Lettuce is one of the most popular crops for urban farming because it is easy to grow and it has high nutritional value. Moreover, it is adaptable and can be combined with other food options, or it can be eaten alone without too much preparation. Predicting lettuce growth can be crucial to find the optimum maturity and harvest time. This paper proposed to use a model of a hybrid tree-fuzzy logic approach, the classification tree was used to select the most significant features from the texture features then the fuzzy inference system was utilized in predicting the lettuce growth stage classification. The hybrid system produced accurate results with low percentage error and correct classifications. Based on these results, the most accurate prediction can be observed in the head development growth stage; the harvest growth stage has a slight variance, while the vegetative stage has the most variance. Overall, the trained hybrid system is reliable in predicting and identifying lettuce growth stage classification.
Article title: Design of A Nutrient Film Technique Hydroponics System with Fuzzy Logic Control
Authors: John Carlo Velasco Puno, Jenskie Jerlin I. Haban, Jonnel Alejandrino, Argel Bandala, et al.
Conference title: 2020 IEEE Region 10 Conference (TENCON), November 2020
Abstract:
This study presents the design and development of a nutrient film technique hydroponics system for lettuce. Hydroponics is a method of cultivating crops with the use of water with nutrient solutions as medium. This nutrient film technique hydroponics system was built as an alternative to traditional farming that requires a lot of space. This system can produce a good number of crops without consuming large land area. The system also features monitoring of the key parameters needed for by the crop to survive. A fuzzy logic control will also be used to maintain the level of the parameters. Data from the sensors for measuring electrical conductivity, pH, and as well as the water level of the mixing tank will be the input of the fuzzy logic and will control the pumps of fresh water and nutrient concentrate reservoir, and the drain of the mixing tank. The optimum values for electrical conductivity, pH, water flow rate, and temperature were all based on the existing studies that also cultivate lettuce as their primary crop.
Full text available upon request to the author
Article title: Fuzzy Irrigation System with Rain Detection and Fertilizer Control
Authors: Michael Pareja and Argel Bandala
Conference title: 2020 IEEE Region 10 Conference (TENCON), November 2020
Abstract:
Irrigation is essential for growing crops and leads to gradual growth in the economy. This research proposal aims to resolve the issue of scarcity and proper water management in the tank system through the Fuzzy Irrigation System. Fuzzy logic improves the irrigation system that includes three input parameters, such as soil moisture, soil temperature, and the water level. The combinations of these parameters will produce the time duration to have an efficient flow of water to the crop fields. Likewise, the Rain Detection Model (RDM) and the Fertilizer Control Model (FCM) are other features that support, strengthen, and innovate the system. The pilot test conducted by the researcher through MATLAB simulations were performed to check the effectiveness of the proposed system before its actual implementation.
Full text available upon request to the author
Article title: Grape Leaf Multi-disease Detection with Confidence Value Using TransfLearning Integrated to Regions with Convolutional Neural Networks
Authors: Sandy C. Lauguico, Ronnie Concepcion II, Argel Bandala, Rogelio Ruzcko Tobias, et al.
Conference title: 2020 IEEE Region 10 Conference (TENCON), November 2020
Abstract:
Identifying variant diseases in leaves is a significant method for optimizing food production. As the global population continues to arise and agricultural space continues to decline, every possible way of increasing the supply of food in any given condition and limited resources will address the above-mentioned problems. This study proposes a way for detecting three different diseases from grape leaves apart from the healthy leaves and considers the confidence value of the system in correctly identifying the classes. The diseases are namely: Black Rot, Black Measles, and Isariopsis. The system conducted a comparative analysis to determine which among the three pre-trained networks, AlexNet, GoogLeNet, and ResNet-18 will be the most suitable network to be integrated with Regions with Convolutional Neural Networks (RCNN) in performing multiple object detection in a given image. The data used in training the models comprised of annotated image data represented as a ground truth table with image files and their corresponding bounding boxes coordinates. The models evaluated resulted to AlexNet being the best pre-trained network to be working on the RCNN with an accuracy of 95.65%. The other two models from GoogLeNet and ResNet-18 only obtained accuracies of 92.29% and 89.49% respectively.
Article title: Vision-based Shrimp Feed Type Classification using Fuzzy Logic
Authors: Rex Paolo C. Gamara, Argel Bandala, Pocholo James Loresco
Conference title: 2020 IEEE Region 10 Conference (TENCON), November 2020
Abstract:
Shrimp farming is a major industry covering 23% of Philippine annual aquaculture production, which requires performing better management practices (BMPs) including growth monitoring and feed management. Traditionally, growth is monitored manually using analog weighing scale and caliper; but the manual measurement is a tedious task for large-scale farming. Feed management entails providing the most suitable feed type based on the shrimp’s current growth stage; furthermore, it addresses issues of underfeeding and overfeeding. The limitations of manual method led to the implementation of computer vision applications for growth measurement. However, existing vision-based measurement studies are not yet applied for feed management. This paper presented a fuzzy-logic based shrimp feed type classification system utilizing Mamdani’s methodology. The output classes are Starter, Grower, and Finisher based onthe three inputs: pixel area, length, and weight. The system was developed using the FIS feature of the MATLAB Fuzzy Logic toolbox. The classification system was evaluated and resulted to 93.33% correct classification accuracy. Based on these results, it can be concluded that fuzzy logic can be utilized to determine the suitable shrimp feed type corresponding to the input features.
Full text available upon request to the author
Article title: Soil Fertilizer Recommendation System using Fuzzy Logic
Authors: Jenskie Jerlin I. Haban, John Carlo Velasco Puno, Argel Bandala, Robert Kerwin Dela Cruz Billones, et al.
Conference title: 2020 IEEE Region 10 Conference (TENCON), November 2020
Abstract:
Soil nutrients and season have direct impact on the growth and yield of a crop. Deficiency on the nutrient level of the soil may result to plant disease while applying excessive amount of soil fertilizer on the other hand, may also cause negative results to the development of the crop. Nutrients on the soil also changes as the season changes from wet season to dry season. This study aims to develop a fuzzy logic-based program that will provide an appropriate amount of fertilizer to soil. The parameters such as season, nitrogen, phosphorus and potassium level are the inputs used on the fuzzy logic system. The researchers proposed four kinds of fertilizer to use in this paper such as Complete, Urea, Solophos and Muriate of Potash. Combination and amount of these fertilizers will be based on the input parameters and fuzzy rules. These soil fertilizer recommendations can be used for rice in an inbred light soil.
Full text available upon request to the author
Article title: Implementation of Automated Annotation through Mask RCNN Object Detection model in CVAT using AWS EC2 Instance
Authors: Marielet Guillermo, Robert Kerwin Dela Cruz Billones, Argel Bandala, Ryan Rhay Vicerra, et al.
Conference title: 2020 IEEE Region 10 Conference (TENCON), November 2020
Abstract:
With machine learning-based innovations becoming a trend, practical resolutions of its implementation to large-scale data and computing problems must be able to cope up as well. Currently, Graphic Processing Units (GPUs) are being chosen over other available physical devices due to its powerful computing capability and easier handling. Several cloud service providers also made it possible for these to be accessible online allowing higher serviceability and lower cost upfront for businesses. With this said, the proponent would implement a common machine learning-based application, automated annotation through Mask RCNN Object Detection Model in CVAT, using AWS instance. The key purpose is to showcase the viability of deploying data and computing intensive system on the cloud.
Full text available upon request to the author
Article title: A Smart Space with Music Selection Feature Based on Face and Speech Emotion and Expression Recognition
Authors: Jose Martin Maningo, Argel Bandala, Ryan Rhay P. Vicerra, Elmer P. Dadios, et al.
Conference title: 2020 IEEE Region 10 Conference (TENCON), November 2020
Abstract:
The technological capabilities of computers in today's time continues to improve in ways that seemed impossible before. It is common knowledge that most people use computers to make everyday lives easier. Therefore, it is vital to bridge the gap between humans and computers to provide more suitable aid to the user. One way to do this is to use emotion recognition as a tool to make the computer understand and analyze how it can help its user on a much deeper level. This paper proposes a way to use both face and speech emotion recognition as a basis for selecting an appropriate music that can improve or relieve one's emotion or stress. To accomplish this, Support Vector Machine with different kernels are used to create the models for validation and testing on both the face and speech emotion recognition. The final integrated system yielded an accuracy rate of 78.5%.
Full text available upon request to the author
Article title: Prediction of Total Body Water using Scaled Conjugate Gradient Artificial Neural Network
Authors: Marife Rosales, Maria Gemel B. Palconit, Argel Bandala, Ryan Rhay P. Vicerra, et al.
Conference title: 2020 IEEE Region 10 Conference (TENCON), November 2020
Abstract:
The study aims to design an intelligent total body water measuring device which will help to determine the total body water level or percentage of an individual using ultrasonic sensor, load cell and bioelectric impedance analysis (BIA). The system utilized the Scaled Conjugate Gradient Artificial Neural Network (ANN) as the machine learning algorithm. The system used the dataset splitting of 70%-15%15% for training, validation and testing. Different hidden neurons were used and compared during neural network training and found out that using 10 neurons will provide the lowest mean square error (MSE) with best value of Pearson’s correlation (R). Based on the results, using 10 neurons, Scaled Conjugate Gradient algorithm has better performance as compared to Levenberg-Marquardt algorithm with MSE equal to 0.180033, 0.118954, 0.529157 while the R value is equal to 0.997887, 0.997488, 0.99644 for training, validation and testing.
Full text available upon request to the author
Article title: Transfer Learning Approach for the Classification of Conidial Fungi (Genus Aspergillus) Thru Pre-trained Deep Learning Models
Authors: Matt Ervin Gatchalian Mital, Rogelio Ruzcko Tobias, Herbert Villaruel, Argel Bandala, et al.
Conference title: 2020 IEEE Region 10 Conference (TENCON), November 2020
Abstract:
The Aspergillus genus is deemed relevant for distinction and classification in the field of food, agriculture and medicine. As there are harmful and useful ones, it adds to the necessity of correct classification. Categorization of this conidial fungi is usually done through manual microscopical procedures which apparently has a degree of subjectiveness. In order to classify Aspergillus samples faster and more accurately, technology, specifically image processing and machine learning are incorporated in this study. Pre-trained deep learning models are employed in classifying 9 kinds of Aspergillus. The methodology is generally comprised of preprocessing, deep-learning (training) and performance evaluation. Performance evaluation pertains to the validation accuracy and running times of the system after training through visual display of graphs and tabulation of acquired data. This study achieved a 93.3333% testing accuracy proving that the transferred knowledge is accurate, compatible and reliable.
Full text available upon request to the author
Article title: Fuzzy Power Control for Non-linear Distortion Suppression in MIMO-OFDM Systems
Authors: Genesis Marr N. Principe, Ryan Rhay R. Vicerra, Argel Bandala
Conference title: 2020 IEEE Region 10 Conference (TENCON), November 2020
Abstract:
The hybridization of MIMO-OFDM systems became one of the most used wireless communication model for broadband, mobile, and multimedia applications because of its high bandwidth efficiency, bandwidth capacity, and robustness to fading. However, it suffers from the underlying disadvantage of OFDM system which is having a high peak-to-average-power ratio (PAPR) due to large envelope variations. These variations cause non-linear distortion when the OFDM signal is amplified for transmission. Hence, in order to eliminate the non-linear distortion effects of the high power amplifier in MIMO-OFDM systems, the input signal power must have an appropriate power level to satisfy an optimal input back off (IBO) value that also contributes to an amplifier’s maximum efficiency. A Fuzzy Logic Controller is used to control the IBO of the system as well as the signal power level. Results shows that using the proposed Single-Input Single-Output (SISO) Fuzzy Power Controller reduces the bit error rate (BER) significantly compared to the traditional scheme.
Full text available upon request to the author
Article title: Identification of Corn Plant Leaf Diseases through Web Server using Image Processing and Artificial Neural Network
Authors: Dailyne Macasaet, Edwin Sybingco, Argel Bandala, Ana Antoinette Cabantug Illahi, et al.
Conference title: 2020 IEEE Region 10 Conference (TENCON), November 2020
Abstract:
This study centers on the design and development of a microcontroller based hardware interface that connects the serial camera, the processor, the WiFi module, and the LCD screen and identification software for corn plant diseases through web-server using image processing and artificial neural network. This is done by capturing and displaying the image of the leaf inside the box and transmits it to the web server as an input image; process, analyze and interpret the data through image processing. The result of the processed image will be sent to the displaying microcontroller based hardware interface through the web-server and display the Pest Management Recommendations.
Full text available upon request to the author
Article title: Visual Classification of Lettuce Growth Stage based on Morphological Attributes using Unsupervised Machine Learning models
Authors: Jonnel Alejandrino, Ronnie Concepcion II, Sandy C. Lauguico, Rogelio Ruzcko Tobias, Argel Bandala, et al.
Conference title: 2020 IEEE Region 10 Conference (TENCON), November 2020
Abstract:
Food shortage is a serious problem facing the world and is prevalent in urban areas. The scarcity of food is mainly caused by crop failure. Environmental factors offered by the rural areas determine the condition of crops to be produced. This scenario pomps, the explication of urban farming. However, urban farming requires all-out monitoring and control. This study specifically solves the predicament of identifying the developmental growth of plants from seed leaf to amend the techniques of plant science and cultivation management. With a view to this, the paper shows coupled color-based superpixels and multifold watershed transformation in segmenting the lettuce image from the background. To fathom it out, a comparative analysis of three unsupervised machine learning algorithms: Self Organizing Map (SOM), Hierarchical, and K - means algorithms were conducted. These were done by modeling each algorithm from the features extracted from morphological computations of the lettuce images raised in a smart aquaponics setup. Each of the models was optimized to increase cross and hold-out validations. The results showed that K – means algorithm having the parameters of algorithm = ‘auto’, copyx= ‘True’, init = ‘K- means++’, maxiter = ‘1000’, nclusters = ‘3’, ninit = ‘15’, n_jobs = ‘1’, precompute_distance = ‘auto’, random_state = ‘10’, tol = ‘0.000001’, verbose = ‘1’, leaf_size = ‘10’ was the most effective model for the given dataset, yielding a high precision and recall unsupervised clustering percentage of 91%.
Article title: Battery Management System with Temperature Monitoring Through Fuzzy Logic Control
Authors: Hilario Calinao, Argel Bandala, Reggie Gustilo, Elmer P. Dadios, et al.
Conference title: 2020 IEEE Region 10 Conference (TENCON), November 2020
Abstract:
Batteries are very important in many different applications. In the solar energy system, the batteries are used as power storage when solar energy is not available especially during night time. Batteries need to be maintained and closely monitor their condition. Battery management systems are normally used for this application but many of them are not monitoring the battery’s temperature. This study will use a fuzzy logic-controlled system to manage the operation of the battery. This system will maintain the operation of the battery in the allowed operating temperature to prevent it from damaged caused by excessive internal temperature.
Full text available upon request to the author
Article title: Crack Detection With 2D Wall Mapping For Building Safety Inspection
Authors: Jose Martin Maningo, Argel Bandala, Then Anjerome Bedruz, Elmer P. Dadios
Conference title: 2020 IEEE Region 10 Conference (TENCON), November 2020
Abstract:
In the Philippines, the number of earthquakes occurring has risen to an alarming rate. ’The Big One’ is one of the biggest expected catastrophes that is undoubtedly going to occur in the next decade as said by various experts. Buildings that were able to withstand the upcoming earthquakes, are to be inspected by engineers without knowing if the safety of the building is compromised. Thus, there is a need for a system that can inspect the cracks on the wall for faster and safer inspection. The objective of this study is to develop a crack detecting system capable of analyzing physical characteristics of cracks and mapping the surface of the wall. The model to be used for classifying and determining what cracks are, was trained with the use of Faster R-CNN machine learning architecture. Trained using the SDNET2018 combined with actual data generated by the proponents, the resulting system can detect cracks with an accuracy of 90% and classify the cracks according to the shape. The system also calculates its physical properties and has a recommender system that provides remarks on what necessary actions can be done.
Full text available upon request to the author
Article title: Human Presence Detection using Ultra Wide Band Signal for Fire Extinguishing Robot
Authors: Argel Bandala, Edwin Sybingco, Jose Martin Maningo, Elmer P. Dadios, et al.
Conference title: 2020 IEEE Region 10 Conference (TENCON), November 2020
Abstract:
Fire incidents often result to associated deaths, injuries, and losses occurring structures and properties, particularly in homes every year. In this study, the researchers proposed a 4-wheeled fire extinguishing robot with the ability to detect human presence in the area even when there is fire. Multiple sensors are utilized in this study to detect nearby flame, smoke, temperature and humidity, and obstacles through integration with Arduino and Raspberry Pi. The proposed robot is remotely controlled by the user over Wi-Fi through the graphical user interface created by the researchers in Python for easy monitoring of data and control. A camera is also mounted to the robot for surveillance purposes. The human detection system of the robot is implemented through using ultra-wide band radar (UWB) by utilizing the X4M300 presence sensor, which could detect human presence based on their respiration movement. Initial testing and four experiments were conducted to test the radar sensor's capabilities compared to the existing methods of human detection. The researchers yielded an accuracy of 97.29% in the testing of human detection system, proving that the implementation of UWB radar sensor is successful.
Full text available upon request to the author
Article title: Particle Swarm Optimization-based Dark Channel Prior Parameters Selection for Single Underwater Image Dehazing
Authors: Vincent Jan Dela Cruz Almero, Jonnel Alejandrino, Ronnie Concepcion II, Argel Bandala, et al.
Conference title: The 9th International Symposium on Computational Intelligence and Industrial Applications (ISCIIA 2020), November 2020
Abstract:
Underwater images are confronted with blurriness and poor color consistency due to the haze produced by the absorption and scattering effects of the turbid water. Dark Channel Prior (DCP) is the state-of-the-art and the algorithmic basis to solve underwater image restoration. However, the default parameters of DCP may not be applicable to underwater images with different levels of degradation. The selection of the appropriate DCP parameters for each underwater image is considered as an optimization problem and can be solved using Particle Swarm Optimization (PSO). The proposed PSO-based selection algorithm is defined by its operators: objective function, swarm size, inertial weights and acceleration coefficients. Obtaining appropriate combination of these ope rators are elaborated. The qualitative and quantitative evaluations observed acceptable visual improvements and measurements to underwater images applied with DCP at optimally selected parameters, in comparison to underwater images applied with DCP at default parameters. Hence, the proposed algorithm provides good adaptability and effectivity to the exhaustive search of appropriate DCP parameters.
Full text available upon request to the author
Article title: Segmentation of Aquaculture Underwater Scene Images based on SLIC Superpixels Merging-Fast Marching Method Hybrid
Authors: Vincent Jan Dela Cruz Almero, Jonnel Alejandrino, Argel Bandala, Elmer P. Dadios
Conference title: 2020 IEEE Region 10 Conference (TENCON), November 2020
Abstract:
Segmentation is a challenging task for the complex and low-quality underwater images, as this is prerequisite to advanced tasks in fish monitoring such as fish detection and classification. A demand exists for underwater image segmentation algorithms that can robustly segment fish from its background. A competitive approach is the integration of states-of-the-art image segmentation algorithms: SLIC superpixels merging by KAZE Keypoints clustering and Fast Marching Method (FMM) to a single framework. The combination of these established methods offers robustness towards underwater images of different visual qualities. First, a locally acquired underwater image is represented as superpixels. Then, the KAZE features of an underwater image is extracted. Such features are utilized by the k-means clustering to group superpixels which contains fish pixels into a region. Lastly, the merged region is further segmented with Fast Marching Method and corresponding morphological processes. The study presents the viability of the integration of different image segmentation techniques for localized application. The number of superpixels, KAZE Keypoint score threshold and FMM threshold are identified to affect the performance of the proposed algorithm. Qualitative observations and quantitative measures validate the robustness of this generated algorithm to address this difficult and persistent task.
Full text available upon request to the author
Article title: Genetic Algorithm-based Dark Channel Prior Parameters Selection for Single Underwater Image Dehazing
Authors: Vincent Jan Dela Cruz Almero, Ronnie Concepcion II, Jonnel Alejandrino, Argel Bandala, et al.
