Neuro-fuzzy systems have been used for robot navigation applications because of their ability to exert human like expertise and to utilize acquired knowledge to develop autonomous navigation strategies. In this paper, neuro-fuzzy based system is proposed for reactive navigation of a mobile robot using behavior based control. The proposed algorithm uses discrete sampling based optimal training of neural network. With a view to ascertain the efficacy of proposed system; the proposed neuro-fuzzy system’s performance is compared to that of neural and fuzzy based approaches. Simulation results along with detailed behavior analysis show effectiveness of our algorithm in all kind of obstacle environments.
IRJET- Path Finder with Obstacle Avoidance RobotIRJET Journal
This document presents a robot that can find a safe path and avoid obstacles. It uses an infrared sensor to detect obstacles in its path. When an obstacle is detected, the robot changes direction to avoid the obstacle and moves towards its destination. The system architecture includes infrared sensors, a microcontroller, and motors. When an obstacle is detected by the infrared sensor, the microcontroller processes the input and redirects the robot using motors controlled by motor drivers, allowing the robot to avoid collisions and safely reach its target location.
A Multi-robot System Coordination Design and Analysis on Wall Follower Robot ...IJECEIAES
In this research, multi-robot formation can be established according to the environment or workspace. Group of robots will move sequently if there is no space for robots to stand side by side. Leader robot will be on the front of all robots and follow the right wall. On the other hand, robots will move side by side if there is a large space between them. Leader robot will be tracked the wall on its right side and follow on it while every follower moves side by side. The leader robot have to broadcast the information to all robots in the group in radius 9 meters. Nevertheless, every robot should be received information from leader robot to define their movements in the area. The error provided by fuzzy output process which is caused by read data from ultrasound sensor will drive to more time process. More sampling can reduce the error but it will drive more execution time. Furthermore, coordination time will need longer time and delay. Formation will not be establisehed if packet error happened in the communication process because robot will execute wrong command.
An overview on Advanced Research Works on Brain-Computer InterfaceWaqas Tariq
A brain–computer interface (BCI) is a proficient result in the research field of human- computer synergy, where direct articulation between brain and an external device occurs resulting in augmenting, assisting and repairing human cognitive. Advanced works like generating brain-computer interface switch technologies for intermittent (or asynchronous) control in natural environments or developing brain-computer interface by Fuzzy logic Systems or by implementing wavelet theory to drive its efficacies are still going on and some useful results has also been found out. The requirements to develop this brain machine interface is also growing day by day i.e. like neuropsychological rehabilitation, emotion control, etc. An overview on the control theory and some advanced works on the field of brain machine interface are shown in this paper.
International Journal of Image Processing (IJIP) Volume (1) Issue (2)CSCJournals
This document summarizes a research paper on a face recognition system that uses a multi-local feature selection approach. The proposed system consists of five stages: face detection, extraction of facial features like eyes, nose and mouth, generation of moments to represent the features, classification of facial features using RBF neural networks, and face identification. The system was tested on over 3000 images from three facial databases and achieved recognition rates over 89%, outperforming global feature-based and single local feature approaches. The technique was also found to be robust to variations in translation, orientation and scaling.
Identifying Gender from Facial Parts Using Support Vector Machine ClassifierEditor IJCATR
Gender classification can be stated as inferring female or male from a collection of facial images. There exist different
methods for gender classification, such as gait, iris, hand shape and hair, it is probably better way to find out gender based on facial
features. In this paper SVM basic kernel function has been employed firstly to detect and classify the human gender Image into
two labels i.e. (1) male and (2) female. The gender classifier achieves over 96% accuracy.
IRJET- Behavior Analysis from Videos using Motion based Feature ExtractionIRJET Journal
This document proposes a technique for analyzing human behavior in videos using motion-based feature extraction. It discusses how previous approaches have used spatial and temporal features to detect abnormal behaviors. The proposed approach extracts motion features from videos to represent each video with a single feature vector, rather than extracting features from each individual frame. This reduces the feature space and unnecessary information. The technique involves preprocessing videos into frames, extracting motion features, using KNN classification on the features to classify behaviors as normal or abnormal, and evaluating the method's performance on various metrics like accuracy, recall, and precision. Testing on fight and riot datasets showed the motion-based approach achieved higher accuracy, recall, precision and F-measure than a non-motion based approach.
Visual Saliency Model Using Sift and Comparison of Learning Approachescsandit
This document discusses a study that aims to develop a visual saliency model to predict where humans look in images. It uses the SIFT feature in addition to low, mid, and high-level image features to train machine learning models on an eye-tracking dataset. Support vector machines (SVM) achieved the best performance, accurately predicting fixations 88% of the time. Including the SIFT feature further improved SVM performance to 91% accuracy. The study evaluates different machine learning methods and determines SVM to be best suited for this binary classification task using high-dimensional image data.
Project on collision avoidance in static and dynamic environmentgopaljee1989
This document proposes a method for real-time path planning of mobile robots to avoid collisions in static and dynamic environments. It involves using an artificial neural network (ANN) for obstacle detection and a vector field histogram method with decision trees for local path planning. When all paths are blocked, fuzzy logic will be used to choose a path. The objective is for the robot to safely navigate around obstacles without communication delays. The proposed method aims to overcome limitations of previous techniques by applying ANN, decision trees, and fuzzy logic for robust, real-time collision avoidance.
IRJET- Path Finder with Obstacle Avoidance RobotIRJET Journal
This document presents a robot that can find a safe path and avoid obstacles. It uses an infrared sensor to detect obstacles in its path. When an obstacle is detected, the robot changes direction to avoid the obstacle and moves towards its destination. The system architecture includes infrared sensors, a microcontroller, and motors. When an obstacle is detected by the infrared sensor, the microcontroller processes the input and redirects the robot using motors controlled by motor drivers, allowing the robot to avoid collisions and safely reach its target location.
A Multi-robot System Coordination Design and Analysis on Wall Follower Robot ...IJECEIAES
In this research, multi-robot formation can be established according to the environment or workspace. Group of robots will move sequently if there is no space for robots to stand side by side. Leader robot will be on the front of all robots and follow the right wall. On the other hand, robots will move side by side if there is a large space between them. Leader robot will be tracked the wall on its right side and follow on it while every follower moves side by side. The leader robot have to broadcast the information to all robots in the group in radius 9 meters. Nevertheless, every robot should be received information from leader robot to define their movements in the area. The error provided by fuzzy output process which is caused by read data from ultrasound sensor will drive to more time process. More sampling can reduce the error but it will drive more execution time. Furthermore, coordination time will need longer time and delay. Formation will not be establisehed if packet error happened in the communication process because robot will execute wrong command.
An overview on Advanced Research Works on Brain-Computer InterfaceWaqas Tariq
A brain–computer interface (BCI) is a proficient result in the research field of human- computer synergy, where direct articulation between brain and an external device occurs resulting in augmenting, assisting and repairing human cognitive. Advanced works like generating brain-computer interface switch technologies for intermittent (or asynchronous) control in natural environments or developing brain-computer interface by Fuzzy logic Systems or by implementing wavelet theory to drive its efficacies are still going on and some useful results has also been found out. The requirements to develop this brain machine interface is also growing day by day i.e. like neuropsychological rehabilitation, emotion control, etc. An overview on the control theory and some advanced works on the field of brain machine interface are shown in this paper.
International Journal of Image Processing (IJIP) Volume (1) Issue (2)CSCJournals
This document summarizes a research paper on a face recognition system that uses a multi-local feature selection approach. The proposed system consists of five stages: face detection, extraction of facial features like eyes, nose and mouth, generation of moments to represent the features, classification of facial features using RBF neural networks, and face identification. The system was tested on over 3000 images from three facial databases and achieved recognition rates over 89%, outperforming global feature-based and single local feature approaches. The technique was also found to be robust to variations in translation, orientation and scaling.
Identifying Gender from Facial Parts Using Support Vector Machine ClassifierEditor IJCATR
Gender classification can be stated as inferring female or male from a collection of facial images. There exist different
methods for gender classification, such as gait, iris, hand shape and hair, it is probably better way to find out gender based on facial
features. In this paper SVM basic kernel function has been employed firstly to detect and classify the human gender Image into
two labels i.e. (1) male and (2) female. The gender classifier achieves over 96% accuracy.
IRJET- Behavior Analysis from Videos using Motion based Feature ExtractionIRJET Journal
This document proposes a technique for analyzing human behavior in videos using motion-based feature extraction. It discusses how previous approaches have used spatial and temporal features to detect abnormal behaviors. The proposed approach extracts motion features from videos to represent each video with a single feature vector, rather than extracting features from each individual frame. This reduces the feature space and unnecessary information. The technique involves preprocessing videos into frames, extracting motion features, using KNN classification on the features to classify behaviors as normal or abnormal, and evaluating the method's performance on various metrics like accuracy, recall, and precision. Testing on fight and riot datasets showed the motion-based approach achieved higher accuracy, recall, precision and F-measure than a non-motion based approach.