Conference title: 2020 IEEE Region 10 Conference (TENCON), November 2020
Abstract:
Dehazing through Dark Channel Prior (DCP), originally developed for land-based images, has translated its potential for improving the quality of underwater images. However, the DCP default parameters, which are just adapted from land-based applications, may not be applicable for underwater images. Such constraint limits the capability of this restoration algorithm to improve the quality of an underwater image; the values of these parameters must be searched for each underwater image. A proposed approach on the parameter values assignment problem is to conduct an optimized search based on Genetic Algorithm. The presentation of this proposed approach focuses on the Genetic Algorithm processes: chromosome encoding, fitness function development, and selection, mutation, and crossover, to perform an effective search of the best solution out of a pool of possible solutions. Qualitative and quantitative evaluations show that utilization of optimized combination of DCP parameters, achieves images of higher quality in comparison to the utilization of established default DCP parameters.
Full text available upon request to the author
Article title: Estimation of Photosynthetic Growth Signature at the Canopy Scale Using New Genetic Algorithm-Modified Visible Band Triangular Greenness Index
Authors: Ronnie Concepcion II, Sandy C. Lauguico, Rogelio Ruzcko Tobias, Elmer P. Dadios, Argel Bandala, et al.
Conference title: 2020 International Conference on Advanced Robotics and Intelligent Systems, August 2020
Abstract:
Greenness index has been proven sensitive to vegetation properties for multispectral and hyperspectral imaging. However, most controlled microclimatic cultivation chambers are equipped with low-cost RGB camera for crop growth monitoring. The lack of camera credentials specially the wavelength sensitivity of visible band provides added challenge in materializing greenness index. The proposed method in this study compensates the unavailability of generic camera peak wavelength sensitivities by employing genetic algorithm (GA) to derive a visible band triangular greenness index (TGI) based on green waveband signal normalized TGI model called gvTGI. The selection, mutation and crossover rates used in configuring the GA model are 0.2, 0.01 and 0.8 respectively. Lettuce images are captured from an aquaponic cultivation chamber for 6-week crop life cycle. The annotated and extracted gvTGI channels are inputted to deep learning models of MobileNetV2, ResNet101 and InceptionResNetV2 for estimation of photosynthetic growth signatures at canopy scale. In predicting cultivation period in weeks after germination, MobileNetV2 bested other image classification models with accuracy of 80.56%. In estimating canopy area, MobileNetV2 bested other image regression models with R 2 of 0.9805. The proposed gvTGI proved to be highly accurate on estimation of photosynthetic growth signatures by using generic RGB cameras, thus providing a low-cost alternative for crop phenotyping.
Full text available upon request to the author
Article title: Android Application for Chest X-ray Health Classification From a CNN Deep Learning TensorFlow Model
Authors: Rogelio Ruzcko Tobias, Luigi Carlo De Jesus, Matt Ervin Gatchalian Mital, Argel Bandala, et al.
Conference title: 2020 IEEE 2nd Global Conference on Life Sciences and Technologies, March 2020
Abstract:
One of the problems in the medical field is incorrect diagnosis, particularly over-diagnosis and under diagnosis. One of the illnesses that is currently researched upon is pneumonia. Several methodologies are employed to further validate this diagnosis. Often, to achieve the goal, medical experts rely on an x-ray image. In this study, the basis is still x-ray images with the incorporation of image processing and machine learning. MobileNetV2 is utilized as the convolution neural network model. The produced frozen graph is injected to Android Studio to produce an android mobile application which will serve as a diagnostic tool. The mobile application has high accuracy and considered reliable because of testing and validation results. This study generally aims to provide a reliable low-cost aid for medical professionals in diagnosing pneumonia.
Article title: Faster R-CNN Model With Momentum Optimizer for RBC and WBC Variants Classification
Authors: Rogelio Ruzcko Tobias, Luigi Carlo De Jesus, Matt Ervin Gatchalian, Argel Bandala, et al.
Conference title: 2020 IEEE 2nd Global Conference on Life Sciences and Technologies, March 2020
Abstract:
Since many diseases and infections are dependent on the count and type of Red Blood Cells (RBCs) and White Blood Cells (WBCs) present in the blood stream, detection and classification pertaining to them is necessary and relevant. Based from existing related literature, ordinary Neural Networks are usually employed. Also, in existing researches, RBC types are the main focus. Hence, after observing research gaps, a Faster Region-based Convolutional Neural Network (Faster R-CNN) was utilized for this study, focusing not only on RBCs but also on the variants of WBCs. The aim is to have a fast and reliable system in order to achieve the goal of aiding the medical field in the classification of RBCs and WBCs.
Article title: Utilization of Genetic Algorithm in Classifying Filipino and Korean Music through Distinct Windowing and Perceptual Features
Authors: Matt Ervin Gatchalian Mital, Rogelio Ruzcko Tobias, Argel Bandala, Robert Kerwin Dela Cruz Billones, et al.
Conference title: 2019 International Conference on Contemporary Computing and Informatics, December 2019
Abstract:
Classification of songs or music in terms of genre, era and any other categories has been sought to be one of the most common yet significant research fields in digital signal processing. Usually, the aim to distinguish musical patterns is only limited to one general type (e.g. American Music). The objective of this study is to perceive the differences and similarities between two general categories namely: OPM (Original Pilipino Music), the apparent representative music of the Philippines and one of the fastest growing music industries K-POP, a general term for contemporary Korean Music. Through the features acquired from jAudio and aid of a genetic algorithm model constructed in Python with the accompaniment of the TPOT library, this research is successful in classifying the music under various settings and desired outputs.
Article title: Throat Detection and Health Classification Using Neural Network
Authors: Rogelio Ruzcko Tobias, Luigi Carlo De Jesus, Matt Ervin Gatchalian Mital, Argel Bandala, Sandy C. Lauguico, et al.
Conference title: 2019 International Conference on Contemporary Computing and Informatics, December 2019
Abstract:
In some instances, physicians’ diagnosis may not be accurate; they may over or under diagnose a patient with throat conditions resulting to improper medications and antibiotics. In order to aid them, a vision system that focused on Histogram of Gradients (HOG) and later on integrated in a neural network is implemented. The system made is in accordance to what is desired with accurate values for testing and validation. Pre-processing of images are done by employing Cascade Trainer; on the other hand, the main training, detection, and classification are implemented in MATLAB.
Article title: Dynamic Peloton Formation Configuration Algorithm of Swarm Robots for Aerodynamic Effects Optimization
Authors: Rhen Anjerome Bedruz, Jose Martin Maningo, Arvin Fernando, Argel Bandala, et al.
Conference title: 2019 7th International Conference on Robot Intelligence Technology and Applications, November 2019
Abstract:
This paper presents a flocking and formation algorithm adapted from the flocking behavior of cycling team or pelotons. Several multi agent applications require efficient positioning of the agents in static and dynamic tasks. It was verified physically that an optimal distance in a peloton formation, the agents take reduced drag due to the inherent drag resistant characteristic of the formation. The said conditions were implemented in an algorithm in a swarm of wheeled robots. Experiment results show that the optimal distance between agents were attained. It was shown that the adaptation of peloton behavior in artificial agents brought efficient formation and foraging trajectories and behaviors.
Full text available upon request to the author
Article title: Longitudinal Wheel Slip Regulation using Nonlinear Autoregressive-Moving Average (NARMA-L2) Neural Controller*
Authors: Ryan Christopher R. Dajay, Jason Española, Argel Bandala, Then Anjerome Bedruz, et al.
Conference title: 2019 7th International Conference on Robot Intelligence Technology and Applications, November 2019
Abstract:
In this study, the implementation of a nonlinear autoregressive-moving average model ( NARMA-L2) neural network controller to maximize the traction of tires during braking scenarios was explored. The proposed controller and system dynamics was done in Simulink. All in all, the neural network controller shows good stability and good response in following the reference trajectory or desired slip ratio. It has experienced the peak worst error of around 2%, its best performance was reached after 89 epochs and it can reach around 99.5% of the reference trajectory or desired slip ratio. Further research should focus on hardware implementation, integration with slip estimation techniques , and, better sets of training data to make the controller more adaptive to different environment and road surface characteristics.
Full text available upon request to the author
Article title: Design of a Robot Controller for Peloton Formation Using Fuzzy Logic
Authors: Rhen Anjerome Bedruz, Argel Bandala, Ryan Rhay P. Vicerra, Ronnie Concepcion II, et al.
Conference title: 2019 7th International Conference on Robot Intelligence Technology and Applications, November 2019
Abstract:
This paper presents a controller for the optimization of flocking and formation algorithm adapted from the flocking behavior of cycling team or pelotons. The controller developed is a fuzzy-logic controller for each of the robotic agent in order for them to perform a peloton formation. Results from the simulation shows that the developed fuzzy logic controller is slightly better than the mathematical models in maintaining a small and optimal position for the peloton formation which results to a more efficient and robust swarm system.
Full text available upon request to the author
Article title: Development of Leap Motion Capture Based - Hand Gesture Controlled Interactive Quadrotor Drone Game
Authors: Argel Bandala, Jose Martin Maningo, Edwin Sybingco, Ryan Rhay P. Vicerra, et al.
Conference title: 2019 7th International Conference on Robot Intelligence Technology and Applications, November 2019
Abstract:
This paper presents an interactive drone game which is controlled by bare hand gestures. A typical handheld or gamepad remote controller bids inconvenience due to unnatural hand movements of the pilot. This study removes the use of handheld controllers and instead employ a contact less hand gesture device for drone control. Several applications can be explored like task designation in a multi-drone system for construction, inspection and relief operations. The drone is controlled using a PID controller which maintains a given trajectory which is determined by the hand gesture of the controller. The platform chosen by this research is an interactive game of two drones which fire IR lasers. The objective of the game is to hit the opponent with the laser emission. This platform showcases the accurateness and reliability of the hand gesture control system by controlling a real time drone application. The accuracy and precision of the movements based on the gesture of the pilot will be tested through simulations and actual implementations.
Full text available upon request to the author
Article title: Innovating Academic Writing through Flipped Classroom Instruction
Authors: Dylyn A. Junio and Argel Bandala
Conference title: 2019 IEEE 11th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM), November 2019
Abstract:
Flipped Classroom Instruction has become a trend and a popular instructional model of the recent time in which the typical lecture and homework features of a course are inverted. This method transforms classroom pedagogy into dynamic and interactive manner where the teacher guides and facilitates students' learning process. The present study investigates the students' perceptions on flipped classroom instruction in academic writing class in contrast to traditional teaching. To achieve this, an instructional design infused with multimedia tools in writing lessons were provided and implemented in writing classes. The mixed method was used to collect qualitative and quantitative data. The results revealed a positive impact of flipped learning to students which also indicate their improved academic writing skills in certain ways.
Full text available upon request to the author
Article title: Design of Controller and PWM-enabled DC Motor Simulation using Proteus 8 for Flipper Track Robot
Authors: Cyrus Lawrence Camancho Bual, Rachel D. Cunanan, Argel Bandala, Then Anjerome Bedruz
Conference title: 2019 IEEE 11th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM), November 2019
Abstract:
The developed tracked mobile robot such as flipper track robot increases its ability and capability in overcoming more challenges in urban environment context and rough terrains. In addition, flippers are its support in dealing with this circumstances. The configuration of flipper tracked robots came from the extended version of conventional two-tracked mobile robot such as two and four-tracked robots. Then, the study aims to create a dedicated controller for the modified flipper track robot. Correspondingly, the target instruments, display and analog control are identified for adept monitoring of the robot status while doing its intended function. Afterwards, using Proteus 8 Professional simulation software, the Arduino UNO controller as main MCU, 16x2 LCD, analog joysticks in terms of analog resistors, and virtual terminal for serial print monitoring are attached and wired accurately. The nine speed level is established and paralleled to the required PWM output for the fine movement of flipper track robot and also the map function of Arduino IDE for degree manipulation of servo motor of two flipper arms. Finally, the results are shown in LCD which matches the established logical conditions of nine speed level as well as the status movement of the flipper track robot. The functionality and feasibility of the controller is verified and exhibited.
Full text available upon request to the author
Article title: Detection of Gas Harmful Effect using Fuzzy Logic System
Authors: Ana Antoniette Cabantug Illahi, Argel Bandala, Elmer P. Dadios
Conference title: 2019 IEEE 11th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM), November 2019
Abstract:
The researcher focuses on determining the harmful effect of gases in human. Using fuzzy logic is the core process of the study. It has an ability to turn the computer to think as a human because of the rules that is embedded in the whole system. Carbon Monoxide, Cyanide, Formaldehyde, Ammonia, and Hydrogen Sulfide was use as a sample of harmful gases. There are certain level of concentration and length of time a human can tolerate the harmful gas. Without any detection of it a person will suffer a bad effect of the gas. The system can tell if the gas in certain level and time is harmful or not. The result of the system is correct in correlation with the rules set in the system and the graph that proves the validity of the system.
Full text available upon request to the author
Article title: Development of a Fuzzy Logic Controller for a Smart RGB Lighting System
Authors: Neil Oliver M. Velasco, Jay Robert Rosario, Argel Bandala
Conference title: 2019 IEEE 11th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM), November 2019
Abstract:
The smart RGB lighting system is a smart lighting system built on a fuzzy logic controller that adjusts the RGB lighting to suit the environment. The concept aims to make lighting systems more efficient in power, and intelligent with the color adjustment from the fuzzy system. This research aims to develop a fuzzy logic controller that aims to control the output RGB light intensity based on the current luminance of the environment and the activity color classification within the room. The membership functions and rules in the system were designed in MATLAB Fuzzy Logic Designer. The system was tested with test inputs into the system.
Full text available upon request to the author
Article title: Classification of Confusion Level Using EEG Data and Artificial Neural Networks
Authors: Claire Receli Morales Renosa, Argel Bandala, Ryan Rhay P. Vicerra
Conference title: 2019 IEEE 11th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM), November 2019
Abstract:
The purpose of this study is to create an artificial neural network (ANN) that can classify a person’s level of confusion using Electroencephalography (EEG) data, more specifically, using the power spectrum of all the brain wave frequencies. This could help people in understanding the complicated mechanisms present in the brain, including the role that each specific brain wave signal plays in the formation of different cognitive activities in one’s mind such as confusion and workload. This study is categorized as a cognitive-affective state research, inspired by its current possible application to different existing societal fields such as education and gaming industries. The processing platforms used to process and interpret the dataset used in this research are Microsoft Excel and MATLAB software, applying frequency-based analysis and standard averaging methods fit for EEG data classification and artificial neural network modeling.
Article title: A Fuzzy Logic-Based Stock Market Trading Algorithm Using Bollinger Bands
Authors: Sandy C. Lauguico, Ronnie Concepcion II, Jonnel Alejandrino, Argel Bandala, et al.
Conference title: 2019 IEEE 11th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM), November 2019
Abstract:
Stock market price forecasting with the use of Technical Analysis is not precise Mathematics. Mostly, prediction is only based on probabilities supported by historical data and patterns. With these, several technical strategies were made by traders to produce signals on trading execution. This study proposes an algorithm that undergoes a certain trading strategy using three fuzzy logic controllers. Technical indicators such as candlestick parameters and Bollinger Bands (BB) were used for triggering the strength of buy, hold, and sell signals. Stock price data were gathered from a certain stock company. These data contain the opening and closing prices that are utilized for computing the BB. The raw and the computed values are the crisp input parameters for the Fuzzy Inference System (FIS). The membership functions were classified to very low, low, high, and very high levels depending on the input default parameters used by traders. Sets of rules were created fuzzy logically to produce signals indicating the strength of an execution recommendation. The system is implemented using NI LabVIEW and MATLAB, proving that the tests are yielding acceptable result of about 94.44%.
Article title: Neural Network Modeling for Fuel Consumption Base on Least Computational Cost Parameters
Authors: Ana Antoniette Cabantug Illahi, Argel Bandala, Elmer P. Dadios
Conference title: 2019 IEEE 11th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM), November 2019
Abstract:
Fuel consumption are important in every vehicle. This study investigates the performance of fuel usage in an engine. The investigation is done through a combination of experimental data analysis and artificial neural network (ANN) modeling. Back propagation neural network was used to determine the optimized fuel consumption. There is a lot of factor which has an effect to fuel consumption in conventional drive procedure, however in this study the factors affecting the fuel consumption are the distance, time, acceleration, and velocity of a car. These parameters are used as input information for the neural network training and fuel consumption prediction as output. This study shows the ANN capability to predict the fuel consumption using MATLAB neural fitting tool. The result demonstrated that the system using neural network is efficient for predicting fuel consumption of an engine.
Full text available upon request to the author
Article title: A Bi-Objective Optimization Model for a Retail Inventory System with Perishable Products
Authors: Phoebe Lim Ching, Dennis Cruz, John Anthony Cheng Jose, Argel Bandala, et al.
Conference title: 2019 IEEE 11th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM), November 2019
Abstract:
For perishable inventories, cost alone is not a sufficient performance indicator. Minimizing costs can yield scenarios where demand is serviced with older units to allow for less frequent orders and maximize the utility of available inventories. This is contrary to the customers' needs, as perishables may have a tendency to deteriorate over time. This study proposes a bi-objective model for managing perishables, with cost and freshness as the system objectives. The proposed model may be used to develop policies for ordering and issuance, which directly affect quality. These would allow for the purposeful movement of inventory units across the system. When the results of the single-objective and bi-objective models were run, it was found that a cost-centric model had a tendency to accumulate older inventory, which could be used to service periods with low demand. This allowed it to work within its capacity constraints while negating the need to order during periods with low demand. The introduction of a quality objective removed this tendency, resulting in fresher inventory and lower inventory levels on the average. The model serves as a base model for further studies, to determine how new policies and technology may be employed to achieve higher quality as well as minimal costs.
Full text available upon request to the author
Article title: YOLO-based Threat Object Detection in X-ray Images
Authors: Reagan L. Galvez, Elmer P. Dadios, Argel Bandala, Ryan Rhay P. Vicerra
Conference title: 2019 IEEE 11th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM), November 2019
Abstract:
Manual detection of threat objects in an X-ray machine is a tedious task for the baggage inspectors in airports, train stations, and establishments. Objects inside the baggage seen by the X-ray machine are commonly occluded and difficult to recognize when rotated. Because of this, there is a high chance of missed detection, particularly during rush hour. As a solution, this paper presents a You Only Look Once (YOLO)based object detector for the automated detection of threat objects in an X-ray image. The study compared the performance between using transfer learning and training from scratch in an IEDXray dataset which composed of scanned Xray images of improvised explosive device (IED) replicas. The results of this research indicate that training YOLO from scratch beats transfer learning in quick detection of threat objects. Training from scratch achieved a mean average precision (mAP) of 45.89% in 416×416 image, 51.48% in 608×608 image, and 52.40% in a multi-scale image. On the other hand, using transfer learning achieved only an mAP of 29.54% while 29.17% mAP in a multi-scale image.
Full text available upon request to the author
Article title: Physiological-Based Smart Stress Detector using Machine Learning Algorithms
Authors: Marife Rosales, Argel Bandala, Ryan Rhay P. Vicerra, Elmer P. Dadios
Conference title: 2019 IEEE 11th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM), November 2019
Abstract:
This paper is focused on the development of an intelligent system to identify if one person is stress or not stress using physiological parameters through machine learning. In this study, the dataset was acquired from three hundred (300) male and female participants ages 18 to 25. The gathered dataset is composed of five (5) features (i.e. heart rate, systolic blood pressure, diastolic blood pressure, galvanic skin response and gender). An intelligent system was developed using machine learning algorithms for classification such as Linear Regression (LR), K-Nearest Neighbor (KNN), and Support Vector Machine (SVM) using Python IDE with sci-kit learn machine learning libraries. Google Colaboratory (Colab) was utilized to perform optimization using Gridsearch to identify the best parameters of each algorithm. Feature selection methods are implemented to identify the most significant features related to stress condition of one person. After optimization, the results showed that SVM has the best performance to classify if one person is stress or not stress with optimized training-testing accuracy score of 95.00% - 96.67%.