Visual Saliency Model Using Sift and Comparison of Learning Approachescsandit
This document discusses a study that aims to develop a visual saliency model to predict where humans look in images. It uses the SIFT feature in addition to low, mid, and high-level image features to train machine learning models on an eye-tracking dataset. Support vector machines (SVM) achieved the best performance, accurately predicting fixations 88% of the time. Including the SIFT feature further improved SVM performance to 91% accuracy. The study evaluates different machine learning methods and determines SVM to be best suited for this binary classification task using high-dimensional image data.
Project on collision avoidance in static and dynamic environmentgopaljee1989
This document proposes a method for real-time path planning of mobile robots to avoid collisions in static and dynamic environments. It involves using an artificial neural network (ANN) for obstacle detection and a vector field histogram method with decision trees for local path planning. When all paths are blocked, fuzzy logic will be used to choose a path. The objective is for the robot to safely navigate around obstacles without communication delays. The proposed method aims to overcome limitations of previous techniques by applying ANN, decision trees, and fuzzy logic for robust, real-time collision avoidance.
EFFECTIVE REDIRECTING OF THE MOBILE ROBOT IN A MESSED ENVIRONMENT BASED ON TH...Wireilla
The use of fuzzy logic in redirecting mobile robot is based on two sets of received information. First set is
the instantaneous distance of the robot from the obstacle and second set is the instantaneous information of
the robot's position. For this purpose, the fuzzy rules base consists of forty-two bases, which is extracted
based on the robot's distance from obstacles, and the target position relative to the instantaneous
orientation of the robot. In the structure of fuzzy systems, minimal inference engine are considered. Also,
Extended Kalman filter is used for localization in a noisy environment. Accordingly, the inputs of the fuzzy
systems are determined based on the estimation of the localization process, the information of the obstacles
center and the target position. Also, the linear acceleration and instantaneous orientation of the mobile
robot are determined by the desired fuzzy structures which are applied to its kinematic model.
EFFECTIVE REDIRECTING OF THE MOBILE ROBOT IN A MESSED ENVIRONMENT BASED ON TH...ijfls
The use of fuzzy logic in redirecting mobile robot is based on two sets of received information. First set is the instantaneous distance of the robot from the obstacle and second set is the instantaneous information of the robot's position. For this purpose, the fuzzy rules base consists of forty-two bases, which is extracted based on the robot's distance from obstacles, and the target position relative to the instantaneous orientation of the robot. In the structure of fuzzy systems, minimal inference engine are considered. Also, Extended Kalman filter is used for localization in a noisy environment. Accordingly, the inputs of the fuzzy systems are determined based on the estimation of the localization process, the information of the obstacles center and the target position. Also, the linear acceleration and instantaneous orientation of the mobile robot are determined by the desired fuzzy structures which are applied to its kinematic model.
MULTI-LEVEL FEATURE FUSION BASED TRANSFER LEARNING FOR PERSON RE-IDENTIFICATIONijaia
Most of the currently known methods treat person re-identification task as classification problem and used commonly neural networks. However, these methods used only high-level convolutional feature or to express the feature representation of pedestrians. Moreover, the current data sets for person reidentification is relatively small. Under the limitation of the number of training set, deep convolutional networks are difficult to train adequately. Therefore, it is very worthwhile to introduce auxiliary data sets to help training. In order to solve this problem, this paper propose a novel method of deep transfer learning, and combines the comparison model with the classification model and multi-level fusion of the convolution features on the basis of transfer learning. In a multi-layers convolutional network, the characteristics of each layer of network are the dimensionality reduction of the previous layer of results, but the information of multi-level features is not only inclusive, but also has certain complementarity. We can using the information gap of different layers of convolutional neural networks to extract a better feature expression. Finally, the algorithm proposed in this paper is fully tested on four data sets (VIPeR, CUHK01, GRID and PRID450S). The obtained re-identification results prove the effectiveness of the algorithm.
This document summarizes a research paper that proposes a method for detecting and recognizing faces using the Viola Jones algorithm and Back Propagation Neural Network (BPNN).
The paper first discusses face detection and recognition challenges. It then provides background on Viola Jones algorithm and BPNN. The proposed methodology uses Viola Jones for face detection, converts the image to grayscale and binary, then trains segments or the whole image with BPNN. Results are analyzed using training, testing and validation curves in the MATLAB neural network tool to minimize error. In under 3 sentences, this document outlines the key techniques, proposed method, and analysis approach discussed in the source research paper.
Vision Based Gesture Recognition Using Neural Networks Approaches: A ReviewWaqas Tariq
The aim of gesture recognition researches is to create system that easily identifies gestures, and use them for device control, or convey in formations. In this paper we are discussing researches done in the area of hand gesture recognition based on Artificial Neural Networks approaches. Several hand gesture recognition researches that use Neural Networks are discussed in this paper, comparisons between these methods were presented, advantages and drawbacks of the discussed methods also included, and implementation tools for each method were presented as well.
A Study of Mobile User Movements Prediction Methods IJECEIAES
For a decade and more, the Number of smart phone users count increasing day by day. With the drastic improvements in Communication technologies, the prediction of future movements of mobile users needs also have important role. Various sectors can gain from this prediction. Communication management, City Development planning, and locationbased services are some of the fields that can be made more valuable with movement prediction. In this paper, we propose a study of several Location Prediction Techniques in the following areas.
Optimizing Mobile Robot Path Planning and Navigation by Use of Differential E...IOSR Journals
Abstract: Path planning and navigation is essential for an autonomous robot which can move avoiding the
static obstacles in a real world and to reach the specific target. Optimizing path for the robot movement gives
the optimal distance from the source to the target and save precious time as well. With the development of
various evolutionary algorithms, the differential evolution is taking the pace in comparison to genetic algorithm.
Differential evolution has been deployed quite successfully for solving global optimization problem. Differential
evolution is a very simple yet powerful metaheuristics type problem solving method. In this paper we are
proposing a Differential Evolution based path navigation algorithm for mobile path navigation and analyze its
efficiency with other developed approaches. The proposed algorithm optimized the robot path and navigates the
robot to the proper target efficiently.
EFFICIENT FEATURE SUBSET SELECTION MODEL FOR HIGH DIMENSIONAL DATAIJCI JOURNAL
This paper proposes a new method that intends on reducing the size of high dimensional dataset by
identifying and removing irrelevant and redundant features. Dataset reduction is important in the case of
machine learning and data mining. The measure of dependence is used to evaluate the relationship
between feature and target concept and or between features for irrelevant and redundant feature removal.
The proposed work initially removes all the irrelevant features and then a minimum spanning tree of
relevant features is constructed using Prim’s algorithm. Splitting the minimum spanning tree based on the
dependency between features leads to the generation of forests. A representative feature from each of the
forests is taken to form the final feature subset
ROBUST STATISTICAL APPROACH FOR EXTRACTION OF MOVING HUMAN SILHOUETTES FROM V...ijitjournal
Human pose estimation is one of the key problems in computer visionthat has been studied in the recent
years. The significance of human pose estimation is in the higher level tasks of understanding human
actions applications such as recognition of anomalous actions present in videos and many other related
applications. The human poses can be estimated by extracting silhouettes of humans as silhouettes are
robust to variations and it gives the shape information of the human body. Some common challenges
include illumination changes, variation in environments, and variation in human appearances. Thus there
is a need for a robust method for human pose estimation. This paper presents a study and analysis of
approaches existing for silhouette extraction and proposes a robust technique for extracting human
silhouettes in video sequences. Gaussian Mixture Model (GMM) A statistical approach is combined with
HSV (Hue, Saturation and Value) color space model for a robust background model that is used for
background subtraction to produce foreground blobs, called human silhouettes. Morphological operations
are then performed on foreground blobs from background subtraction. The silhouettes obtained from this
work can be used in further tasks associated with human action interpretation and activity processes like
human action classification, human pose estimation and action recognition or action interpretation.
Crowd Recognition System Based on Optical Flow Along with SVM classifierIJECEIAES
The manuscript discusses about abnormalities in a crowded scenario. To prevent the mishap at a public place, there is no much mechanism which could prevent or alert the concerned authority about suspects in a crowd. Usually in a crowded scene, there are chances of some mishap like a terrorist attack or a crime. Our target is finding techniques to identify such activities and to possibly prevent them. If the crowd members exhibit abnormal behavior, we could identify and say that this particular person is a suspect and then the concerned authority would look into the matter. There are various methods to identify the abnormal behavior. The proposed approach is based on optical flow model. It has an ability to detect the sudden changes in motion of an individual among the crowd. First, the main region of motion is extracted by the help of motion heat map. Harris corner detector is used for extracting point of interest of extracted motion area. Based on the point of interest an optical flow is estimated here. After analyzing this optical flow model, a threshold value is fixed. Basically optical flow is an energy level of individual frame. The threshold value is forwarded to SVM classifier, which produces a better result with 99.71% accuracy. This approach is very useful in real time video surveillance system where a machine can monitor unwanted crowd activity.