Article title: Genetic Algorithm Based 3D Motion Planning for Unmanned Aerial Vehicle
Authors: Maverick C. Rivera, Jay Robert Del Rosario, Argel Bandala
Conference title: 2019 IEEE 11th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM), November 2019
Abstract:
Development of Unmanned Aerial Vehicle (UAV) is now a popular field in research. In most of its applications, a pathfinding algorithm is needed in order to find the optimal path and avoid obstacles. In this paper, a genetic algorithm is implemented in order to determine the optimal path for a UAV that will avoid obstacles along the way. The genetic algorithm implemented uses variable-length chromosomes to solve the problem. The results of the simulation of the system yield an average of 29 generations and avoided 53, 500 collisions to find the best path.
Full text available upon request to the author
Article title: A Soft Robotic Tentacle Robot Arm for Inspection Sytem on Manufacturing Lines
Authors: Tristan Joseph C. Limchesing, Then Anjerome Bedruz, Argel Bandala, Nilo T. Bugtai, et al.
Conference title: 2019 IEEE 11th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM), November 2019
Abstract:
With the emerging sector of new industrial age is a foot, innovations are made every day in order to cope with new technologies. Maintenance and regular repairs are key roles in keeping the industry as efficient as possible. With maintenance, safety is paramount in order to prevent accidents and further delays in production. In order to tackle issues for inspecting hazardous and tight spaces, this study utilizes soft robotics technology in aiming to achieve an efficient and cost-effective means to safely inspect tight and hazardous spaces. Soft Robotics is a field in robotics that specializes in materials that are flexible and elastic. Their movements mimic movements that are often found in nature. The soft robotic arm that will be accomplished in this study is a soft robotic tentacle arm with a mounted camera for inspection. This is especially good in reaching places that have limited spaces. The soft robotic arm will be actuated pneumatically and will have an electronic pneumatic microcontroller for its activation. A minicamera will also be mounted on the tip of the robotic arm for the inspection system. With the results and data gathered, it shows that this system can effectively maneuver using its electronic pneumatic controllers. The soft robot arm is also stable enough for the mini camera to be mounted on.
Full text available upon request to the author
Article title: Optimization of Vehicle Classification Model using Genetic Algorithm
Authors: Cyril Dale L. Cero, Edwin Sybingco, Allysa Kate M. Brillantes, Argel Bandala, et al.
Conference title: 2019 IEEE 11th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM), November 2019
Abstract:
This paper focuses on classifying vehicle types into car, van, motorcycle, bus, light truck, multi-axle truck and determine its class based on the Philippine Toll Regulatory Board's vehicle classification. This study utilized DEvol, an open-source tool that uses genetic algorithm for evolving number of filters and nodes, optimizer, activation, dropout rate. The model attained the best accuracy with 78.53% using 9000 images from MIO-TCD dataset.
Article title: Blended Learning Approach to Teaching Oral Communication: Video Blogging in ESL Classroom
Authors: Dylyn A. Junio and Argel Bandala
Conference title: 2019 IEEE 11th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM), November 2019
Abstract:
With the rapid advancement of modern technology, the restructuring of pedagogical teaching is necessitated. Blended learning is one of the recent approaches to teaching which gained massive popularity. In recent years, many studies have been conducted in English teaching instruction; however, research on its application to English as a second language (ESL) oral communication instruction is scarce. The present study reports on the effectiveness of blended learning using video blogs to ESL's oral communication skills. The mixed-methods design was used to collect quantitative and qualitative data. To determine the impact of a blended learning model to ESL's oral communication ability, pretest and posttest were administered to measure improvement. The results of the quantitative data revealed that the implementation of blended learning using video blogs improved students' oral proficiency in terms of pronunciation, fluency, syntax, lexical range and general use of the English language. The qualitative data, on the other hand, indicated the positive attitude of students toward the blended learning experience.
Full text available upon request to the author
Article title: Hazard Classification of Toluene, Methane and Carbon Dioxide for Bomb Detection Using Fuzzy Logic
Authors: Dailyne Macasaet, Argel Bandala, Ana Antoniette Cabantus Illahi, Elmer P. Dadios, et al.
Conference title: 2019 IEEE 11th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM), November 2019
Abstract:
This paper intends to explore bomb detection technology by employing fuzzy logic in classifying Toluene, Carbon dioxide and Methane which are commonly used gases in bombs and other flammables. This research uses Matlab Fuzzy Logic toolbox in classifying gases into three hazard classifications- Safe, Hazardous, and Deadly based on gas concentration and exposure time. Provided in this paper are the standard gas levels which are considered safe to human with respect to exposure time. The output of the classification will help develop a more improved and accurate bomb detection system which is of great importance in today's world.
Full text available upon request to the author
Article title: Analysis of Big Data Technologies for Policy Building in the Philippines
Authors: Rhen Anjerome Bedruz, Ronnie Concepcion II, Argel Bandala, Ryan Rhay P. Vicerra, et al.
Conference title: 2019 IEEE 11th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM), November 2019
Abstract:
Big data is one of the rising technologies in the world today. This is due to the fact that data in general is very useful because there are tons of information to be obtained out of it. This paper analyzes and discusses the rise of big data as one of the most important technologies today and actions that the Philippines can do about it. Current trends in the global scale and in the Philippines are also discussed. Subsequently, the general applications of big data and the how the Philippines adapt the technology are analyzed. Most importantly, its impact to the society, economy and the industry is examined. From this, policy recommendations are given which would help the nation adapt this technology for nation building.
Full text available upon request to the author
Article title: A Fuzzy Logic Approach for Fish Growth Assessment
Authors: Jo-Ann Magsumbol, Vincent Jan Dela Cruz Almero, Marife Rosales, Argel Bandala, et al.
Conference title: 2019 IEEE 11th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM), November 2019
Abstract:
Smart aquaculture is making a name these days due to an ever-increasing demand for an alternative source of protein, fatty acids, vitamins, minerals and essential nutrients, which make it superior over animal meat. To address the rising demand for healthy source of meat, aqua farmers adapt methods wherein they can increase the fish supply all year round. This paper makes use of the fuzzy logic system to identify the current growth stage of carp fish in the pond. The output of the system will be used as an actuator for the feeder system in the aquafarm. Results show that the system successfully identified the current status of the fish in the study.
Full text available upon request to the author
Article title: An Aquaculture-Based Binary Classifier for Fish Detection using Multilayer Artificial Neural Network
Authors: Vincent Jan Dela Cruz Almero, Ronnie Concepcion II, Marife Rosales, Ryan Rhay P. Vicerra
Conference title: 2019 IEEE 11th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM), November 2019
Abstract:
Fish detection, a specific task in computer vision system for fish monitoring, is challenging due to the complex characteristics of the captured images. A proposed approach in tackling this challenging task was to incorporate a multilayer artificial neural network to a computer vision system algorithm, implemented in aquaculture. This computer vision system algorithm captured the images from the aquaculture setup. Then, these captured images were processed. After that, the features out of these processed images were extracted and utilized to develop this multilayer artificial neural network. The best configuration, which is trained with the least learning time and tested with least mean square error and highest accuracy, was determined by adjusting the number of neurons in the two hidden layers. The multilayer artificial neural network with 50 neurons in the first hidden layer and 10 neurons in the second layer was considered the best configuration; it has achieved learning time of 3.374 ms, mean square error of 0.2315, and accuracy of 79.00%, hence, proving the competitiveness of this approach.
Full text available upon request to the author
Article title: Fuzzy Classification Approach on Quality Deterioration Assessment of Tomato Puree in Aerobic Storage using Electronic Nose
Authors: Ronnie Concepcion II, Argel Bandala, Then Anjerome Bedruz, Elmer P. Dadios
Conference title: 2019 IEEE 11th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM), November 2019
Abstract:
Food safety heavily deals with food spoilage that may yield food poisoning. Tomato-based dishes have different shelf-life leading to unique acceptable standards for a person in determining the food condition, and sometimes misclassification due to confusion. To address this problem, a proposed solution is the development of an intelligent electronic nose (eNose) system that will discriminate the condition of tomato puree using fuzzy logic. This system is composed of two sections: the development of electronic nose using Gizduino microcontroller and Mĭngăn Qǐ lai (MQ) gas sensors, and the implementation of fuzzy logic system for classification of food condition. Fuzzy logic resembles human reasoning that yields definite output based on ambiguous input. The collection data rate was set to 2 Hz for tomato puree-emitted gas samples with varying shelf life considering outdoor aerobic storage. Combined Min-Max method and Mamdani inference system was used for the inference engine, and centroid method for defuzzification. The system classifies the tomato puree sample as not spoiled, partially spoiled, and spoiled. The smellprint of each food condition was generated and the tomato puree-spoilage determinant parameters were characterized. Through embedded fuzzy logic, an accuracy of 90.00 % was yielded for tomato puree quality deterioration classification. The developed mechanism is a potential application in domotics.
Full text available upon request to the author
Article title: Motion Planning of a Robotic Arm using an Adaptive Linear Interpolation Crossover and Variable-Length Move Sequence Genome
Authors: Dino Dominic Forte Ligutan, Jason Española, Argel Bandala, Alexander Co Abad, et al.
Conference title: 2019 IEEE 11th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM), November 2019
Abstract:
The ability to perform robotic arm motion planning is a necessity in the design of autonomous and intelligent robotic systems. Motion planning allows the autonomous robotic arm to maneuver its end-effector in an unstructured environment whilst avoiding obstacles on the workspace. This ability is particularly important in processes with pick-and-place operations and varying object positions. In this study, a genetic algorithm-based motion planning for a 4-DOF robotic arm was developed. The developed genetic algorithm operates on a variable-length genome that consists of changes in joint angles. These changes in joint angles represent the end-effector's move sequence. The results show that adaptive linear interpolation crossover (ALIX) improves the convergence of the motion path towards minimization of end-effector error and path length. On average, the end-effector error is 1.4 mm with a maximum path length deviation from a straight line of about 50.4 mm tested on extreme target points. Testing with obstacles present in the workspace shows the ability of the algorithm to generate solution paths to avoid them as well.
Full text available upon request to the author
Article title: Soil Nutrient Detection using Genetic Algorithm
Authors: John Carlo Velasco Puno, Then Anjerome Bedruz, Allysa Kate M. Brillantes, Argel Bandala, et al.
Conference title: 2019 IEEE 11th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM), November 2019
Abstract:
Genetic Algorithm is the method used in this study in classifying the qualitative level of the soil nutrients. The data set includes images coming from the result of the soil testing. The extracted features were the HSV values and the LAB values color space. Out of the six extracted features from the data set, the B from LAB color space is the most linear so with that, it is the input of genetic algorithm in identifying the qualitative level of the soil nutrients. For the run of the program using python programming language and pyCharm CE as IDE, the values of each parameters follow: the population size is 10, mutation rate is 0.01, the number of cross over points is 2 and the maximum number of generations is 1000. The population's final best fitness has 98.2609% that proves that Genetic Algorithm is an effective method in classifying the qualitative level of the soil nutrients.
Article title: Philippine License Plate Character Recognition using Faster R-CNN with InceptionV2
Authors: Mari Christine E. Amon, Allysa Kate M. Brillantes, Ciprian D. Billones, Robert Kerwin Dela Cruz Billones, Argel Bandala, et al.
Conference title: 2019 IEEE 11th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM), November 2019
Abstract:
This research proposes a method for automatic license plate recognition (ALPR) using the Faster R-CNN with InceptionV2 feature extractor that works in the Philippines. While there exist character recognition systems, there still remains difficulty in recognition due to different variations of Philippine license plates. By training a deep neural network in the extraction of the features in images of the different types of Philippine license plates - 1981, 2003, 2014, and others - our proposed multi-class detection system can recognize the alphanumeric characters in the license plate images. The system was tested on actual traffic images in the Philippines that contains different types of license plates, and achieved the detection rate of 90.011%, recognition rate of 93.21% and an overall accuracy of 83.895%.
Article title: Neural Network Utilization for Flagged Words Detection thru Distinct Audio Features
Authors: Matt Ervin Gatchalian Mital, Herbert Villaruel, Rommel Lim, Rogelio Ruzcko Tobias, Argel Bandala, et al.
Conference title: 2019 IEEE 11th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM), November 2019
Abstract:
This research paper employed a method of detecting a given flagged word that would possibly trigger a machine and at the same time, being able to separate such sound source in a given real world environment. As part of the experimentation done, the flagged words were recorded by 3 different individuals. To make sure that only the flagged words would be detected by the robot’s auditory signal processor, the 3 individuals were also asked to record random words that would be used to test whether the robot’s detector responds even in random words being heard. By utilizing the neural networks concepts and processes, detection of flagged words was made possible. After the results has been produced, the researchers arrived to a conclusion that even in the middle of a noisy and reverberant surroundings and situations, the robot can capture the flagged words coming from the crowd by allowing the neural network to perform its function.
Article title: Solving 3D Coverage Problem using Genetic Algorithms in Wireless Camera-Based Sensor Network Modelling
Authors: Neil Oliver M. Velasco, Jay Robert Del Rosario, Argel Bandala
Conference title: 2019 IEEE 11th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM), November 2019
Abstract:
Wireless camera sensor networks are used mostly in research today. However, the camera location in a space of study is a problem in maximizing the coverage of the camera. Instead of being a 2D computation, this computation is modeled in 3D projected to 2D walls. This research makes use of Genetic Algorithms - a search optimization algorithm to find the best placement of camera which will yield to a maximum coverage ratio. The results of the experiment show the length of time the algorithm computed, and the obtained the least number of cameras needed for the most coverage.
Full text available upon request to the author
Article title: Unmanned Aerial Vehicle (UAV) Attitude Estimation Using Artificial Neural Network Approach
Authors: Marc Francis Q. Say, Edwin Sybingco, Argel Bandala, Ryan Rhay P. Vicerra, et al.
Conference title: 2019 IEEE 11th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM), November 2019
Abstract:
There is a growing interest in Unmanned Aerial Vehicles (UAV) which are used in various applications such as cinematography, security, entertainment, and research and development. For a UAV to be able to these applications, stability is a vital aspect. Inertial Measurement Unit (IMU) which is composed of accelerometers, and gyroscopes, and separate magnetometer give data for the attitude position of the UAV to be known and maintain a steady flight. Attitude estimation can be done by various techniques such as using an Extended Kalman Filter (EKF) to predict and estimate angular positions based on the sensor data. In this paper, an Artificial Neural Network (ANN) approach is used to estimate the angular positions as an option for the EKF. A nonlinear autoregressive with exogenous inputs (NARX) is used to create the attitude estimation to investigate the performance compared to the EKF.
Full text available upon request to the author
Article title: Estimation of Triangular Greenness Index for Unknown PeakWavelength Sensitivity of CMOS-acquired Crop Images
Authors: Anton Louise De Ocampo, Argel Bandala, Elmer P. Dadios
Conference title: 2019 IEEE 11th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM), November 2019
Abstract:
Notable works on the use of the triangular greenness index (TGI) to estimate vegetation fraction of croplands or chlorophyll content of crops, proved that relevant metrics on crop health monitoring can be derived from images at the visible spectrum. However, the performance of the TGI-based metric in crop health monitoring greatly depends on knowledge of wavelength sensitivities of the CMOS sensors used to obtain the RGB images of the crop. This becomes a problem when generic digital cameras are used and the specifications of the CMOS sensors are not available. The proposed method in this study compensates for the lack of information on the peak wavelength sensitivities of generic CMOS sensors by performing a parametric sweep on the proportions of 670nm-, 550nm-, and 480nm-peak wavelengths to derive a TGI equation normalized by the green signal. This allows the use of any available digital cameras even without prior knowledge of the wavelength sensitivity at the visible spectrum of the installed CMOS sensors.
Article title: Moving Particle Semi-Implicit Method for Control of Swarm Robotic Systems
Authors: Joseph Aldrin Chua, Laurence A. Gan Lim, Gerardo Lumagbas Augusto, Argel Bandala, et al.
Conference title: 2019 IEEE 11th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM), November 2019
Abstract:
The advancements of Multiple Robot Systems (MRS) have shown advantages over single robot systems. Generally, the motivations for the development of MRS are task flexibility, time efficiency, and single-point failure resiliency. The challenge in MRS, however, is the control and coordination of all the members in the system when performing tasks. Swarm robotics is a branch of MRS that deals with groups of homogeneous robots. The goal of swarm robotics is to produce systems that are scalable, flexible, and robust. The control of swarm robotic systems, however, looks to be one of the main challenges. These control concepts are inspired by biological swarms and, more recently, physics concepts. The success of the swarm's control algorithm will also lead to the swarm's ability to perform cooperative tasks. The use of homogeneous robots in swarm systems makes it advantageous to model the swarm robots as particles in a fluid. The Moving Particle Semi-Implicit (MPS) Method, a particle-based method in fluid dynamics, is proposed to be used as a control algorithm for swarm robotics.
Full text available upon request to the author
Article title: Implementation of a Closed Loop Control System for the Automation of an Aquaponic System for Urban Setting
Authors: Alec Zandra Mae H. Ambrosio, Lanz Harvey M. Jacob, Lea Anne R. Rulloda, Argel Bandala, et al.
Conference title: 2019 IEEE 11th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM), November 2019
Abstract:
An aquaponic system addressing the limited farming space available in condominiums by using a 43cm by 60cm by 80cm stackable design was implemented. The aquaponic system performs the monitoring and control of environmental parameters and data display. The monitoring system also observes the air and water temperature, humidity, pH, and water level. Based on the data gathered from these parameters, Arduino microcontroller determines the necessary output response such as lighting, fish feeding, mist making, and water circulation. Moreover, LED grow lights are used for faster growth rate of plants. Lastly, The data from the sensors, actuators, and growth monitoring system are logged in a micro sd card through a micro sd card module for further analysis.
Article title: Use of Artificial Neural Network in the Estimation of Detonation Velocity for Tetranitromethane-Nitrobenzene Mixture
Authors: Danielle Grace Evangelista, Ryan Rhay P. Vicerra, Argel Bandala
Conference title: 2019 IEEE 11th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM), November 2019
Abstract:
Detonation velocity or rate of energy release is an important property to consider when rating an explosive. It is a critical parameter used for estimating explosive performance as it can indicate the intensity of detonation. The purpose of this research study is to propose an artificial neural network model that would aid in the estimation of detonation velocities of a high explosive specifically, tetranitromethane-nitrobenzene (TNM/NB) mixture, with varying parameters.
Full text available upon request to the author
Article title: Performance Comparison of Classification Algorithms for Diagnosing Chronic Kidney Disease
Authors: Justin De Guia, Ronnie Concepcion II, Argel Bandala, Elmer P. Dadios
Conference title: 2019 IEEE 11th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM), November 2019
Abstract:
Chronic kidney disease (CKD) is one of the diseases with high mortality rate. It is a disease resulted from kidney function loss over a long period of time. The disease shows no symptoms during initial stage. When left not medicated, a person may suffer from other complications such as high blood pressure, anemia, malnutrition, increased risk of cardiovascular disease, cognitive impairment and impaired physical function. Automated diagnosis by using classification algorithms has been an interest of researchers. In this study, six machine learning algorithms were used for classification and its prediction performance was compared based on training time and F1 score, with and without hypertuning the parameters. Of all the six algorithms, KNN has the best F1 score of 0.992248 and minimal training time of 46.999ms. The performance of decision trees was improved with hypertuning, having a F1 score from 0.96 to 0.99. Overall, machine learning algorithms are significant tool to assess chronic kidney disease.
Article title: Implementation of Inverse Kinematics for Crop-Harvesting Robotic Arm in Vertical Farming
Authors: Sandy C. Lauguico, Ronnie Concepcion II, Dailyne Macasaet, Argel Bandala, et al.