Measuring similarity between mobility models and real world motion trajectoriescsandit
Various mobility models have been proposed to represent the motion behaviour of mobile nodes
in the real world. Selection of the most similar mobility model to a given real world environment
is a challenging issue which has a significant impact on the quality of performance evaluation
of different network protocols. In this paper we propose a methodology for measurement of
similarity between mobility models used in mobile networks simulation and real world mobility
scenarios with different transportation modes. We explain our mobility metrics we have used for
analysis of motion behavior of mobile nodes and a pre-processing method which makes our
trajectories suitable for extraction and calculation of these metrics considering shape of the
road networks and GPS noise. Then we use a feature selection method to find the most
discriminative features which are able to distinguish between trajectories with different
transportation modes using a supervised learning and feature ranking method. Subsequently,
using our selected feature space we perform Fuzzy C-means Clustering to find the degree of
similarity between each of our mobility models and real world trajectories with different
transportation modes. Our methodology can be used to select the most similar mobility model
suitable for simulation of mobile network protocols (such as DTN and MANETs protocols) in a
particular real world area.
Schematic model for analyzing mobility and detection of multipleIAEME Publication
The document discusses a schematic model for analyzing mobility and detecting multiple objects in traffic scenes. It aims to not only detect and count moving objects, but also understand crowd behavior and reduce issues with objects occluding each other. Previous work on object detection is reviewed, noting that most approaches do not integrate detecting multiple objects simultaneously or address problems of object occlusion. The proposed model uses background subtraction and unscented Kalman filtering to increase detection accuracy and reduce false positives when analyzing image sequences of traffic scenes to detect multiple moving objects. It was tested in MATLAB and results showed highly accurate detection rates.
The document describes a new hierarchical deep learning algorithm for facial expression recognition (FER). The algorithm extracts appearance features from preprocessed LBP images using a convolutional neural network. It also extracts geometric features by tracking the coordinates of action unit landmarks, which are facial muscles involved in expressions. These two types of features are fused in a hierarchical structure. The algorithm combines the softmax outputs of each network by considering the second-highest predicted emotion. It also uses an autoencoder to generate neutral expression images to help extract dynamic features between neutral and emotional expressions. The algorithm achieved 96.46% accuracy on the CK+ dataset and 91.27% on the JAFFE dataset, outperforming other recent FER methods.
Survey on video object detection & trackingijctet
This document summarizes previous work on video object detection and tracking techniques. It discusses research papers that used techniques like active contour modeling, gradient-based attraction fields, neural fuzzy networks, and region-based contour extraction for object tracking. Background subtraction, frame differencing, optical flow, spatio-temporal features, Kalman filtering, and contour tracking are described as common video object detection techniques. The challenges of multi-object data association and state estimation for tracking multiple objects are also mentioned.
ADOPTING AND IMPLEMENTATION OF SELF ORGANIZING FEATURE MAP FOR IMAGE FUSIONijistjournal
A different image fusion algorithm based on self organizing feature map is proposed in this paper, aiming to produce quality images. Image Fusion is to integrate complementary and redundant information from multiple images of the same scene to create a single composite image that contains all the important features of the original images. The resulting fused image will thus be more suitable for human and machine perception or for further image processing tasks. The existing fusion techniques based on either direct operation on pixels or segments fail to produce fused images of the required quality and are mostly application based. The existing segmentation algorithms become complicated and time consuming when multiple images are to be fused. A new method of segmenting and fusion of gray scale images adopting Self organizing Feature Maps(SOM) is proposed in this paper. The Self Organizing Feature Maps is adopted to produce multiple slices of the source and reference images based on various combination of gray scale and can dynamically fused depending on the application. The proposed technique is adopted and analyzed for fusion of multiple images. The technique is robust in the sense that there will be no loss in information due to the property of Self Organizing Feature Maps; noise removal in the source images done during processing stage and fusion of multiple images is dynamically done to get the desired results. Experimental results demonstrate that, for the quality multifocus image fusion, the proposed method performs better than some popular image fusion methods in both subjective and objective qualities.
International Journal of Engineering Research and Applications (IJERA) is an open access online peer reviewed international journal that publishes research and review articles in the fields of Computer Science, Neural Networks, Electrical Engineering, Software Engineering, Information Technology, Mechanical Engineering, Chemical Engineering, Plastic Engineering, Food Technology, Textile Engineering, Nano Technology & science, Power Electronics, Electronics & Communication Engineering, Computational mathematics, Image processing, Civil Engineering, Structural Engineering, Environmental Engineering, VLSI Testing & Low Power VLSI Design etc.
This document presents an audio-visual emotion recognition system that uses multiple modalities and machine learning techniques. It extracts audio features like MFCCs and visual features like facial landmarks from video clips. It uses classifiers like CNNs and stacks their confidence outputs to predict emotions. The system achieves state-of-the-art performance on several databases according to experiments. It represents an improvement over previous work by combining audio, visual and classifier fusion approaches for multimodal emotion recognition.
IRJET- Recognition of Human Action Interaction using Motion History ImageIRJET Journal
This document discusses a method for recognizing human actions and interactions using motion history images and machine learning. It begins with an abstract describing the goal of accurately detecting human activities from videos using data mining techniques. It then provides more details on the proposed method, which uses background subtraction, motion history images, local binary patterns, histogram of oriented gradients, k-means clustering, and support vector machines. Key frames are extracted and motion history images are generated only for those frames to improve recognition rates. The method is evaluated on a dataset containing videos of different human interactions like beating, caring, pushing and punching.
This document compares the dimensions of a ping pong table and tennis court. A ping pong table measures 5 feet by 9 feet while a tennis court is 27 feet by 78 feet. These dimensions are not proportional as dividing the tennis court measurements by the ping pong table measurements results in ratios of 1:5.4 for width and 1:8.6 for length.
EFFECTIVE REDIRECTING OF THE MOBILE ROBOT IN A MESSED ENVIRONMENT BASED ON TH...Wireilla
The use of fuzzy logic in redirecting mobile robot is based on two sets of received information. First set is
the instantaneous distance of the robot from the obstacle and second set is the instantaneous information of
the robot's position. For this purpose, the fuzzy rules base consists of forty-two bases, which is extracted
based on the robot's distance from obstacles, and the target position relative to the instantaneous
orientation of the robot. In the structure of fuzzy systems, minimal inference engine are considered. Also,
Extended Kalman filter is used for localization in a noisy environment. Accordingly, the inputs of the fuzzy
systems are determined based on the estimation of the localization process, the information of the obstacles
center and the target position. Also, the linear acceleration and instantaneous orientation of the mobile
robot are determined by the desired fuzzy structures which are applied to its kinematic model.
EFFECTIVE REDIRECTING OF THE MOBILE ROBOT IN A MESSED ENVIRONMENT BASED ON TH...ijfls
The use of fuzzy logic in redirecting mobile robot is based on two sets of received information. First set is the instantaneous distance of the robot from the obstacle and second set is the instantaneous information of the robot's position. For this purpose, the fuzzy rules base consists of forty-two bases, which is extracted based on the robot's distance from obstacles, and the target position relative to the instantaneous orientation of the robot. In the structure of fuzzy systems, minimal inference engine are considered. Also, Extended Kalman filter is used for localization in a noisy environment. Accordingly, the inputs of the fuzzy systems are determined based on the estimation of the localization process, the information of the obstacles center and the target position. Also, the linear acceleration and instantaneous orientation of the mobile robot are determined by the desired fuzzy structures which are applied to its kinematic model.
MULTI-LEVEL FEATURE FUSION BASED TRANSFER LEARNING FOR PERSON RE-IDENTIFICATIONijaia
Most of the currently known methods treat person re-identification task as classification problem and used commonly neural networks. However, these methods used only high-level convolutional feature or to express the feature representation of pedestrians. Moreover, the current data sets for person reidentification is relatively small. Under the limitation of the number of training set, deep convolutional networks are difficult to train adequately. Therefore, it is very worthwhile to introduce auxiliary data sets to help training. In order to solve this problem, this paper propose a novel method of deep transfer learning, and combines the comparison model with the classification model and multi-level fusion of the convolution features on the basis of transfer learning. In a multi-layers convolutional network, the characteristics of each layer of network are the dimensionality reduction of the previous layer of results, but the information of multi-level features is not only inclusive, but also has certain complementarity. We can using the information gap of different layers of convolutional neural networks to extract a better feature expression. Finally, the algorithm proposed in this paper is fully tested on four data sets (VIPeR, CUHK01, GRID and PRID450S). The obtained re-identification results prove the effectiveness of the algorithm.
This document summarizes a research paper that proposes a method for detecting and recognizing faces using the Viola Jones algorithm and Back Propagation Neural Network (BPNN).
The paper first discusses face detection and recognition challenges. It then provides background on Viola Jones algorithm and BPNN. The proposed methodology uses Viola Jones for face detection, converts the image to grayscale and binary, then trains segments or the whole image with BPNN. Results are analyzed using training, testing and validation curves in the MATLAB neural network tool to minimize error. In under 3 sentences, this document outlines the key techniques, proposed method, and analysis approach discussed in the source research paper.
Vision Based Gesture Recognition Using Neural Networks Approaches: A ReviewWaqas Tariq
The aim of gesture recognition researches is to create system that easily identifies gestures, and use them for device control, or convey in formations. In this paper we are discussing researches done in the area of hand gesture recognition based on Artificial Neural Networks approaches. Several hand gesture recognition researches that use Neural Networks are discussed in this paper, comparisons between these methods were presented, advantages and drawbacks of the discussed methods also included, and implementation tools for each method were presented as well.