Conference title: 2019 IEEE 11th International Conference on Cybernetics and Intelligent Systems (CIS) and IEEE Conference on Robotics, Automation and Mechatronics (RAM), November 2019
Abstract:
The world population is expected to increase to 9.8 billion in 2050 according to United Nations. With this, scarcity of food and space will further be a major concern. This study proposes a framework which used initializing, processing, and directing applied to an inverse kinematics based robotic arm. An automatized approach in addressing the foreseeable problem on providing nutritional plant-based food considering that cities are becoming highly-urbanized was developed. Wall gardening used for vertical farming or urban farming is a technique by which there are sets of rows and columns of pockets installed over a wall. These pockets are filled with soil or other planting bases (i.e. water for hydroponics) for the seedlings to grow. A robotic arm is manually set to point on a specific pocket where a crop has grown. Using inverse kinematics, the set points determine the joint angles. This then targets the pockets and the end-effector of the robot arm performs a grip to the roots of the crops. The robotic arm then moves to its initial point, technically pulling up the crop. After positioning to the initial point, the arm directs to the side of the wall, where a container is located. The end-effector opens to drop the crop carefully into the container. The research study is simulated using MATLAB and Universal Robots. The results show that it can only yield 85.42% of the crops.
Full text available upon request to the author
Article title: Quality Assessment of Mangoes using Convolutional Neural Network
Authors: John Carlo Velasco Puno, Robert Kerwin Dela Cruz Billones, Argel Bandala, Elmer P. Dadios, et al.
Conference title: 2019 IEEE 11th International Conference on Cybernetics and Intelligent Systems (CIS) and IEEE Conference on Robotics, Automation and Mechatronics (RAM), November 2019
Abstract:
The Philippines is one of the countries in the world known for exporting good quality crops. Mangoes in the Philippines are very popular for its good sweet taste and considerably one of the best. Hence, ensuring the quality of the crop to be exported is essential. The study focused on utilizing convolutional neural network in determining the quality of carabao mango (Mangifera Indica). To make sure that all sides of the mango is going to be considered for the quality assessment, a mechanical system that uses conveyor belt, rollers, and camera was used to gather videos for training and validation of the model. The videos were extracted into frames and gone through image processing to remove the background and retain the mango only. The dataset is composed of different mangoes having both good and bad qualities. The implemented model used a total of 5550 training samples with 94.99% accuracy and a total of 2320 samples used for validation with an accuracy of 97.21%.
Article title: Optimization of Extracted Features from an Explosive-Detecting Electronic Nose Using Genetic Algorithm
Authors: Jason Española, Argel Bandala, Ryan Rhay P. Vicera, Elmer P. Dadios
Conference title: 2019 IEEE 11th International Conference on Cybernetics and Intelligent Systems (CIS) and IEEE Conference on Robotics, Automation and Mechatronics (RAM), November 2019
Abstract:
The use of an electronic nose in detecting explosives has gained attention among researchers. This paper aims to optimize the extraction of features generated from a predetermined explosive-detecting electronic nose setup by using a genetic algorithm. A genetic algorithm (GA) is used to minimize the errors such as the mean error within explosive types, the mean error between explosive types and the classification error. The GA optimization program is implemented for each feature extraction technique, namely, principal component analysis (PCA) and linear discriminant analysis (LDA). As a result, the proponents were able to optimize the extracted features into a single point that can truly classify each explosive type. PCA is more preferred than LDA for practical purposes.
Full text available upon request to the author
Article title: Tomato Fruit Image Dataset for Deep Transfer Learning-based Defect Detection
Authors: Robert de Luna, Elmer P. Dadios, Argel Bandala, Ryan Rhay P. Vicerra
Conference title: 2019 IEEE 11th International Conference on Cybernetics and Intelligent Systems (CIS) and IEEE Conference on Robotics, Automation and Mechatronics (RAM), November 2019
Abstract:
Tomato is considered as one of the vegetable crops with highest demand in the Philippines. Job of the farmers does not end after harvesting since the harvested tomatoes needed to be sorted according to its size. Manual sorting is the most widely recognized strategy in sorting but is very dependent on human interpretation and thus, very prone to error. This research proposed a solution that provides sorting of tomato fruit by detection of presence of defect. The study presented the generation of image dataset for a deep learning approach detection of defects based from a single tomato fruit image. Models were implemented using OpenCV libraries and Python programming. A total of 1200 tomato images classified as no defect and with defect are gathered using the improvised image capturing box. These images are used for the training, validation, and testing of the three deep learning models namely; VGG16, InceptionV3, and ResNet50. From this, 240 images are used as testing images to assess independently the performance of the trained models using accuracy and F1-score as performance metrics. Experiment results shown that VGG16 has 95.75-95.92-98.75 training-validation-testing accuracy percentage performance, 56.38-59.24-58.33 for the InceptionV3 model, and 90.58-58.46-64.58 for the ResNet50. Comparative analysis revealed that VGG16 is the best deep learning model to be used in the detection of presence of defect in the tomato fruit based from the dataset gathered.
Full text available upon request to the author
Article title: A Vision-Based Detection and Tracking Algorithm for a Child Monitoring Robot
Authors: John Anthony Cheng Jose, Justine Veronica Basco, Jomar Kenneth Jolo, Argel Bandala, et al.
Conference title: 2019 4th Asia-Pacific Conference on Intelligent Robot Systems, July 2019
Abstract:
Accidents have been found to be one of the leading causes of both fatal and non-fatal injuries to children. Though some accidents that occur are often unavoidable, more often than not these injuries can be prevented by giving the child proper attention. The researchers intend to address certain gaps in stationary monitoring solutions by adding abilities such as an insured way of continuously monitoring the test subject and a real time notification feature to a mobile spherical robot. This research presents the software division of a technological solution to child monitoring by developing a computer vision algorithm for following and monitoring children indoors utilizing an RGB-D camera. This algorithm will work hand in hand with a hardware design of a spherical robot that utilizes microcontrollers, RFID technology and GSM system. An Android application will also be created to provide the users the means of manually overriding the spherical robot, color calibration and location indicator as a part of the robot's notification system. The detection and tracking ability of the algorithm is tested by using objects with varying characteristics. The autonomous navigation testing of the robot is performed at two controlled test setups: living room and child's playroom.
Full text available upon request to the author
Article title: Spherical Mobile Robot for Monitoring and Tracking Children Indoors
Authors: John Anthony Cheng Jose, Justine Veronica Basco, Jomar Kenneth Jolo, Argel Bandala, et al.
Conference title: 2019 4th Asia-Pacific Conference on Intelligent Robot Systems, July 2019
Abstract:
Families around the world continue to suffer the loss of a child due to unintentional injuries caused by accidents that could have been prevented. Stationary monitoring solutions are widely used to aid in the prevention of such situations. However, these technologies present certain gaps that the researchers would like to address by adding a real time notification ability and an ensured way of continuously monitoring by making sure that the test subject will never be lost by the intended solution. This research paper presents the hardware division of a technological solution to child monitoring by developing a semi-autonomous spherical robot to follow a child as the subject moves throughout the room. The spherical robot would have the ability to manually navigate around two controlled test setups: living room and child's playroom. The robot would also be able to distinguish designated safe zones and danger zones with the help of the RFID technology. The real time notification ability will be highlighted by giving the robot the feature of sending SMS messages to the subject's parent or guardian indicating the time and place of where the child last exited. The manual navigation was tested with the use of two controlled test setups and the notification system utilizing the RFID technology was tested thirty times in six various places having different signal strengths ranging from -50 dBm to -120 dBm.
Full text available upon request to the author
Article title: Categorizing License Plates Using Convolutional Neural Network with Residual Learning
Authors: Argel Bandala, John Anthony Cheng Jose, Jose Martin Maningo, et al.
Conference title: 2019 4th Asia-Pacific Conference on Intelligent Robot Systems, July 2019
Abstract:
Like other countries, the Philippines uses various license plate standards wherein some purely text while some are hybrid graphic-text plates. And to harness its generalizability, this study developed a classification algorithm utilized as a pre-processing scheme for the multi-standard license plate. With an input image captured at a different perspective, it was feed into the neural network and classify as Rizal monument series (2001 base and 2003 base), 2014 series and conduction sticker for new vehicles. In total, there are 303 different images captured for this study. Around 100 conduction sticker images, 103 Rizal Monument images, 100 black and white images. Furthermore, this study focused on using transfer learning technique, wherein a trained network utilized, then only the last layer was reset and retrained on the new dataset. To measure the performance of the classification model and optimized it cross-entropy and stochastic gradient descent was employed respectively at a learning rate of 0.001 and reduced by 10 for every seven (7) epochs. The progression of accuracy results in increasing the epochs, and for the 25 epochs, the training completed in 4 minutes and 7 seconds with the best validation accuracy of 82.61%.
Full text available upon request to the author
Article title: Control and Mechanical Design of a Multi-diameter Tri-Legged In- Pipe Traversing Robot
Authors: Argel Bandala, Jose Martin Maningo, Arvin Fernando, et al.
Conference title: 2019 IEEE/SICE International Symposium on System Integration (SII), January 2019
Abstract:
In this paper a versatile and adaptive in-pipe robot is designed and tested The existing problem of versatility and adaptability of in pipe inspection robots are addressed in this study. The robot is equipped with a screw type assembly, which uniformly contracts and retracts a tri-arm assembly of wheels connected to it. This mechanism ensures that the robot grips the pipe walls when the arms are expanded. The dynamic model of the robot is derived and implemented in a proportional integral controller. The robot can maintain vertical position by maintaining the force exerted on the screw system Simulations and experiments were conducted to determine the robustness and stability of the robot system In addition, the robot is also capable of rust mapping, which enables easier pipe monitoring. The rust mapping yielded an S3.2% success rate.
Full text available upon request to the author
Article title: Formation-based 3D Mapping of Micro Aerial Vehicles
Authors: Mark Lester F. Padilla, Pakpong Chirarattananon, Argel Bandala, et al.
Conference title: 2019 IEEE/SICE International Symposium on System Integration (SII), January 2019
Abstract:
Micro Aerial Vehicles have brought tremendous interests to the research community, particularly in localization and mapping. While there are many commercially available sensors, such as Laser Range Finders (LRF) and RGBD cameras, that provide accurate 3D maps, they usually have significant power and payload requirements. This means, small flying robots are unable to handle such sensors. This study explores the possibility of collaborative mapping using formations from multiple simple cameras to obtain an accurate map similar to that of the LRF and RGBD cameras. By using multiple small robots and integrating them as one, we have created a platform for 3D reconstruction in which formations can be incorporated. Thus, the proposed method can be used with a low-cost system for surveying, disaster management, and surveillance in the future.
Full text available upon request to the author
Article title: DeepTronic: An Electronic Device Classification Model using Deep Convolutional Neural Networks
Authors: Argel Bandala, Rodolfo C. Salvador, Irister M. Javel, et al.
Conference title: 2018 IEEE 10th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment and Management (HNICEM)
Abstract:
This paper presents a novel and straightforward way of classifying discrete and surface-mount electronic components found on electronic prototypes using transfer learning and deep convolutional neural networks (DCNN). The goal of this study is to precisely classify images of electronic components into six classes: resistor, capacitor, inductor, transformer, diode, or integrated circuit. Each class of electronic components has over 100 images which are augmented and preprocessed to match the input layer requirements of the deep learning models used. The dataset was divided into a ratio of 70:30, where 70% was used for training and 30% was used for testing and validation. Transfer Learning (TL) was done using three pre-trained deep learning models that are available on MATLAB's Neural Network Toolbox: Inception-v3, GoogleNet, and Resnet101. Using this approach provides faster deployment and only requires fewer lines of coding compared to typical deep learning classification methods which make use of Python, Tensorflow, and Keras. The results of the experiment showed that Inception-v3 has the highest validation accuracy of 94.64% in classifying electronic components.
Full text available upon request to the author
Article title: Vehicle-Pedestrian Classification with Road Context Recognition Using Convolutional Neural Networks
Authors: Robert Kerwin Dela Cruz Billones, Argel Bandala, et al.
Conference title: 2018 IEEE 10th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment and Management (HNICEM)
Abstract:
In road traffic scene analysis, it is important to observe vehicular traffic and how pedestrian foot traffic affects the over-all traffic situation. Road context is also significant in proper detection of vehicles and pedestrians. This paper presents a vehicle-pedestrian detection and classification system with road context recognition using convolutional neural networks. Using Catch-All traffic video data sets, the system was trained to identify vehicles and pedestrians in four different road conditions such as low altitude view T-type road intersection (DS0), mid-altitude view bus stop area in day-time (DS4-1) and night-time (DS43) condition, and high-altitude view wide intersection (DS31). In the road context recognition, the system was first tasked to identify in which of the four road conditions the current traffic scene belongs. This is designed to ensure a high detection rate of vehicles and pedestrians in the mentioned road conditions. Road context recognition has 98.64% training accuracy with 2800 sample images, and 100% validation accuracy with 1200 sample images. After road context recognition, a detection algorithm for vehicle and pedestrians was trained for each condition. In DS0, the training accuracy is 97.75% with 1200 image samples, while validation accuracy is 94.75% with 400 image samples. In DS3-1, the training accuracy is 98.63% with 1400 image samples, while validation accuracy is 98.29% with 600 image samples. In DS4-1, the training accuracy is 99.43% with 1400 image samples, while validation accuracy is 99.83% with 600 image samples. In DS4-3, the training accuracy is 97.77% with 1400 image samples, while validation accuracy is 98.29% with 600 image samples.
Article title: Smart Wound Dressing with Arduino Microcontroller
Authors: Argel Bandala
Conference title: 2018 IEEE 10th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment and Management (HNICEM)
Abstract:
Wound dressing are necessity in wound management and quick healing. Commonly used dressings are simple and affordable, but healing may not result to optimum healing. Moisture must be maintained and should be replaced when moisture is no longer present. This paper designed a moisture monitoring wound dressing using Arduino microcontroller since there is no cost effective, biocompatible, and mass manufacturable wound dressing that can monitor conditions continuously while keeping foreign pathogens out at the same time [1]. Using biocompatible materials to make sensors, physicians will able to track the status of the wound through one or many variables including temperature, pH, moisture level, oxygen level etc. The status could be accessed through mobile phones using wireless connectivity through Bluetooth.
Full text available upon request to the author
Article title: A Robotic Model Approach of an Automated Traffic Violation Detection System with Apprehension
Authors: Argel Bandala, Then Anjerome Bedruz, et al.
Conference title: 2018 IEEE 10th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment and Management (HNICEM)
Abstract:
This study suggests a robotic model approach to create a sample traffic violation detection system with apprehension scenario before such system can be implemented in a real road. The model used two robots, one for the moving object which will be detected by the camera and one for the robot that will follow the moving object if its speed reaches a certain limit. The captured images from the camera were fed to an algorithm which detects the centroid of the moving object to track its speed, thereby deciding if it is moving beyond the reference speed. The result of this algorithm was fed to the tracker robot, which then mobilizes and follows the moving object when the moving object exceeds the speed limit.
Article title: Application of Artificial Neural Networks in prediction of pyrolysis behavior for algal mat (LABLAB) biomass
Authors: Argel Bandala, Andres Philip Mayol, et al.
Conference title: 2018 IEEE 10th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment and Management (HNICEM)
Abstract:
Pyrolysis kinetics is one way to produce bio-oil and biochar from a biomass product. It is a method to harvest clean energy from a biomass product. Moreover, kinetics and thermal composition of the biomass product is essential for pyrolysis design and optimization. However, industrial pyrolysis process is up to 200°C/min and lab scale pyrolysis temperature is up to 100°C/min. In this study, data from thermogravimetric analysis (TGA) has been utilized and gathered to provide data on algal pyrolysis kinetics. To predict the pyrolysis kinetics at a heating rate of 200°C/min, artificial neural networks (ANN) has been utilized. Results show that ANN predicted the outcome of pyrolysis kinetics which had a correlation with heating rates (10°C, 25°C, and 50°C) of the sample. This is quantified by the correlation coefficient during training which is 0.9972. The average fit quality of the derived model with respect to the experimental data is 98.51%. This work can be improved by considering other hyperparameters for the neural network. This work can also be extended to other compounds besides lablab biomass.
Full text available upon request to the author
Article title: Threat Object Classification in X-ray Images Using Transfer Learning
Authors: Argel Bandala, Reagan L. Galvez, Elmer P. Dadios, et al.
Conference title: 2018 IEEE 10th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment and Management (HNICEM)
Abstract:
Automatic classification of threat objects in X-ray images is important because of terrorist incidents happening in every country especially in the Philippines. Manual inspection using X-ray machine is prone to human error due limited amount of time given to the operator to check the baggage. This task is also stressful because there are lots of objects to be identified and needs full attention. Over long period of time, the performance of human inspector decreases and the chance of missed detection increases. As a solution to the problem, this paper used the concept of transfer learning in classification of threat objects. The threat objects used in the experiment consists of 4 classes such as blade, gun, knife and shuriken. The dataset came from the GDXray database, a public database of X-ray images. Experiment results showed that by using the concept of transfer learning with data augmentation and fine-tuning, threat objects can be classified at 99.5% accuracy.
Full text available upon request to the author
Article title: Characterization And Effect Of Enhanced Flipped Classroom Implementation
Authors: Argel Bandala and Dylyn A. Junio
Conference title: 2018 IEEE 10th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment and Management (HNICEM)
Abstract:
The advent of digital technology paves the way for the innovative classroom pedagogy, one such innovation in is the use of flipped learning. In the recent years, many researches have been conducted to prove its effectiveness however few studies have been done in the practice of teaching English as a second language in the Philippines. The current study explores flipped classroom instruction to teach oral communication skills to senior high school students. To achieve this, a 6-week flipped classroom instruction was designed to provide students with technology enhanced lessons. The mixed method was used to collect quantitative and qualitative data. Findings of the study revealed the positive impact of flipped classroom instruction to the teaching and learning of oral communication in senior high school.
Full text available upon request to the author
Article title: Design of the Philippine Jeepney for Crashworthiness Analysis: A Finite Element Analysis Approach
Authors: Argel Bandala
Conference title: 2018 IEEE 10th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment and Management (HNICEM)
Abstract:
In this study, the crashworthiness analysis of the Philippine jeepney was successfully demonstrated using the finite element analysis approach. The Philippine jeepneys, or sometimes-called jeeps, are the most popular means of public transportation in the country. They are often known in the country as “King of the Road.” Though commuting via jeepney is the cheapest option, there are a lot of cons. Jeepneys are often mechanically unsound due to their balding tires, crabbing and yawing from distorted subframes with poor emission. The FEA Design Center Facility of the Metals Industry Research and Development Center (MIRDC) of the Department of Science and Technology was able to evaluate the vehicle crashworthiness using computer-aided design (CAD) and FEA models developed in SIEMENS NX. From the simulated impact analysis results, the current jeepney design is not well designed to absorb such crash impact. Thus, resulting in fatal injuries that may cause harm to its passengers.
Full text available upon request to the author
Article title: Coding-based Traffic Warning System Using GSM
Authors: Rhen Anjerome Bedruz, Aaron Parayno Uy, Argel Bandala, et al.
Conference title: 2018 IEEE 10th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment and Management (HNICEM)
Abstract:
This study presents a coding-based traffic warning system using the GSM technology. With a specific traffic density of cars, the implemented system decides and sends a warning message alerting the nearby drivers about the traffic condition along the Vito Cruz Taft Avenue Street. The warning messages were differentiated in 5 categories. The system in particular, was modelled with source encoding (Huffman), and Channel encoding (Hamming), and that the GSM technology was applied thereafter. The traffic warning system modelled was found to have an average compression ratio of 59.37 %, and BER of 0.916 %. These results show that the system is well-suited for real application of traffic warning system as it provided a reliable means of communication.
Full text available upon request to the author
Article title: Multi-Scale Vehicle Classification Using Different Machine Learning Models
Authors: Edison Roxas, Ryan Rhay P. Vicerra, Laurence A. Gan Lim, Argel Bandala, et al.
Conference title: 2018 IEEE 10th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment and Management (HNICEM)
Abstract:
The focus of this paper is to explore multi-scale vehicle classification based on the histogram of oriented gradient features. Several literatures have used these features together with different classification models, however, there is a need to compare different models suited for vehicle classification application. In order to quantify the results a common dataset was used for the machine learning models: logistic regression, k-nearest neighbor, and support vector machine. However, since the classification of the support vector machine is based on the type of kernel (linear, polynomial, and Gaussian) used, additional tests were conducted. Thus, this study provides the following contributions: (1) comparison of machine learning models for vehicle classification; and (2) comparison of the best type of kernel function.