A Study of Mobile User Movements Prediction Methods IJECEIAES
For a decade and more, the Number of smart phone users count increasing day by day. With the drastic improvements in Communication technologies, the prediction of future movements of mobile users needs also have important role. Various sectors can gain from this prediction. Communication management, City Development planning, and locationbased services are some of the fields that can be made more valuable with movement prediction. In this paper, we propose a study of several Location Prediction Techniques in the following areas.
Optimizing Mobile Robot Path Planning and Navigation by Use of Differential E...IOSR Journals
Abstract: Path planning and navigation is essential for an autonomous robot which can move avoiding the
static obstacles in a real world and to reach the specific target. Optimizing path for the robot movement gives
the optimal distance from the source to the target and save precious time as well. With the development of
various evolutionary algorithms, the differential evolution is taking the pace in comparison to genetic algorithm.
Differential evolution has been deployed quite successfully for solving global optimization problem. Differential
evolution is a very simple yet powerful metaheuristics type problem solving method. In this paper we are
proposing a Differential Evolution based path navigation algorithm for mobile path navigation and analyze its
efficiency with other developed approaches. The proposed algorithm optimized the robot path and navigates the
robot to the proper target efficiently.
EFFICIENT FEATURE SUBSET SELECTION MODEL FOR HIGH DIMENSIONAL DATAIJCI JOURNAL
This paper proposes a new method that intends on reducing the size of high dimensional dataset by
identifying and removing irrelevant and redundant features. Dataset reduction is important in the case of
machine learning and data mining. The measure of dependence is used to evaluate the relationship
between feature and target concept and or between features for irrelevant and redundant feature removal.
The proposed work initially removes all the irrelevant features and then a minimum spanning tree of
relevant features is constructed using Prim’s algorithm. Splitting the minimum spanning tree based on the
dependency between features leads to the generation of forests. A representative feature from each of the
forests is taken to form the final feature subset
ROBUST STATISTICAL APPROACH FOR EXTRACTION OF MOVING HUMAN SILHOUETTES FROM V...ijitjournal
Human pose estimation is one of the key problems in computer visionthat has been studied in the recent
years. The significance of human pose estimation is in the higher level tasks of understanding human
actions applications such as recognition of anomalous actions present in videos and many other related
applications. The human poses can be estimated by extracting silhouettes of humans as silhouettes are
robust to variations and it gives the shape information of the human body. Some common challenges
include illumination changes, variation in environments, and variation in human appearances. Thus there
is a need for a robust method for human pose estimation. This paper presents a study and analysis of
approaches existing for silhouette extraction and proposes a robust technique for extracting human
silhouettes in video sequences. Gaussian Mixture Model (GMM) A statistical approach is combined with
HSV (Hue, Saturation and Value) color space model for a robust background model that is used for
background subtraction to produce foreground blobs, called human silhouettes. Morphological operations
are then performed on foreground blobs from background subtraction. The silhouettes obtained from this
work can be used in further tasks associated with human action interpretation and activity processes like
human action classification, human pose estimation and action recognition or action interpretation.
Crowd Recognition System Based on Optical Flow Along with SVM classifierIJECEIAES
The manuscript discusses about abnormalities in a crowded scenario. To prevent the mishap at a public place, there is no much mechanism which could prevent or alert the concerned authority about suspects in a crowd. Usually in a crowded scene, there are chances of some mishap like a terrorist attack or a crime. Our target is finding techniques to identify such activities and to possibly prevent them. If the crowd members exhibit abnormal behavior, we could identify and say that this particular person is a suspect and then the concerned authority would look into the matter. There are various methods to identify the abnormal behavior. The proposed approach is based on optical flow model. It has an ability to detect the sudden changes in motion of an individual among the crowd. First, the main region of motion is extracted by the help of motion heat map. Harris corner detector is used for extracting point of interest of extracted motion area. Based on the point of interest an optical flow is estimated here. After analyzing this optical flow model, a threshold value is fixed. Basically optical flow is an energy level of individual frame. The threshold value is forwarded to SVM classifier, which produces a better result with 99.71% accuracy. This approach is very useful in real time video surveillance system where a machine can monitor unwanted crowd activity.
Measuring similarity between mobility models and real world motion trajectoriescsandit
Various mobility models have been proposed to represent the motion behaviour of mobile nodes
in the real world. Selection of the most similar mobility model to a given real world environment
is a challenging issue which has a significant impact on the quality of performance evaluation
of different network protocols. In this paper we propose a methodology for measurement of
similarity between mobility models used in mobile networks simulation and real world mobility
scenarios with different transportation modes. We explain our mobility metrics we have used for
analysis of motion behavior of mobile nodes and a pre-processing method which makes our
trajectories suitable for extraction and calculation of these metrics considering shape of the
road networks and GPS noise. Then we use a feature selection method to find the most
discriminative features which are able to distinguish between trajectories with different
transportation modes using a supervised learning and feature ranking method. Subsequently,
using our selected feature space we perform Fuzzy C-means Clustering to find the degree of
similarity between each of our mobility models and real world trajectories with different
transportation modes. Our methodology can be used to select the most similar mobility model
suitable for simulation of mobile network protocols (such as DTN and MANETs protocols) in a
particular real world area.