Full text available upon request to the author
Article title: Vision-Based Passenger Activity Analysis System in Public Transport and Bus Stop Areas
Authors: Robert Kerwin Dela Cruz Billones, Edwin Sybingco, Laurence A. Gan Lim, Argel Bandala, et al.
Conference title: 2018 IEEE 10th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment and Management (HNICEM)
Abstract:
This study presents the development of a vision system for passenger activity analysis in public transport and bus stop areas. The vision system used people detection and counting algorithm to track the flow of boarding and alighting passengers in a bus stop area. A fuzzy logic controller used inputs from the vision system to determine boarding frequency and alighting frequency for analysis of bus route and dwell time to avoid long queueing that usually cause traffic congestion. People detection and counting result using DS6 dataset (indoor) have 96.81% accuracy with 97.93% precision. People detection and counting result using DS4-1 dataset (outdoor, bus stop area) have 80.39% accuracy with 87.13% precision. Fuzzy simulation results show a boarding frequency of 22 passengers/minute and alighting frequency of 12 passengers/minute. The vision system also analyzed the boarding and alighting of passengers in no loading and unloading areas. This event usually caused traffic bottleneck due to road blockage and long bus queues. In the analysis of DS4-1 (24-hr length) videos, a total of 212 no loading/unloading violations were recorded.
Article title: Coconut Fruit Maturity Classification using Fuzzy Logic
Authors: Iristed M. Javel, Argel Bandala, Rodolfo C. Salvador
Conference title: 2018 IEEE 10th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment and Management (HNICEM)
Abstract:
The coconut is one of the most useful trees in the world. Its fruit, with the scientific name cocos nucifera, is one of the major agricultural products of the Philippines. The coconut fruit depending on its maturity is used as a food or as a beverage. There are three stages of maturity namely: malauhog, malakanin, and malakatad. The classification into each stage may be based on the color and hardness of its shell so as the amount and tenderness of its meat. To categorize maturity stage, this paper uses fuzzy logic with color and sound as fuzzy inputs. Image color analysis for determining the percentage brown in the shell. Sound spectral analysis for relating the shell hardness and meat condition. Fuzzy inference system for evaluating the relationship of sound and color with the maturity of a coconut fruit. This study is able to simulate coconut fruit maturity classification system using a fuzzy logic approach.
Full text available upon request to the author
Article title: Performance Evaluation of 12Hp 4-stroke Single Cylinder Diesel Engine based on the Philippine Standards
Authors: Jonathan Q. Puerto, Allan John Sala Limson, Fred P. Liza, Argel Bandala, et al.
Conference title: 2018 IEEE 10th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment and Management (HNICEM)
Abstract:
Presently, local demands for single cylinder engines used in agricultural equipment served thru importation. Our country is in complete dependence on other countries concerning supplying the prime movers for its primary source of power. The Department of Science and Technology initiated developing a 12Hp single cylinder diesel engine. And to make it acceptable to the user, performance evaluation of engine was conducted as necessary for their commercial operation. PAES 117:2000 is the basis of assessing the engine performance and tested at starting condition, varying load performance and during a continuous run. Based on the result, the average maximum power was rated 93.9% (8.42 kW). Likewise, the average fuel consumption was 3.15 L/hr. Also, the average continuous power as a percentage of the rated maximum power was 83.6% (7.48 kW). During the continuous running test, the average maximum noise level of the prototypes was 90.6 dB(A). It showed that the engine developed achieved the standard performance conditions and comparable to the leading brands of commercial engines of the same power rating.
Full text available upon request to the author
Article title: Arduino-based Chug Meter With Force Sensing Resistor And Accelerometer
Authors: Gabriela Eustaquio, Colin Velasco, Bernard Chi, Jarred Cheng, Argel Bandala
Conference title: 2018 IEEE 10th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment and Management (HNICEM)
Abstract:
With just about anything and everything being automated or technologized, there remains endless ordinary and mundane objects at easy disposal. An arduino platform was utilized for its many advantages including fast processing and simple interface. The important age-old pub debate of who can chug beer the fastest is to be put to an end with a revolutionary chug meter. To recognize when the beer is lifted up from the coaster to start the timer and when the beer is placed back on the coaster to stop the timer, a force sensing resistor was used. Moreover, the modification of this prototype involves a validation that the beer is being “chugged” or consumed with the incorporation of an accelerometer to sense that the beer is tilted the entire time after it is lifted from the coaster.
Full text available upon request to the author
Article title: Vision System for Soil Nutrient Detection Using Fuzzy Logic
Authors: John Carlo Velasco Puno, Argel Bandala, Elmer P. Dadios, Edqin Sybingco
Conference title: 2018 IEEE 10th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment and Management (HNICEM)
Abstract:
Several methods exists to identify the nutrient content of the soil. The most popular method is by using Soil Test Kit (STK). STK gives soil qualitative level of macronutrients and pH. Chemicals that change color upon reaction with soil samples can determine macronutrients such as nitrogen, phosphorus, and potassium. These chemicals are going to be processed based on the method given by the kit. With the use of different algorithms that is commonly used for classification, mostly, a vision system is required. In this study, the development of the vision system that will capture the image of the soil sample after conducting soil testing will be tackled together with the image processing and feature extraction. Using the extracted features as the input of the fuzzy logic gives accurate result in determining the nutrient level of the soil.
Article title: Automated Image Capturing System for Deep Learning-based Tomato Plant Leaf Disease Detection and Recognition
Authors: Robert de Luna, Elmer P. Dadios, Argel Bandala
Conference title: TENCON 2018-IEEE Region 10 Conference
Abstract:
Smart farming system using necessary infrastructure is an innovative technology that helps improve the quality and quantity of agricultural production in the country including tomato. Since tomato plant farming take considerations from various variables such as environment, soil, and amount of sunlight, existence of diseases cannot be avoided. The recent advances in computer vision made possible by deep learning has paved the way for camera-assisted disease diagnosis for tomato. This study developed the innovative solution that provides efficient disease detection in tomato plants. A motor-controlled image capturing box was made to capture four sides of every tomato plant to detect and recognize leaf diseases. A specific breed of tomato which is Diamante Max was used as the test subject. The system was designed to identify the diseases namely Phoma Rot, Leaf Miner, and Target Spot. Using dataset of 4,923 images of diseased and healthy tomato plant leaves collected under controlled conditions, we train a deep convolutional neural network to identify three diseases or absence thereof. The system used Convolutional Neural Network to identify which of the tomato diseases is present on the monitored tomato plants. The F-RCNN trained anomaly detection model produced a confidence score of 80 % while the Transfer Learning disease recognition model achieves an accuracy of 95.75 %. The automated image capturing system was implemented in actual and registered a 91.67 % accuracy in the recognition of the tomato plant leaf diseases.
Article title: Payload Lift and Transport Using Decentralized Unmanned Aerial Vehicle Quadcopter Teams
Authors: Argel Bandala, Aldrin G. Chua, Ryan R. Dajay, Rafael D. Rabacca
Conference title: TENCON 2018-IEEE Region 10 Conference
Abstract:
This paper presents a decentralized and cooperative load lifting and transportation system using unmanned aerial vehicle quadcopters. The limitation of a single UAV to carry load is addressed in this study by creating a cooperative lifting system that can accommodate varying load weight. Cooperative, independent and scalable agents were implemented with decision making algorithm embedded in each agents. Decentralized sensing of load is done by the UAV and the group consensually decides if another UAV is needed to carry the load. The system can lift different weight by autonomously sending appropriate number of UAV depending on the load. Experiments were conducted to determine the responsiveness of the system in varying load weights. Experiment results showed that the developed system is robust and scalable.
Full text available upon request to the author
Article title: Object Detection Using Convolutional Neural Networks
Authors: Reagan L. Galvez, Argel Bandala, Elmer P. Dadios, Ryan Rhay P. Vicerra, et al.
Conference title: TENCON 2018-IEEE Region 10 Conference
Abstract:
Vision systems are essential in building a mobile robot that will complete a certain task like navigation, surveillance, and explosive ordnance disposal (EOD). This will make the robot controller or the operator aware what is in the environment and perform the next tasks. With the recent advancement in deep neural networks in image processing, classifying and detecting the object accurately is now possible. In this paper, Convolutional Neural Networks (CNN) is used to detect objects in the environment. Two state of the art models are compared for object detection, Single Shot Multi-Box Detector (SSD) with MobileNetV1 and a Faster Region-based Convolutional Neural Network (Faster-RCNN) with InceptionV2. Result shows that one model is ideal for real-time application because of speed and the other can be used for more accurate object detection.
Full text available upon request to the author
Article title: Development of an Adaptive In-Pipe Inspection Robot with Rust Detection and Localization
Authors: Julianne Diaz, Manuel I. Ligeralde, Micah Antoinette B. Antonio, Argel Bandala, et al.
Conference title: TENCON 2018-IEEE Region 10 Conference
Abstract:
In response to addressing the issue of pipe quality checking, the researchers developed an adaptive in-pipe inspection robot that is able to detect rust as well as map the rust on the pipe network. The robot is traversed in a pipe network of horizontal, vertical, elbow, and tee type with diameters of 8, 10 and 12 inches for all. Hence, the test features the versatility, adaptability, and robustness of the robot. The leg expansion of the robot is inspired by the scissors mechanism. On the other hand, rust detection was done through a per pixel classification via image processing. To effectively map the rust, checkpoints were used as a guide of the robot. Testing of the robot were supported in both simulation and actual testing, wherein it yields a 96.45% success rate on the site. Likewise, its rust detection program proved to be successful with a high percentage accuracy of 99.18%. The localization on the other hand yielded an accuracy of 85%. Given the obtained data and results, the researchers were able to go beyond their target objective of 70%.
Full text available upon request to the author
Article title: Detection of Fonts and Characters with Hybrid Graphic-Text Plate Numbers
Authors: Allysa Kate M. Brillanteas, Argel Bandala, Elmer P. Dadios, John Anthony Cheng Jose
Conference title: TENCON 2018-IEEE Region 10 Conference
Abstract:
Philippine license plates have different plate styles and character fonts making the plate character recognition challenging. This paper focuses on improving the segmentation method to recognize characters of different formats of Philippine license plates. The proposed system comprises of license plate classification, character segmentation and character recognition. License plate series was classified using color level of pixels in the image. Plate characters were segmented using 3-Class Fuzzy Clustering with Thresholding and Connected Component Analysis and were recognized using Template Matching. The system achieved an accuracy of 95% and 70% for the 2003 plate series and 2014 plate series, respectively, having tested 20 license plates from each series.
Article title: Design of a Fuzzy-Genetic Controller for an Articulated Robot Gripper
Authors: Jason Española, Argel Bandala, Ryan Rhay P. Vicerra, Elmer P. Dadios
Conference title: TENCON 2018-IEEE Region 10 Conference
Abstract:
In this study, a fuzzy logic controller (FLC) was designed to manipulate an articulated robot gripper. An idea from a previous study was utilized to enhance the performance of the FLC using genetic algorithms by optimizing newly-introduced coefficients in the membership functions of the FLC. The proposed controller was applied on a robot gripper model in Simulink. All in all, the genetic algorithm was able to come up with optimized parameters after an average of at least eight (8) generations and the proposed controller was able to follow the reference trajectory more accurately than the simple fuzzy controller. Further research will be necessary for physical implementation and possible improvement of the utilized genetic algorithm.
Full text available upon request to the author
Article title: Vehicle Classification Using AKAZE and Feature Matching Approach and Artificial Neural Network
Authors: Rhen Anjerome Bedruz, Arvin Fernando, Argel Bandala, Edwin Sybingco, et al.
Conference title: TENCON 2018-IEEE Region 10 Conference
Abstract:
This research proposes a method in order to classify vehicles in a highly congested roads , a robust technique for vehicle classification with low computational power must be used. So, a proposed solution is to embed an AKAZE feature matching extraction which is ran in an artificial neural network will be used. AKAZE was used because it is faster than SIFT. The features extracted from the AKAZE algorithm will be grouped according to the type of vehicle where it was used and be placed to an Artificial Neural Network (ANN) for the training of the network itself. The results yielded good for real-time Vehicle Classification.
Full text available upon request to the author
Article title: Human Gesture Recognition Using Computer Vision for Robot Navigation
Authors: Pocholo James Loresco, Argel Bandala
Conference title: 5th International Conference on Communication and Computer Engineering (ICOCOE’2018)
Abstract:
Robot navigation is one of the significant requirements of human-computer interaction (HCI). Gesture recognition in robots is the control of its movement by gestures priori information. Gesture recognition methods employing wearable technologies are usually not natural and not barrier-free in interaction. Existing computer vision based gesture recognition required full body vector data demanding higher computational complexity. This paper presented hand gesture recognition to navigate a robot using computer vision focused on the hand image only. The system provided hand gesture recognition algorithms to control robot navigation for 4 dynamic gestures, namely 'Go Left', 'Go Right', 'Go Backwards', and 'Go Forward' and 2 static gestures, namely 'Stop' and 'Turn around'. Tests gave a high identification rate for hand gestures. Future work will involve implementing the proposed gesture control along with other sensor technologies and other computer vision algorithms to enable self-localization and positioning.
Article title: Coverage Path Planning on Multi-Depot, Fuel Constraint UAV Missions for Smart Farm Monitoring
Authors: Anton Louise De Ocampo, Argel Bandala, Elmer P. Dadios
Conference title: 2018 IEEE Region 10 Symposium
Abstract:
UAVs used in monitoring crop fields are flying higher than 6 meters and capture telemetric data that provides information on the general condition of the plants in the field. But, in order to obtain specific information on the actual conditions of the plants based on individual morphological aspects, lower altitude monitoring, at most 3 meters, is required. Low-altitude missions cover less than high-altitude and requires UAVs to fly longer to cover more area. In this paper, an approach for multi-depot, fuel constrained coverage path planning is presented. First, target coverage is segmented into smaller regions based on the number of available charging depots. Then, each region is further decomposed into multitude of cells with area equivalent to the camera FOV when UAV is flying at 3 meters above the field. All possible routes are generated and fed into evolutionary optimization in aim to identify the optimal path considering the fuel constraints and availability of recharging depots. The optimization yields a significant improvement in obtaining the route that will provide the minimum distance that the UAV should traverse to cover the entire Area-of-Interest. This approach proved to be useful for crop field monitoring using UAVs.
Full text available upon request to the author
Article title: Development of an Adaptive Pipe Inspection Robot with Rust Detection
Authors: Argel Bandala, Jose Martin Maningo, John Anthony Cheng Jose, Arvin Fernando, et al.
Conference title: 2018 IEEE Region 10 Symposium
Abstract:
In response to addressing the issue of pipe quality checking, the researchers developed an adaptive in-pipe inspection robot that is able to detect rust. The robot is traversed in a pipe network of horizontal, vertical, elbow, and tee type with diameters of 8, 10 and 12 inches for all. Hence, the test features the versatility, adaptability, and robustness of the robot. As for the leg expansion of the robot, it is inspired by the scissors mechanism that is achieved by using of linked, folding supports in a crisscross pattern. In this paper, the traversing of the robot was supported in both simulation and actual testing, wherein it yield a 97.2167% success rate on the site. Likewise, Rust Detection proved to be successful with its high percentage accuracy of 95%. Given the obtained data and results, the researchers were able to go beyond their target objective of 70%.
Full text available upon request to the author
Article title: Vision-based traffic sign compliance evaluation using convolutional neural network
Authors: Edison Roxas, Joshua N. Acilo, Ryan Rhay P. Vicerra, Argel Bandala, et al.
Conference title: 2018 IEEE International Conference on Applied System Innovation (ICASI)
Abstract:
Manual monitoring of road signs compliance procedures are adapted by developing countries. As effective as this method is, the amount of time and funds needed to cover a large area is quite alarming. Thus, a need for a vision - based traffic sign detection and recognition system. However, while a majority of researches using machine vision focuses on the development of a robust real - time traffic sign recognition system, researches addressing the issue of the sign compliance and standardization is lacking.
Full text available upon request to the author
Article title: Vehicle classification method using compound kernel functions
Authors: Edison Roxas, Ryan Rhay P. Vicerra, Elmer P. Dadios, Argel Bandala
Conference title: 2018 IEEE International Conference on Applied System Innovation (ICASI)
Abstract:
The focus of this paper is to explore the use of the Support Vector Machine (SVM) classifier. Though several literatures have already discussed the idea of using this method in vehicle classification, however, SVM accuracy is limited on the type of Kernel function used. Each Kernel functions has their own characteristics and limitations that is highly dependent on its parameter. Thus, in order to overcome these limitations, a method of compounding Kernel function for vehicle classification is hereby implemented.
Full text available upon request to the author
Article title: Cardiovascular health pre-diagnosis based on a BP profile using Artificial Neural Network
Authors: Jackielyn G. Domingo, Sean Harvy S. Geronimo, Gavril Ryan N. Ochoa, Argel Bandala, et al.
Conference title: 2017 IEEE 9th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM)
Abstract:
This study describes the implementation of Artificial Neural Network Pattern Recognition using Back-Propagation algorithm which can perform cardiovascular health pre-diagnosis of a patient through a generated blood pressure profile. The proponents gathered data from institutions that conduct Exercise Stress Testing, specifically the Treadmill Stress Test. The data gathered were age, gender, height, weight, blood pressure and heart rate readings and is considered as blood pressure condition risk factor. They compose the 47 input parameters of the network and was then divided into two - the training data and the testing data. This was put into a database created using Microsoft Excel. The back-propagation neural network model gives an accurate pre-diagnosis. The trained system gives an acceptable pre-diagnosis in reference to the given diagnosis by the attending practitioner that facilitated the collection of data.
Article title: Design, fabrication, and testing of a semi-autonomous wheelchair
Authors: S. Karim, B. D. Que, A. Bandala, J. E. Que, et al.
Conference title: 2017 IEEE 9th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM)
Abstract:
This research project presents a semi-autonomous wheelchair that will move and avoid obstacles autonomously while a remote control will be used to guide the wheelchair in its path to the user. The wheelchair is equipped with two DC motors, batteries, a transmitter-receiver pair (to relate to the remote control), and nine ultrasonic sensors that is controlled by an Arduino microcontroller. The motors used are 24VDC, 250W brushed DC motors, and are independently controlled based on the input provided by the ultrasonic sensors and on-board receiver. The control system implements algorithms in obstacle avoidance for the wheelchair and in the path planning for the remote control. In the testing and performance evaluation, factors such as response time, maneuverability, speed, turning radius, and recommended maximum payload is measured and analyzed.
Full text available upon request to the author
Article title: Development of a text to braille interpreter for printed documents through optical image processing
Authors: Joshua L. Dela Druz, Jonaida Angela D. Ebreo, Reniel Inovejas, Argel Bandala, et al.
Conference title: 2017 IEEE 9th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM)
Abstract:
This paper presents the development of an optical text to braille converter device for aiding visually impaired individuals to read printed materials. This is a solution for the lag or even failure of translating or printing the braille version of everyday reading materials. The system utilized optical character recognition engine in which an image of the text to be translated into braille is captured. The digitized texts are then transferred electronically in a braille haptic device. This device are piezoelectric based haptic system which is composed of several haptic pins arranged in a way to resemble the braille writing system. Several experiments were conducted to determine the performance of the system. The overall system reliability obtained was 95.68%. The system is also capable of processing speed of 1 word in 2 seconds. The system performs at its best with a letter sized page reading material within the range of 15 to 20 cm from the camera, with the camera positioned at 0 degrees.
Full text available upon request to the author
Article title: Path planning for mobile robots using genetic algorithm and probabilistic roadmap
Authors: Robert Martin Cahanding Santiago, Anton Louise De Ocampo, Aristotle Ubando, Argel Bandala, et al.
Conference title: 2017 IEEE 9th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM)
Abstract:
Mobile robots have been employed extensively in various environments which involve automation and remote monitoring. In order to perform their tasks successfully, navigation from one point to another must be done while avoiding obstacles present in the area. The aim of this study is to demonstrate the efficacy of two approaches in path planning, specifically, probabilistic roadmap (PRM) and genetic algorithm (GA). Two maps, one simple and one complex, were used to compare their performances. In PRM, a map was initially loaded and followed by identifying the number of nodes. Then, initial and final positions were defined. The algorithm, then, generated a network of possible connections of nodes between the initial and final positions. Finally, the algorithm searched this network of connected nodes to return a collision-free path. In GA, a map was also initially loaded followed by selecting the GA parameters. These GA parameters were subjected to explorations as to which set of values will fit the problem. Then, initial and final positions were also defined. Associated cost included the distance or the sum of segments for each of the generated path. Penalties were introduced whenever the generated path involved an obstacle. Results show that both approaches navigated in a collision-free path from the set initial position to the final position within the given environment or map. However, there were observed advantages and disadvantages of each method. GA produces smoother paths which contributes to the ease of navigation of the mobile robots but consumes more processing time which makes it difficult to implement in real time navigation. On the other hand, PRM produces the possible path in a much lesser amount of time which makes it applicable for more reactive situations but sacrifices smoothness of navigation. The presented advantages and disadvantages of the two approaches show that it is important to select the proper algorithm for path planning suitable for a particular application.
Full text available upon request to the author
Article title: Quality assessment of lettuce using artificial neural network
Authors: Ira Valenzuela, John Carlo Velasco Puno, Argel Bandala, Renann Baldovino, et al.
Conference title: 2017 IEEE 9th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM)
Abstract:
The critical features in yield forecasts determination are crop health and seasonal progress. These serve as an indicator for the success of farming. Visual inspection often produces a false assumption on the quality of the lettuce crop health. To address this problem, a proposed solution is the development of a machine vision system for the assessment of the quality of the lettuce crop. This system is composed of two parts: application of digital image processing for the feature extraction of the sample lettuce and implementation of the back propagation artificial neural network for the self-learning classification of the system. ANN is a tool designed like a human brain that can learn patterns and relationship based on the input data. Also, backpropagation has been used because it has the capability to adjust its weights and biases in increasing the efficiency of its learning. A total of 253 images were collected and 70% of these were used for training the network, 15% fro validation and 15% for testing. The developed system produced was able to classify the quality of the lettuce with minimum relative error of 0.051.
Article title: Integrative review of the development of a multi-object tracker from a dual camera system in an unmanned aerial vehicle (UAV)
Authors: Jay Robert Del Rosario, Janela Assumpta L. Angeles, Aldwin Jerome P. Cabebe, Argel Bandala, et al.
Conference title: 2017 IEEE 9th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM)
Abstract:
In the modern age, unmanned aerial vehicles are used in industries for a variety of reasons ranging from surveillance uses to product deliveries. They are also used by people for recreational purposes such as aerial photography or drone racing. However, the capabilities of these UAVs are limited to merely recording and storing the videos. This paper introduces the development of a quadcopter capable of multi-object detection. It also explores the possibility of using a dual camera system for extended range.
Full text available upon request to the author
Article title: Machine vision for rat detection using thermal and visual information
Authors: Georjean D. S. Brown, Argel Bandala, Carlo Enrico A. Latonio, Richard Dean N. Oanes, et al.
Conference title: 2017 IEEE 9th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM)
Abstract:
Pests, particularly rodents, are a major cause of problem to people because of the deadly diseases it spreads and the damage it does on field crops as it decreases billion worth of yield crop production in the Philippines. A detection measure to help eradicate rats is proposed by the researchers to prevent future failures or deficiencies in crop cultivation. Researchers developed a solution to this by designing and developing a Machine Vision System using thermal and visual identification for rodent identification. Thermal imaging uses infrared imaging to detect and record only thermal temperature patterns emitted by an object whereas visual imaging record videos exposed to good lighting and has not been configured for dark environment tracking. The rat detection accuracy of both individual cameras were recorded for data comparison and researchers proved that the use of a thermal camera arise to results that are more accurate than with the use of a visual camera.
Full text available upon request to the author
Article title: Multi-view multi-object tracking in an intelligent transportation system: A literature review
Authors: Jay Robert Del Rosario, Argel Bandala, Elmer P. Dadios
Conference title: 2017 IEEE 9th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM)
Abstract:
A comprehensive and interdisciplinary review of notable literatures conducted were topic closely related to object detection and surveillance, mainly vehicle tracking. This survey of literature is focus on multi view vision system in various platform like: static and dynamic cameras.
Full text available upon request to the author
Article title: Fuzzy-based fault-tolerant control of Micro Aerial Vehicles (MAV) — A preliminary study
Authors: Mark Lester F. Padilla, Selwyn Jenson C. Lao, Renann Baldovino, Argel Bandala, et al.
Conference title: 2017 IEEE 9th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM)
Abstract:
Unmanned Aerial Vehicles (UAV) has gained popularity in the past decades. This has been widely used throughout the world in the fields of military, surveillance, agriculture, and construction. One of the main problems in Micro Aerial Vehicles (MAV), typically smaller version of UAV, is its ability to detect and tolerate faults inside the system. In this paper, a Fault-Tolerant Control (FTC) will be developed using fuzzy logic and uses battery percentage and degree of ability to hover as the crisp inputs. The fuzzy logic will use five and three membership functions for the Battery Percentage and Degree of Ability to Hover respectively. The output of the controller will be the degree of ability to continue a certain mission. Further studies can include other constraints such as mapping efficiency where neural networks and deep learning can be associated. Thus, making a hybrid system.
Full text available upon request to the author
Article title: Real-time vehicle detection and tracking using a mean-shift based blob analysis and tracking approach
Authors: Rhen Anjerome Bedruz, Edwin Sybingco, Argel Bandala, Ana Riza Fernandez Quiros, et al.
Conference title: 2017 IEEE 9th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM)
Abstract:
Implementing computer vision on traffic scenarios are one of the most widely sought area in the field of vision research. In dealing with the surveillance in traffic scenarios, every vehicle in the scene must be observed which results to problem arising from instances whenever the traffic density in an area is high due to occlusion caused by the large number of vehicles being observed. Thus, this paper proposes a vehicle detection and tracking algorithm whose main purpose is to detect and track vehicles entering an intersection and track them robustly in real-time. The algorithm which was used is a blob analysis and tracking based on a mean-shift kernel. The blob approach acts as the main tracking and will use the mean-shift in the event of blob merging or occlusion. In this paper, the proposed tracking method is tested using a CCTV camera on an intersection with high traffic density to illustrate the capability of solving occlusion and observe the robustness of the algorithm in the scene. The results show that the proposed system successfully tracks the vehicles during and after occlusion with other vehicles or other types of objects in the scene.
Full text available upon request to the author
Article title: Usage prediction of appliances in filipino households using Bayesian algorithm
Authors: Ian Joseph J. Pastorfide, Jua Franco M. Revilla, Chantel Kim D. Santos, Jennica Tsubasa F. Takada
Conference title: 2017 IEEE 9th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM)
Abstract:
The standby power accumulated after some time contributes to the wasted energy of a household and can be noticeable in a home's power consumption. In this study, the group aims to devise a standby power management system that is able to adapt constantly with one's changing lifestyle. To know the appliances available in households, a survey with 230 respondents was conducted and the most common appliances were taken into consideration. The power measurements of the appliances were also recorded using a power meter. The data log was conducted by members of different households for the activation of the appliances, the users, and the occupancy of the household. The mentioned factors from the usage log was then used on the Bayesian algorithm, which was used to calculate the probability of usage of the appliances. This learning prediction, in addition, to a power management system will minimize the power consumed by appliances in standby mode, thus saving energy and income.
Full text available upon request to the author
Article title: Design of a fuzzy logic controller for a vent fan and growlight in a tomato growth chamber
Authors: Arvin Fernando, Argel Bandala, Laurence A. Gan Lim, Maglaya Archie, et al.
Conference title: 2017 IEEE 9th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM)
Abstract:
In this paper, a fuzzy logic controller was design and develop to control the temperature, relative humidity and Carbon Dioxide (CO 2 ) inside the prototype tomato growth chamber. The model was develop to automatically adjust the inside parameters to obtain the optimum tomato plant environment condition. The growth chamber fuzzy logic controller was modeled using the MATLAB fuzzy logic tool box. In this research we design a fuzzy logic controller (FLC) to control the environment parameters in the growth chamber. In order to provide the most suitable conditions for the growth of the tomato plant and might minimize energy consumption.
Article title: Quadrotor system for gathering discomfort index and amount of air pollutants
Authors: Junlae Cheong, Rohit P. Nihalani, Noel B. Paulino, Argel Bandala, et al.
Conference title: 2017 IEEE 9th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM)
Abstract:
One can often feel some form of discomfort when going outdoors either from the heat or from the pollutants present in the air such as humidity, temperature, and the presence of air pollutants such as carbon monoxide and particulate matter. In the national setting, there has not been much awareness of the factors of discomfort yet these still prevail. The proposed solution for this problem is a system that is able to measure different parameters of discomfort. To provide safety for the measurer and the mobility of the system, the group has implemented the system on an unmanned aerial vehicle, specifically, a quadrotor. Here, the controller can manipulate the quadrotor system to go up at a desired altitude for measurements. This has been tested quarterly during the daytime and is able to maintain 85% accuracy but during testing, this may not be the case since there are other external factors that can affect the measurements such as the input power and wind. An Android application has been developed for the purpose of updating and viewing the data recorded.
Full text available upon request to the author
Article title: Artificial neural network model for solar resource assessment: An application to efficient design of photovoltaic system
Authors: Robert Martin Cahanding Santiago, Argel Bandala, Elmer P. Dadios
Conference title: TENCON 2017-IEEE Region 10 Conference
Abstract:
The power output of solar energy conversion facilities such as photovoltaic systems is highly dependent and proportional to the amount of solar radiation absorbed on the collecting surface. In order to have an efficient design of these systems, it is essential to perform solar resource assessment on the intended location prior to installation. Advancements in computational intelligence led to applications of artificial neural networks for solar resource assessment which outperforms existing empirical models in terms of speed and accuracy and overcomes the cost of using expensive solar radiation sensors. In this study, a single recurrent or feedback network is developed and assessed for efficacy in estimating the daily sum of solar radiation in the Philippines using meteorological data such as daily sum of sunshine duration, daily mean air temperature, daily mean air pressure, and daily mean air humidity. The collected data used in this study were obtained for the year 2014 from the Bureau of Soils and Water Management (BSWM) Agro-meteorological Station Lufft sensors in three locations: (1) Tanay, Rizal, (2) Barili, Cebu, and (3) Sto. Tomas, Davao del Norte. The developed model responded with mean squared error (MSE) values of 0.1491, 0.1679, and 0.2297 and regression values of 0.9146, 0.9313, and 0.9277 for the training, validation, and testing phases. The error histogram also shows that low values of error exist for each dataset and most errors fall between the ranges of -0.4581 to 0.5646. Results may further be improved by having larger data for training, validation, and testing phases for the neural network which can make the model more robust for larger variations in the weather patterns.
Full text available upon request to the author
Article title: Rust detection using image processing via Matlab
Authors: Julianne Diaz, Manuel I. Ligeralde, John Anthony Cheng Jose, Argel Bandala
Conference title: TENCON 2017-IEEE Region 10 Conference
Abstract:
This research attempted to create a program that is capable of detecting rust through image processing. Image processing is known for the manipulation of image through quantizing the image itself in matrix form. Through this quantization, it gives opportunity to not only manipulate the image but also detect a particular subject on the image as well, such as rust. Through setting the threshold values and the use of edge detection and segmentation, rusts on the image can be detected. The threshold values will set the parameters and characterize what a rust is. The edge detection will check for the sudden changes of colors in the images. The segmentation will then determine the colors on the image. The results in the edge detection and segmentation will be integrated to determine the rust on the image. The results of the program yield a success rate 90% in detecting rust on images with rusts and did not obtain any errors on images with no rust.
Article title: Development and implementation of swarm sweep cleaning protocol for quadrotor unmanned aerial vehicle (QUAV) swarm
Authors: Christian Kyle Y. Fermin, Arthur Lanz L. Imperial, Karlo Feliper D. L. Molato, Argel Bandala, et al.
Conference title: TENCON 2017-IEEE Region 10 Conference
Abstract:
This study aims to implement the SWEEP cleaning protocol in a swarm of flying robots. The dynamic cleaners' problem is one of the most popular application of swarm intelligence. The swarm is tasked to cover the target area in an optimized manner. It starts with an area where in a contamination spreads. The swarm is tasked to decontaminate the area and suppress the contamination. The accuracy and speed of cleaning is measured in static and dynamic contamination spreading with varying spreading time. Also, swarm members are increase in the given conditions. Experiment shows that there is an average decrease in cleaning time of 12.87% for every increase in swarm member number. The accuracy of the system is at cleaning the area is 89.8%. The completion of this research paves the way for several real life applications. Some of this will include search and rescue, target searching, and surveillance operations.
Full text available upon request to the author
Article title: A kNN-based approach for the machine vision of character recognition of license plate numbers
Authors: Ana Riza Fernandez Quiros, Rhen Anjerome Bedruz, Aaron Parayno Uy, Argel Bandala, et al.
Conference title: TENCON 2017-IEEE Region 10 Conference
Abstract:
This research proposes to automate the plate recognition process by installing an IP camera on a road and analyzing the video-feed to capture the vehicles along that road. The contours of the characters in a given plate image are detected, violated and isolated from the parent image. This results to segmented characters. Each of the characters are identified using a k nearest neighbors (kNN) algorithm. The kNN algorithm was trained using different sets of training data containing 36 characters each. The algorithm was tested on the previously segmented characters. The simulations show that an accuracy of 87.43% was achieved for the plate recognition algorithm using kNN at k = 1. Compared against existing character recognition techniques such as artificial neural networks (ANN), the difference in the accuracy is minimal. Moreover, the average processing time was 0.034 s.
Full text available upon request to the author
Article title: Development of a biomorphic and hyper-redundant caecilian based robots
Authors: Karl Fabico, Jan Karlo M. Hernandez, Simon Joseph P. Plata, Argel Bandala, et al.
Conference title: TENCON 2017-IEEE Region 10 Conference
Abstract:
This paper presents the development and design of a mobile multilink robot. Because of its multilink property, snake robots are appropriate in tight and hard to reach places. This is robot is intended to be deployed in such areas specifically in disaster rubbles. The robot is composed of 10 segments each of which has one degree of freedom. The robot is equipped with proximity sensors for obstacle detection and IMU for orientation sensing. In front of the robot, a wireless camera is attached so that the environment where the robot operates is viewed by the base controller. Experiments showed that the movement algorithm which follows the snake's biological motion is successfully implemented with a maximum movement velocity of 70cm/s. The maximum climb height is 22cm. This study demonstrates the effectiveness of the snake robot in different terrain with obstacles and small obstacle gaps.
Full text available upon request to the author
Article title: Localization of license plates using optimized edge and contour detection technique
Authors: Ana Riz Fernandez Quiros, Then Anjerome Bedruz, Aaron Parayno Uy, Argel Bandala, et al.
Conference title: TENCON 2017-IEEE Region 10 Conference
Abstract:
There are a lot of existing studies in the field of plate detection. However, most of them focused on using still images and only a few have applied the process on video streams. This research proposes to automate the plate detection process by the use of intelligent transport system through image processing techniques such as edge detection and contour matching. The region of interest of the vehicle image was reduced to the lower half since statistically, license plates were located on the bottom half of a vehicle to improve the computational complexity of the system. The edges of the vehicle image were computed, from which contours were calculated. The detected contours were filtered based on three parameters - area, aspect ratio and diagonal. The results show that the system achieved 96.67% accuracy.
Full text available upon request to the author
Article title: Obstacle avoidance algorithm for swarm of quadrotor unmanned aerial vehicle using artificial potential fields
Authors: Reagan L. Galvez, Gerard Ely Ucab Faelden, Jose Martin Maningo, Argel Bandala, et al.
Conference title: TENCON 2017-IEEE Region 10 Conference
Abstract:
Unmanned aerial vehicle that is moving from one place to another needs to have a real-time obstacle avoidance controller to prevent collisions in the obstacles around it. In this paper, the concept of artificial potential field is proposed to implement obstacle avoidance in swarm of quadrotors. This is based on the assumptions that the target and obstacle will introduce a certain force that will direct the robot to its destination. The effectiveness of this method was tested in a computer simulation and verified using real quadrotors.
Article title: Road lane reconstruction using vision — based macro block spatial predictions
Authors: Edison Roxas, Rhay Rhay P. Vicerra, Arvin Fernando, Argel Bandala, et al.
Conference title: TENCON 2017-IEEE Region 10 Conference
Abstract:
Vision - based road lane detection and reconstruction is a very common interest in the field of computer vision (CV). It has numerous application ranging from autonomous vehicle to driver assist and support systems technology. These researches are always focusing on both accuracy and complexity of the system's output; however, none of these uses Macro Block (MB) method. This paper introduces the characteristics of MB method used for spatial road lane detection and reconstruction subjected to different environment conditions; different MB size; and different function approximations.
Article title: Optimization of Photosynthetic Rate Parameters using Adaptive Neuro-Fuzzy Inference System (ANFIS)
Authors: Ira Valenzuela, Renann Baldovino, Argel Bandala, Elmer P. Dadios
Conference title: 2017 International Conference on Computer and Applications (ICCA)
Abstract:
Crop growth is greatly affected by light intensity, temperature and CO 2 concentration. The combinations of these factors are considered in growing crops. In this study, a system was developed using adaptive neuro-fuzzy inference system for the prediction of the photosynthetic rate of lettuce crop based on the temperature, light intensity and CO 2 . A fuzzy inference system is designed to generate the rules for the fuzzy logic where inputs of these are from the output of the trained neural network. Based on the result, the system was able to predict the photosynthetic rate of the lettuce crop based on the three input parameters. The RMSE value for the ANFIS model was found to be 2.7843e-05.
Article title: Vehicle detection and tracking using corner feature points and artificial neural networks for a vision-based contactless apprehension system
Authors: Robert Kerwin Dela Cruz Billones, Argel Bandala. Edwin Sybingco, Laurence A. Gan Lim, et al.
Conference title: 2017 Computing Conference
Abstract:
Blocked intersections have been a contributing factor in the city-wide traffic congestion, especially in metropolitan cities. This research study aims to develop a better traffic violations management system in city-road intersections by using a machine vision system that automatically identifies and tags traffic violations committed in an intersection. The proposed system have three main sub-systems which are the video capture, video analysis, and output sub-systems. This study presents the development and results of a vehicle detection and tracking system using corner feature point detection and artificial neural networks for the vision-based contactless traffic violations apprehension system. This detection and tracking system serves as the front-end processing in the video analysis sub-system. Experiments were conducted for different corner feature-points detection algorithm: Harris, Shi-Tomasi, and Features from Accelerated Segment Test (FAST). The results showed that in the testing phase Harris-ANN have 89.09% accuracy, Shi-TomasiANN have 88.48%, and FAST-ANN have 90.30% accuracy.
Article title: Automated vehicle class and color profiling system based on fuzzy logic
Authors: Aaron Christian P. Uy, Rhen Anjerome Bedruz, Ana Riza Fernandez Quiros, Argel Bandala, et al.
Conference title: 2017 5th International Conference on Information and Communication Technology (ICoIC7)
Abstract:
The study proposes an automated vehicle class and color profiling system to specifically have distinct information on any apprehended car in an intelligent traffic system. The problem arises from the fact that traffic enforcers are sometimes unreliable with apprehending cars due to the lack of information on the violator. The solution is an automated system which consists of background difference method, and fuzzy logic to classify these violators. The general process is as follows: a capture picture from a traffic CCTV camera is subjected to a car detection process, and then the fuzzy inference systems are run to find the class and color of the car, and finally display a cropped picture of it along with the said descriptions. The automated car profiling system was found to have an accuracy of 99.391% for the classification process while 98.580% for the color profiling process. These results show that the algorithm is well-suited for a reliable implementation on intelligent traffic system.
Full text available upon request to the author
Article title: Fuzzy-based Decision Support System for Smart Farm Water Tank Monitoring and Control
Authors: John Dela Cruz, Renann Baldovino, Francisco Culibrina, Argel Bandala, et al.
Conference title: 2017 5th International Conference on Information and Communication Technology (ICoIC7)
Abstract:
Water is considered as the blood of the irrigation system. It is a basic necessity in farms so as to have a high amount of production. For some farms located in higher elevation, electric motor pump is being used to collect water from underground reservoir. Since electric motor pumps will be used, electric consumption issue is also a concern. Proper allocation of available resources is one of the main issues for smart farm. In this paper, the authors proposes the consideration of using of a Fuzzy-based Decision Support System (FDSS) in the Water Tank Monitoring and Control Subsystem (WTMCS) of the Smart Farm Automated Irrigation System (SFAIS) based on the following: (1) water level in the tank storage connecting the motor pump and the irrigation pipes, and (2) availability of electricity from a power source with limited amount of energy available. MATLAB simulations, via Fuzzy Logic Toolbox of Simulink, were done to verify the feasibility of the proposed system before the actual implementation.
Full text available upon request to the author
Article title: Water usage optimization of Smart Farm Automated Irrigation System using artificial neural network
Authors: John Dela Cruz, Renann Baldovino, Argel Bandala, Elmer P. Dadios
Conference title: 2017 5th International Conference on Information and Communication Technology (ICoIC7)
Abstract:
Limited water resources had become the main constraint to be considered in farming. Optimizing this has become one of the interests in researches relating to precision agriculture. In this paper, the researchers use Neural Network in optimizing the water usage in the smart farm by incorporating it to the proposed Smart Farm Automated Irrigation System (SFAIS) by implementing an expert system. Simulations were done using the MATLAB Neural Network toolbox and results show that neural network is a useful tool.
Full text available upon request to the author
Article title: Multiple objective optimization of LED lighting system design using genetic algorithm
Authors: Robert Martin Cahanding Santiago, John Anthony Cheng Jose, Argel Bandala, Elmer P. Dadios
Conference title: 2017 5th International Conference on Information and Communication Technology (ICoIC7)
Abstract:
In order to maximize the advantages of LED lighting systems for controlled environment agriculture (CEA), several considerations must be taken into account such as the achievement of required daily light integral (DLI), uniform light distribution over the plant growing area, and minimize the investment and operating costs associated with the lighting system. This study aims to apply the multiple objective optimization of genetic algorithm in designing a lighting system that meets the mentioned objectives. The optimization variables, number of bits per variable and maximum number of iterations are fixed parameters tuned to the requirements of this application and the population size, mutation rate, and selection rate are genetic parameters for explorations. Results of the algorithm suggest the use of a number of LED lamps that is 31.25% lower than the maximum number of lamps that may be used in the plant growing area and, consequently, reduce the investment and operating costs while maintaining the required light integral capacity and uniformity. This and other studies that aim to develop and optimize LED lighting systems open more possibilities and promote the technology for controlled environment. Moreover, control and optimization of agricultural practices can lead to better plant quality and production even on locations and periods that they do not usually grow.
Full text available upon request to the author
Article title: Obstacle Avoidance for Quadrotor Swarm Using Artificial Neural Network Self-Organizing Map
Authors: Argel Bandala, Jose Martin Maningo, Gerard Ely Ucab Faelden, Reiichiro Christian S. Nakano, et al.
Conference title: 2015 International Conference on Humanoid Nanotechnology, Information Technology, Communication and Control, Environment and Management (HNICEM)
Abstract:
Swarm operation in Unmanned Aerial Vehicles is an emerging technology which has numerous uses. It can be used in industrial, agricultural, and even military applications. However, it must be able to perform formations for it to be effective. Also, countermeasures must be made by the swarm to account for certain obstructions that are present in the environment. This paper aims to address this issue by implementing an artificial neural network self-organizing map to give the correct coordinates to each swarm individual such that the swarm formation would be present in the given space while avoiding the obstructions present. Testing would include subjecting the system to three different obstruction patterns in a given 3D space. The results showed that for all cases, the swarm was able to avoid all the obstructions.
Article title: Philippine vehicle plate localization using image thresholding and genetic algorithm
Authors: Rhen Anjerome Bedruz, Edwin Sybingco, Argel Bandala, Ana Riza Fernandez Quiros, et al.
Conference title: TENCON 2016-2016 IEEE Region 10 Conference
Abstract:
This paper proposes a vehicle plate localization method using genetic algorithm integrated with image thresholding. Image thresholding outputs a value which varies on the time the image is captured. Genetic algorithm on the other hand, executed the license plate region detection of the digital image which depends on the set-level of the image threshold values obtained. Using the proposed algorithm, it was shown how the algorithm was effective on finding the plate location in a given image. Results show that the different parameters tested were successful and converges to a point where the plate locations can be located. The algorithms were tested on an image of a vehicle equipped with a license plate on its frontal view tested on a large number of trials. The genetic algorithm initialized 2000 chromosomes as its initial population and a fixed generation’s count of 100. It was observed that the time it took for the program to locate the plate is about 3 seconds. Another finding observed is that by varying the initial chromosome count and generation count will lead to longer computation time with increased accuracy. On the contrary, if the initial values were lessened, computation time will be less but the accuracy lessen. Results show that this plate localization technique successfully locates the plate and may be calibrated depending on the time of analysis.
Article title: Machine vision of traffic state estimation using fuzzy logic
Authors: Ana Riza Fernandez Quiros, Rhen Anjerome Bedruz, Aaron Parayno Uy, Argel Bandala, et al.
Conference title: TENCON 2016-2016 IEEE Region 10 Conference
Abstract:
One of the problems encountered by motorists are congested roads. Current technology cannot easily broadcast the information about which roads are heavily congested and which are not to the motorists. As such, planning of the route to take to their destinations is compromised. This paper proposes a fuzzy logic method approach to the estimation of the traffic state of a road. Images from IP cameras installed in different roads can be used to determine the state of the traffic in an area at any point in time. The vehicles within the image are needed to be detected first via edge detection. As the vehicles are detected within the image, so are their position and size with respect to the whole image are obtained. As such, three different parameters namely vehicle density, distance between neighboring vehicles and vehicle sizes can be computed. Using these three parameters, a fuzzy logic system can be created. Three degrees of intensity for each parameter was used, creating 27 rules. The center of gravity method was used to defuzzify the traffic density parameter. Based on the results, the designed algorithm was able to identify six different road images of different traffic states accurately.
Full text available upon request to the author
Article title: Formation control in quadrotor swarm aggregation using Smoothed Particle Hydrodynamics
Authors: Jose Martin Maningo, Gerard Ely Ucab Faelden, Reiichiro Christian S. Nakano, Argel Bandala, et al.
Conference title: TENCON 2016-2016 IEEE Region 10 Conference
Abstract:
This paper uses the Smoothed Particle Hydrodynamics technique to perform formation control of quadrotor swarms. The swarm is to be modelled to behave like water. A simple aggregation behavior is exhibited with certain primitives that act as obstacles to force formations from the swarm. Different primitives are implemented to manifest various formations. Results show that SPH outperforms APF by a margin of 7.31% for a cubic container primitive and by a margin of 27.81% for a spherical target enclosure primitive. Formation control was successfully implemented using Smoothed Particle Hydrodynamics and is proven to be more efficient than the benchmark algorithm.
Full text available upon request to the author
Article title: Automated traffic violation apprehension system using genetic algorithm and artificial neural network
Authors: Aaron Parayno Uy, Ana Riz Fernandez Quiros, Rhen Anjerome Bedruz, Argel Bandala, et al.
Conference title: TENCON 2016-2016 IEEE Region 10 Conference
Abstract:
Developing countries face the problem of crowded and congested roads because of inefficient implementation of traffic rules. Motorists ignore the rules because they are not apprehended and can get away easily. This paper proposes an intelligent traffic system that is able to automatically detect and apprehend traffic violators, specifically motorists who either swerve or block the pedestrian lane. The system is designed by integrating three processes: violation detection, plate localization and plate recognition. The violation detection and plate localization were realized using genetic algorithm while the plate recognition process was performed using an artificial neural network. The recognition of the plate number is highly dependent on the position of the detected vehicle with respect to the camera. Thus, the recognized plate number will only be supplementary information about the violator; the physical attributes of the vehicle which is captured by the violation detection process will be the main information on the violator. Based on the results of 48 images tested, the overall system was able to detect the mentioned violations and to identify the plate number of the vehicles that were detected as traffic violators, with an average accuracy of 90.67%, and program runtime of 1.34 seconds.
Full text available upon request to the author
Article title: Implementation of swarm aggregation in quadrotor swarms using an artificial potential function model
Authors: Gerard Ely Ucab Faelden, Jose Martin Maningo, Reiichiro Christian Nakano, Argel Bandala, et al.
Conference title: TENCON 2016-2016 IEEE Region 10 Conference
Abstract:
Swarm robotics is one of the novel approaches being explored in multiple quadrotor. It aims to mimic social behaviors of animals and insects. This paper presents the physical implementation of the swarm behavior aggregation in a quadrotor swarm. It is implemented over a quadrotor swarm testbed that makes use of external motion capture cameras. The completed algorithm makes use of the artificial potential function model with a linear attraction and bounded repulsion. Results show successful demonstration of the aggregation algorithm with minimal error in position. It is tested for an increasing number of quadrotors and errors are seen to increase with swarm size. Results show an error of 3.293 cm from the individual target position for aggregation. It also shows and average aggregation speed of 1.896 secs for all test while having an increase in aggregation speed of about 1.772 sec per increase in swarm size. The time in aggregate is seen to be at an average of 98.5405% of the time. All the tests show successful demonstration of the swarming behavior which can now mark the start of development of implementation of more complex swarming behaviors.
Full text available upon request to the author
Article title: Intelligent system architecture for a vision-based contactless apprehension of traffic violations
Authors: Robert Kerwin Dela Cruz Billones, Argel Bandala, Edwin Sybingco, Laurence A. Gan Lim, et al.
Conference title: TENCON 2016-2016 IEEE Region 10 Conference
Abstract:
The paper presents an intelligent system architecture for detecting traffic violations based on vision. This study aims to better manage traffic conditions in block intersections by employing a computer vision system that facilitates the identification of traffic violations committed in the road intersection. The architecture includes three sub-system: video capture sub-system, intelligent operating architecture (IOA) sub-system, and output sub-system. The IOA manages different algorithms to recognize traffic violations. The algorithms developed are vehicle detection and tracking, plate number localization, plate character recognition, and traffic violations identification. The traffic violations addressed in this study are number coding, over-speeding, and swerving. The research study is in the initial phase of development, and the experiment results showed that optical character recognition have 86.11% accuracy and speed measurement have 88.45% accuracy.
Full text available upon request to the author
Article title: Implementation of Varied Particle Container for Smoothed Particle Hydrodynamics-Based Aggregation for Unmanned Aerial Vehicle Quadrotor Swarm
Authors: Argel Bandala, Gerard Ely Ucab Faelden, Jose Martin Maningo, Reiichiro Christian S. Nakano, et al.
Conference title: 2016 IEEE/RSJ International Conference on Intelligent Robost and Systems (IROS)
Abstract
The property of the Smoothed Particle Hydrodynamics (SPH) method of being mesh free, adaptable and sui table for tracking of individual particles makes it appropriate for approximating swarm behaviors for multi-agent robotics applications. The researchers modeled each of the swarm robots as SPH particles and then subjected them to external forces to exhibit aggregation and force certain formations. The external forces subjected to the SPH particles are gravity forces and container constraints . The containers explored in the study are simple geometrical primitives: sphere and cube . Computer simulations were done to show how SPH can facilitate in forcing swarm formations with the help of bounding primitives. Algorithm benchmarking was done to show how well SPH performs. Results show that SPH performs better than the benchmark algorithm by a margin of 0.703 and 1.016 units for the two set-ups, respectively. Actual robot implementation was also done to verify the effectivity and viability of the proposed algorithm in exhibiting the aggregation behavior. After 15 seconds of system run time, the interparticle distance and motion accuracy reached 96.93% and 91.14%, respectively.
Article title: Utilization of Sensor Network for Combustible Gas Detection and Monitoring Implemented in Household
Authors: Argel Bandala, Kenneth V. Balmes, James Matthew T. Chua, Mary Anne O. De Jesus, et al.
Conference title: 2015 IEEE Region 10 Humanitarian Technology Conference
Abstract
Human negligence and ignorance concerning LPG handling can cause serious risks which may lead to fire and explosion. This study aims to create an automated system that intervenes human processes and implements necessary precautions based on the current state of the concerned area. The fulfillment of this study has produced a small scale system capable of detecting LPG concentration within the accuracy range of 85% to 100%. Results of this study shows that after detecting LPG in the environment, the system then implements specified action and notification procedures with 100% accuracy, thus preventing LPG leakage hazards to aggravate.
Full text available upon request to the author
Article title: A Comparative Study of Swarm Foraging Behaviors; Trophallaxis, Task Allocation and Pheromone
Authors: Argel Bandala, Ryan Rhay P. Vicerra, Elmer P. Dadios
Conference title: 2015 International Conference on Humanoid Nanotechnology, Information Technology, Communication and Control, Environment and Management (HNICEM)
Abstract
A group of algorithms enhancing such collective behavior is inspired by the animals working together as a group such as ants, bees, and etc. In connection, swarm is defined as a set of two or more independent homogeneous or heterogeneous agents acting upon a common environment in a coherent fashion which generates emergent behavior. The development of artificial swarms or robotic swarms has attracted a lot of researchers in the last two decades including pheromone, trophallaxis and task allocation algorithms. However among these swarm based algorithms, the most efficient in terms of group performance, efficiency and interference in collecting the dusts or objects in an environment with variable terrains. With this, the researchers see the need to develop a swarm simulation platform that would compare the swarm- behavior-based algorithms for an ideal use of robots in different environments in dust collection.
Full text available upon request to the author
Article title: Color Quality Assessment of Coconut Sugar Using Artificial Neural Network (ANN)
Authors: Argel Bandala, Aaron Aquino, Mary Grace Ann Bautista, Elmer P. Dadios
Conference title: 2015 International Conference on Humanoid Nanotechnology, Information Technology, Communication and Control, Environment and Management (HNICEM)
Abstract
This paper presents a simple color recognition algorithm using digital image processing techniques and pattern recognition to eliminate the subjectiveness of manual inspection of the quality of coconut sugar based on Philippine National Standard. The image processing was built using MATLAB functions through RGB acquisition. The Backpropagation Artificial Neural Network was used in this project to enhance the accuracy and performance of image processing. The database of the network involved 300 images and 70% of these were used for training the network, 15% for validation and 15% for testing.
Full text available upon request to the author
Article title: Obstacle Avoidance of Hybrid Mobile-Quadrotor Vehicle With Range Sensors Using Fuzzy Logic Control
Authors: Argel Bandala, Ryan Rhay P. Vicerra, Elmer P. Dadios
Conference title: 2015 International Conference on Humanoid Nanotechnology, Information Technology, Communication and Control, Environment and Management (HNICEM)
Abstract
This study presents a fuzzy logic based approach to a hybrid mobile quadrotor vehicle that is able to perform goal seeking and obstacle avoidance, given that the obstacles are nonmoving and are along a fixed path. Two range sensors will be used to construct the input variable of the fuzzy logic control. The algorithms are developed to achieve goal position while avoiding obstacles. Simulations are conducted and the efficiency of the results using the method is proved using MATLAB.
Article title: Predicting the motion of quadrotor using neural network
Authors: Argel Bandala, Reagan L. Galvez, Elmer P. Dadios
Conference title: 2015 International Conference on Humanoid Nanotechnology, Information Technology, Communication and Control, Environment and Management (HNICEM)
Abstract
Accurate control of quadrotors movement is a challenging task. One must have a reliable controller that will manipulate the speed of each motors. It is also important to study the effect of voltage to the motor speed. This paper will use neural network to predict the rolling and pitching motion of the quadrotor based on the voltage inputs of each motor. It will also determine the net force along z-axis acting at the center of mass of the quadrotor. The neural network is a computer algorithm that mimics the biological structure of neurons (nerve cell) and a powerful tool commonly used in fitting a function, pattern recognition, face recognition, clustering and optimization.
Article title: A Multiple Level MIMO Fuzzy Logic Based Intelligence for Multiple Agent Cooperative Robot System
Authors: Argel Bandala, Ryan Rhay P. Vicerra, Kanny Krizzy A. David, Angelo Rejaba dela Cruz, et al.
Conference title: 2015 IEEE Region 10 Conference (TENCON 2015)
Abstract
Fuzzy Logic is a many valued logic and it is very similar to human reasoning which is not binary. It uses approximate measures rather than exact, making it suitable for linguistic variable and analysis. It has been applied to many applications in artificial intelligence, control and robotics. In this paper, the authors develop an artificial intelligence using multiple fuzzy logic for a dynamic multiple agent robot system. The system is made up of multiple robots with multiple identity assignment; which means that each robot will have its distinct behavior. In order to design pure fuzzy logic artificial intelligence, we used fuzzy logic block in different parallel and series configuration making giving it multiple fuzzy logic levels. Furthermore, there is multiple input - multiple output (MIMO) fuzzy logic implementation in one of our several fuzzy logic blocks, this is necessary in order to utilize pure fuzzy logic control in the whole artificial intelligence. The multi agent cooperative robot platform we choose to test our artificial intelligence is a multiple robot system for FIRA Micro-Robot World Soccer Tournament (MiroSot). In our setup, there are three robots to be assigned dynamically with three different identities; the Forward, the Back and the Goal-keeper. Robot identity assignment depends on the position of each robot with respect to the position of the ball. To tune each fuzzy logic block individually isolation is done. Some tuning procedures are performed in a simulator while most of them are tuned in the actual platform. Although tuning procedures are rigorous, the linguistic approach and human reasoning nature of fuzzy logic made it possible to achieve its completion. Overall, the proposed artificial intelligence produced favorable response based on the expected outcome and experimentations.
Article title: Eye State Analysis Using EyeMap for Drowsiness Detection
Authors: Argel Bandala, Jenel Luise C. Bolosan, Mary Lisette L. dela Torre, Josephine Gomez, et al.
Conference title: 2015 IEEE Region 10 Conference (TENCON 2015)
Abstract
Drowsiness has become one of the many reasons of vehicular accidents. This research aims to create a system that can analyze whether the person is drowsy or non- drowsy and send a warning signal whenever it detects signs of drowsiness. This design undergoes several image processing for boosting the systems capability to retain only the region of interest and successfully initiate alarms within minimal time. It utilizes EyeMap mainly for eye localization and windowing and aided by the Circular Hough transform to extract only the eye region - specifically the iris; and classify whether the person is experiencing drowsiness at the moment. The researchers develop an additional device that is equipped with three warning signals and reacts on how the system sees the state of the person. Three setups were implemented in this study: Regular Camera, Infrared Sensitive Camera and Multiple Cameras. All setups were implemented during day and night to test the response of the system to varying lighting conditions. The subjects are tested inside a car and their present state is determined using the Karolinska Sleeping Scale. The current state of the person is then compared to the system's response. The subjects are tested three times under different setups to determine if the system is responding correctly under different condition. The study shows that the system is able to successfully determine whether the person is in the drowsy or non-drowsy state in all of the three setups, multi-camera being the most effective. However, it is limited by the capability of the camera to adapt to different lighting condition. During night time, the ability of the system to determine the state of the system drops.
Article title: Implementation of an Artificial Neural Network in Recognizing in-flight Quadrotor Images
Authors: Argel Bandala, Reiichiro Christian S. Nakano, Gerard Ely Ucab Faelden, Jose Martin Maningo, et al.
Conference title: 2015 IEEE Region 10 Conference (TENCON 2015)
Abstract
This paper shows an implementation of a feedforward artificial neural network capable of recognizing images of the CrazyFlie 2.0 quadrotor during flight. The network is to be used in a real-time quadrotor swarming application and has to be able to successfully differentiate pictures that show a quadrotor in flight versus pictures that do not. The network was trained using a standard backpropagation algorithm and images taken from a video of the said quadrotor in flight. These images were divided into three groups: a training set and validation set for the training stage, and a testing set for verification of the trained neural network. The results showed that the neural network was able to correctly identify the images in the testing phase 100 percent of the time while achieving a 94 percent accuracy for the images in the testing set.
Article title: A Neural Network Approach to a Cooperative Balancing Problem in Quadrotor-Unmanned Aerial Vehicles (QUAVs)
Authors: Argel Bandala, Gerard Ely Ucab Faelden, Jose Martin Maningo, Reiichiro Christian S. Nakano, et al.
Conference title: 2015 International Conference on Humanoid Nanotechnology, Information Technology, Communication and Control, Environment and Management (HNICEM)
Abstract
There is growing interest in unmanned aerial vehicles (UAVs) such as quadrotors over the past several years. Cooperation among multiple quadrotors is one of the areas of focus. This paper proposes a neural network form of control for a cooperative task done by four quadrotors and will be tested through simulations. The task at hand is a ball and plate balancing problem during flight of multiple quadrotors carrying the plate. The objective is to maintain the keep the ball at the center of the plate even if the ball is introduced at different parts of the plate. The neural network controller will output the appropriate motor speeds of the rotors based on the detected area of introduction of the ball. Results show that the artificial neural network controller successfully directs the ball towards the center of the plate. The network outputs an average deviation of 0.00924 units from the expected PWM signal strength which corresponds to a 0.249% error from the expected value.
Article title: Adaptive Aggregation Algorithm for Target Enclosure Implemented in Quadrotor unmanned aerial vehicle (QUAV) Swarm
Authors: Argel Bandala, Ryan Rhay P. Vicerra, Elmer P. Dadios
Conference title: 2014 International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment and Management (HNICEM)
Abstract
This paper presents aggregation behavior algorithm that will be applied for unmanned aerial vehicle quadrotors (QUAV). The most basic behavior for natural swarms is aggregation. Other swarm or social behaviors are derived from the aggregation behavior. Due to the concept of independence, each swarm members are required to collect themselves together to perform a certain task. However the swarm faces different environments thus this behavior is very complex to accomplish. This is the reason why the researchers developed this paper for multi robotic systems. Simulations were done to test the said algorithm and the researchers garnered the accuracy of 90.85%.
Full text available upon request to the author
Article title: A Genetic Algorithm Approach to Swarm Centroid Tracking in Quadrotor Unmanned Aerial Vehicles
Authors: Reiichiro Christian S. Nakano, Argel Bandala, Gerard Ely Ucab Faelden, Jose Martin Maningo, et al.
Conference title: 2014 International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment and Management (HNICEM)
Abstract
One of the trademark behaviors of a swarm is aggregation. Aggregation is the ability to gather swarm members around a specific point in space. The goal is to keep an object, stationary or moving, at the center of the swarm. This paper presents a novel approach to centroid tracking in robotic swarms. Genetic algorithm is used in quadrotor unmanned aerial vehicles to keep the object being tracked at the center while minimizing two parameters: the distance travelled by each quadrotor and the distance of each quadrotor from the object. Centroid tracking was found to have an average error of 0.0623568 units for swarm populations ranging from 10 to 100 with the lower swarm populations exhibiting lower errors. Convergence did not exceed the maximum of 23 milliseconds for populations less than 30. These results show that the algorithm is well-suited for implementation in swarms with lower numbers of quadrotors.
Article title: Optimization of Decentralized Information Dissemination in Quadrotor Swarm Using Genetic Algorithm
Authors: Argel Bandala, Gerard Ely Ucab Faelden, Jose Martin Maningo, Reiichiro Christian S. Nakano
Conference title: 2014 International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment and Management (HNICEM)
Abstract
There is a glaring problem in communication systems when it comes to a decentralized robotic swarm. Since a decentralized swarm would limit the awareness of each agent to its immediate surroundings/neighbors, the exchange of information between agents may now prove to be challenging. An epidemic-based broadcasting technique is then presented to resolve the problem of end-to-end agent communication. This paper aims to optimize the information diffusion by means of implementing genetic algorithm to optimize the time it will take for each quadrotor individual to acquire the information coming from a single source (i.e. the quadrotor who first received the information from an external stimulus). The method by which this is done is epidemic in nature. Due to this, for each time there would be a signal broadcasting, the genetic algorithm would be run to determine the next ideal location of each individual. A genetic algorithm was looped several times to achieve the desired solution. The results showed that for each run of the GA, the number of quadrotors having received the information continually increased until the output converges to a fitness level. However this only worked under certain constraints that need to be weighed out properly. This includes the readjustment of the fitness and crossover functions. Also, the parameters of the GA must be well calibrated for proper output response.
Article title: Blind Localization Method for Quadrotor Unmanned Aerial Vehicle (QUAV) Utilizing Genetic Algortihm
Authors: Argel Bandala, Gerard Ely Ucab Faelden, Jose Martin Maningo, Reiichiro Christian S. Nakano, et al.
Conference title: 2014 International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment and Management (HNICEM), November 2014
Abstract:
There is an increasing research interest in unmanned autonomous vehicles (UAVs) such as quadrotors. These researches applies these quadrotors for much more complicated tasks with most requiring cameras and GPS modules for positioning. This paper presents an alternative way of position localization of a quadrotor without the use of cameras and GPS modules by means of transceivers and Genetic Algorithm (GA). This paper uses the received signals from the transceivers as inputs for the genetic algorithm in order to locate the quadrotor in a xyz axis. Parameters such as location of transceivers, amount of transceivers and population size of the GA are tested in order to determine a successful way of locating the quadrotor. Results show that the different parameters tested were successful and converges to a point usually with a fitness measure greater than 99%. An average fitness measure greater than 99.9900% served as a benchmark for the tests done. The first test achieved this benchmark at about 130 generations and the second test achieved it at 110 generations. The time it took for the program to locate the quadrotor is about 60 milliseconds. Results show that this blind localization technique successfully locates the quadrotor and may be calibrated to one's own need.
Article title: Path Planning for Quadrotor UAV Using Genetic Algorithm
Authors: Argel Bandala, Reagan L. Galvez, Elmer P. Dadios
Conference title: 2014 International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment and Management (HNICEM),
Abstract:
Path planning in quadrotor-typed UAV is essential in navigating from initial to destination point. This will minimize the power consumption of the vehicle which is important to avoid wasted energy in a given amount of time. This paper will use Genetic Algorithm (GA) to determine the shortest path that the quadrotor must travel given one target point to save energy and time without hitting an obstacle. The obstacle is assumed to be any point within the boundary. This algorithm is effective in searching solutions in a given sample space or population. If you know the possible solutions of the problem, you can evaluate it based on its fitness until the fittest individual arrives.
Article title: Modelling and Characterization of an Artificial Neural Network for Infant Cry Recognition Using Mel-frequency Cepstral Coefficients
Authors: Argel Bandala, Allimzon M. Lim, Mark Anthony D. Cai, Allan Jeffrey C. Bacar, et al.
Conference title: TENCON 2014 - 2014 IEEE Region 10 Conference
Abstract:
This paper is about the creation of an artificial neural network (ANN) in MATLAB to analyze the features extracted from calculating the mel-frequency cepstral coefficients (MFCC) of the raw audio data. The paper explains basic concepts about the ANN, as well as the MFCC and other relevant theories. Regarding the design of the ANN, it uses multiple infant crying sounds, as well as non-crying sounds, to create a sample training set with a corresponding target that determines whether the sound is a cry or not. The paper uses relevant concepts heavily utilized in speech recognition for the design of the infant cry recognition, modifies them, and adds a few more calculations to fit the desired application to compensate for the differences present in a cry from human speech.
Article title: Swarm Algorithm Implementation in Mobile Robots for Security and Surveillance
Authors: Argel Bandala, Patricia Marie L. Lapena, Joseph Ian Q. Blanco, Kevin I. Bunda, et al.
Conference title: TENCON 2014 - 2014 IEEE Region 10 Conference
Abstract:
This paper presents the design and construction of a system consisting of five mobile robots (mobots) and a communications system that will serve as a security surveillance system. This is implemented using a microcontroller as the core that enables the mobots to work cooperatively. The mobots are free to move within their designated areas and are capable of relaying messages via ZigBee communication to a base controller system. The purpose of this system is to have an alternative or even a complement to regular CCTV surveillance, especially in buildings with several rooms. This would enhance the security as the system utilizes a database to store the information gathered from intruder alerts. Furthermore, the communication radios used transmit with low power over a long range. The mobots are enabled to relay data via mobot-to-PC and mobot-to-mobot paths up to five hops. The data transmitted allows the base controller system to identify its source for intrusion detection.
Full text available upon request to the author
Article title: Formation Stabilization Algorithm for Swarm Tracking in Unmanned Aerial Vehicle (UAV) Quadrotors
Authors: Argel Bandala, Ryan Rhay P. Vicerra, Elmer P. Dadios
Conference title: TENCON 2014 - 2014 IEEE Region 10 Conference
Abstract:
This paper presents swarm formation algorithm for swarm tracking behavior in multi robotic system of flying quadrotor unmanned aerial vehicles (QUAV). Multi robotic system ensures the success of the task through the increase in members of the swarm. This characteristic is very suitable for tracking moving objects. Another key feature would be the decentralized processing of the swarm. The loss of a swarm member would not contribute significantly to the swarm. The behaviors were patterned to biological traits of insects and animals and are applied to computer applications. Simulations were performed and results showed that swarm tracking accuracy yielded 89.23%. This result is due to the accuracy of 84.65% of the formation behavior of the swarm. Furthermore, the aggregation behavior further contributed with an accuracy of 90.62%.
Article title: Development and Design of Automated Hospital Bed With Incremental Panels for Bedsore Prevention
Authors: Argel Andala, Lance Kevin G. Apelo, Trizia C. Dimalanta, Jan-Anthony Vince V. Macatangay
Conference title: TENCON 2014 - 2014 IEEE Region 10 Conference
Abstract:
The common causes of Bedsores are constant pressure and moisture build-up. To prevent the development of these causes, the researchers have decided to modify the typical hospital bed into an electronically automated prototype. The prototype was built with 50 incremental panels that has the capability to move the patient on the by the combination of the Chain and Sprocket Method and the Lead Screw. These incremental panels were implemented with Temperature Sensors to monitor the ambient temperature on the bed surface and IR Proximity sensors to detect the location of the patient in the prototype. The data gathered by these sensors are processed through the use of a microcontroller. The microcontroller will then relay the processed data to the stepper motor driver which will manipulate the stepper motors to cause movement to the incremental panels.
Article title: Cabling and cost optimization system for IP based networks through Genetic Algorithm
Authors: Charmaine B. Balubal, Angela Rachel D. Bernardo, Argel Bandala, Regina A. Uyehara, et al.
Conference title: 2014 IEEE Region 10 Symposium
Abstract:
The creation of an optimized cabling plan in terms of cost through optimized cable length was introduced in this study. The researchers designed a system that utilized Genetic Algorithm for the said optimization. This system was integrated in a graphical user interface created using visual c# language which enables the users to upload an image representing the floor plan of the desired network to be optimized. The user can then place specified components on the floor plan. Lastly, the system will generate the optimized cabling plan which the user can readily print. Furthermore, a complete bill of materials and costing report will be generated also. The system generated these outputs by using genetic algorithm in the graphical inputs which were processed and converted in numerical representations. Upon accomplishing all the experimentations, the system yielded 99.51% optimization accuracy with 99.02% as the highest optimization level generated after accomplishing 100 trials on 10 different floor plans.
Full text available upon request to the author
Article title: Autonomous parallel parking of four wheeled vehicles utilizing adoptive Fuzzy-Neuro control system
Authors: Jerome T. Marasigan, Iara Buo Saberon, Argel Bandala, Dan Patrick B. San Jose, et al.
Conference title: 2014 IEEE Region 10 Symposium
Abstract:
The study presents an autonomous sensor based parallel parking maneuver on a car-like mobile robot. This project focuses on parallel parking the car-like mobile robot within a given scenario following the fifth degree polynomial reference path in a backward maneuver. Training data, gathered from the fifth degree polynomial path, is subjected to subtractive clustering algorithm to determine the fuzzy controller and trained by the adaptive neurofuzzy inference system. The project uses eight ultrasonic sensors, placed strategically to avoid radial imprecision, to detect the obstacles along its path; an accelerometer is also used to detect the inclination of the car-like mobile robot (CLMR). The sensors acquire necessary sensor data for the Neuro-Fuzzy Inference System to determine the proper motion direction at each sampling point. The efficiency of the proposed Neuro-Fuzzy Controller (NFC) design is revealed through the actual results.
Article title: Unmanned Underwater Vehicle Navigation and Collision Avoidance Using Fuzzy Logic
Authors: Argel Bandala, Kanny Krizzy A. David, Ryan Rhay P. Vicerra, Laurence A. Gan Lim, et al.
Conference title: 2014 IEEE Region 10 Symposium
Abstract:
Unmanned underwater vehicles (UUVs) have become an integral part in helping humans do underwater explorations more efficiently and safely since these vehicles can stay underwater much longer than any human can possibly do and they require little or almost no human interaction. These vehicles are subject to dynamic and unpredictable nature of the underwater environment resulting to complexities in their navigation. This paper proposes a fuzzy logic-based controller to allow the vehicle to navigate autonomously while avoiding obstacles. The said controller is implemented in an actual low-cost underwater vehicle equipped with magnetometer and ultrasonic sensors. The intelligence of the UUV includes a two fuzzy logic block, namely Motion Control block and Heading Correction block. The fuzzy logic controller takes in target positions in X, Y and Z axes. Also, the heading error and rate of heading error are included as inputs in order to correct the bearing or direction of the vehicle. A heuristic and integration stage is also included after these fuzzy logic blocks for vehicle's collision avoidance. The controller output parameters are the adjusted thrusters' speeds which dictate the six thrusters speed and direction. With the proper output commands from this controller, the vehicle is able to navigate in its predefined destination.
Article title: Multiple Level Fuzzy Logic-based Artificial Intelligence for Multi-Agent Cooperative Robot Platform
Authors: Ryan Rhay P. Vicerra, Kanny Krizzy David, Kristan Bryan Caluntad Simbulan, Argel Bandala, et al.
Conference title: 2014 IEEE Region 10 Symposium
Abstract:
Fuzzy Logic is a many valued logic and it is very similar to human reasoning which is not binary. It uses approximate measures rather than exact, making it suitable for linguistic variable and analysis. It has been applied to many applications in artificial intelligence, control and robotics. In this paper, the authors develop a pure fuzzy logic artificial intelligence for a dynamic robot platform with multiple robots and multiple identity assignment which means that each robot will have its distinct behavior. In order to design pure fuzzy logic artificial intelligence, we applied fuzzy logic multiple times calling each of them as a fuzzy logic block. These fuzzy logic blocks can be seen in different parallel and series configuration making it multilevel in structure. Furthermore, there is multiple input - multiple output (MIMO) fuzzy logic implementation in one of our several fuzzy logic blocks, this is necessary in order to utilize pure fuzzy logic control in the whole artificial intelligence. The mult-iagent cooperative robot platform we choose to test our artificial intelligence is a multiple robot system for FIRA Micro-Robot World Soccer Tournament (MiroSot). The system requires complex intelligence as its individual agents must perform specific tasks in a dynamic environment, unlike other systems which duplicates a single task for all the agents. In our setup, we use three robots and gave them three different identities; the Forward, the Back and the Goal-keeper. The robot identity assignment is very dynamic and depends on the position of each robot with respect to the position of the object they are pursuing. The developers have to tune each fuzzy logic blocks individually by isolating each one from the other. Some tuning procedures are performed in a simulator while most of them are tuned in real time in the actual platform. Although tuning procedures are rigorous, the linguistic approach and human reasoning nature of fuzzy logic made it easy for the developers to achieve its completion. The multiple trial and error tuning enhanced the developers understanding of how fuzzy logic and the overall system works. Overall, the proposed artificial intelligence produced favorable response based on the expected outcome and experimentations.
Article title: Simultaneous Face Detection and Recognition Using Viola-Jones Algorithm and Artificial Neural Networks for Identity Verification
Authors: Argel Bandala, Ma. Christina D. Fernandez, Kristina Joyce E. Gob, Aubrey Rose M. Leonidas, et al.
Conference title: 2014 IEEE Region 10 Symposium
Abstract:
The study presented aims to design and develop a face recognition system. The system utilized Viola Jones Algorithm in detecting faces from a given image. Also the system used Artificial Neural Networks in recognizing faces detected from the input. Upon experimentation the system generated can recognize human faces with accuracy of 87.05%. The system performs at its best if the person is around 150cm away from the camera with an accuracy rate of 87.59%. Also, the best amount of lighting for the recognition system is at 480 lumens with an accuracy rate of 88.64%. Lastly, the system also performs at its best if the person is directly facing the camera or at 0 degrees with respect to the camera.
Article title: Object recognition and detection by shape and color pattern recognition utilizing Artificial Neural Networks
Authors: Jerome Paul N. Cruz, Ma. Lourdes Dimaala, Argel Bandala, Erica Joanna S. Franco, et al.
Conference title: 2014 IEEE Region 10 Symposium
Abstract:
A robust and accurate object recognition tool is presented in this paper. The paper introduced the use of Artificial Neural Networks in evaluating a frame shot of the target image. The system utilizes three major steps in object recognition, namely image processing, ANN processing and interpretation. In image processing stage a frame shot or an image go through a process of extracting numerical values of object's shape and object's color. These values are then fed to the Artificial Neural Network stage, wherein the recognition of the object is done. Since the output of the ANN stage is in numerical form the third process is indispensable for human understanding. This stage simply converts a given value to its equivalent linguistic term. All three components are integrated in an interface for ease of use. Upon the conclusion of the system's development, experimentation and testing procedures are initiated. The study proved that the optimum lighting condition opted for the system is at 674 lumens with an accuracy of 99.99996072%. Another finding that the paper presented is that the optimum distance for recognition is at 40cm with an accuracy of 99.99996072%. Lastly the system contains a very high tolerance in the variations in the objects position or orientation, with the optimum accuracy at upward position with 99.99940181% accuracy rate.
Article title: Development and design of mobile robot with IP-based vision system
Authors: Argel Bandala, Elmer P. Dadios
Conference title: TENCON 2012 - 2012 IEEE Region 10 Conference, November 2012
Abstract:
Hardware, firmware and software design of a mobile robot capable of transmitting video information and receiving commands from a controlling point is presented. The hardware design is composed of a PIC18F4620 microcontroller, a UCC27525 MOSFET gate driver, XBee Series 2 OEM RF Module. Firmware design includes the reception, processing and decoding of Zigbee API packets. Based on this decoded information the microcontroller will generate signals to move the motors namely left and right motors with a corresponding direction, either clockwise of counter clockwise. The software part includes the graphical user interface which generates commands sent to the mobile robot. The images from the mobile robot are sent to the central controller. The images are then processed and a command is generated. The command is formatted in API format and then sent to the mobile robot. Testing of the system is done by experimentation. Three parameters are tested which are influenced by four parameters. Image recognition is measured while varying the distance. Also image recognition is measured while varying the luminance of the environment. The received signal level is measured while varying the distance. Lastly the accuracy of the movement of the mobile robot is also measured while varying the target position. The results showed that the distance used by the researcher has no significant effect on image recognition. The results showed also that image recognition is unaffected with the luminance of 230-1590 lumens. The mobile robot will respond in an optimum range of one meter and can respond from one to ten meters.
Full text available upon request to the author
Article title: Vehicle parking inventory system utilizing image recognition through artificial neural networks
Authors: Leo S. Bartolome, Argel Bandala, Cesar Llorente, Elmer P. Dadios
Conference title: TENCON 2012 - 2012 IEEE Region 10 Conference, November 2012
Abstract:
An automated vehicle logging system is introduced in this paper. The system utilizes character recognition through images captured from the entrance of a parking area. These images are processed to extract the licensed plates of any vehicle entering the parking area. Extracted plates images are then converted into numerical forms devised by researchers to fit the requirements of the artificial neural network. From the numbered plate, each character is then extracted to produce their distinct features. Character recognition engine is primarily implemented using feed forward neural networks. There are 50 input neurons which are defined by resizing each character into 25×25 pixel image and summing all the pixel values in each row and each columns resulting to 50 sums. After which a numerical value will be produce and will signify a character equivalent. Characters are recognized separately. This process is done until all of the characters are recognized. Afterwards, these characters are then concatenated to produce the plate number identity. The system is trained using 5860 sets of training data yielding a system with 0.0001645724% error.
Full text available upon request to the author