Schematic model for analyzing mobility and detection of multipleIAEME Publication
The document discusses a schematic model for analyzing mobility and detecting multiple objects in traffic scenes. It aims to not only detect and count moving objects, but also understand crowd behavior and reduce issues with objects occluding each other. Previous work on object detection is reviewed, noting that most approaches do not integrate detecting multiple objects simultaneously or address problems of object occlusion. The proposed model uses background subtraction and unscented Kalman filtering to increase detection accuracy and reduce false positives when analyzing image sequences of traffic scenes to detect multiple moving objects. It was tested in MATLAB and results showed highly accurate detection rates.
The document describes a new hierarchical deep learning algorithm for facial expression recognition (FER). The algorithm extracts appearance features from preprocessed LBP images using a convolutional neural network. It also extracts geometric features by tracking the coordinates of action unit landmarks, which are facial muscles involved in expressions. These two types of features are fused in a hierarchical structure. The algorithm combines the softmax outputs of each network by considering the second-highest predicted emotion. It also uses an autoencoder to generate neutral expression images to help extract dynamic features between neutral and emotional expressions. The algorithm achieved 96.46% accuracy on the CK+ dataset and 91.27% on the JAFFE dataset, outperforming other recent FER methods.
Survey on video object detection & trackingijctet
This document summarizes previous work on video object detection and tracking techniques. It discusses research papers that used techniques like active contour modeling, gradient-based attraction fields, neural fuzzy networks, and region-based contour extraction for object tracking. Background subtraction, frame differencing, optical flow, spatio-temporal features, Kalman filtering, and contour tracking are described as common video object detection techniques. The challenges of multi-object data association and state estimation for tracking multiple objects are also mentioned.
ADOPTING AND IMPLEMENTATION OF SELF ORGANIZING FEATURE MAP FOR IMAGE FUSIONijistjournal
A different image fusion algorithm based on self organizing feature map is proposed in this paper, aiming to produce quality images. Image Fusion is to integrate complementary and redundant information from multiple images of the same scene to create a single composite image that contains all the important features of the original images. The resulting fused image will thus be more suitable for human and machine perception or for further image processing tasks. The existing fusion techniques based on either direct operation on pixels or segments fail to produce fused images of the required quality and are mostly application based. The existing segmentation algorithms become complicated and time consuming when multiple images are to be fused. A new method of segmenting and fusion of gray scale images adopting Self organizing Feature Maps(SOM) is proposed in this paper. The Self Organizing Feature Maps is adopted to produce multiple slices of the source and reference images based on various combination of gray scale and can dynamically fused depending on the application. The proposed technique is adopted and analyzed for fusion of multiple images. The technique is robust in the sense that there will be no loss in information due to the property of Self Organizing Feature Maps; noise removal in the source images done during processing stage and fusion of multiple images is dynamically done to get the desired results. Experimental results demonstrate that, for the quality multifocus image fusion, the proposed method performs better than some popular image fusion methods in both subjective and objective qualities.
International Journal of Engineering Research and Applications (IJERA) is an open access online peer reviewed international journal that publishes research and review articles in the fields of Computer Science, Neural Networks, Electrical Engineering, Software Engineering, Information Technology, Mechanical Engineering, Chemical Engineering, Plastic Engineering, Food Technology, Textile Engineering, Nano Technology & science, Power Electronics, Electronics & Communication Engineering, Computational mathematics, Image processing, Civil Engineering, Structural Engineering, Environmental Engineering, VLSI Testing & Low Power VLSI Design etc.
This document presents an audio-visual emotion recognition system that uses multiple modalities and machine learning techniques. It extracts audio features like MFCCs and visual features like facial landmarks from video clips. It uses classifiers like CNNs and stacks their confidence outputs to predict emotions. The system achieves state-of-the-art performance on several databases according to experiments. It represents an improvement over previous work by combining audio, visual and classifier fusion approaches for multimodal emotion recognition.
IRJET- Recognition of Human Action Interaction using Motion History ImageIRJET Journal
This document discusses a method for recognizing human actions and interactions using motion history images and machine learning. It begins with an abstract describing the goal of accurately detecting human activities from videos using data mining techniques. It then provides more details on the proposed method, which uses background subtraction, motion history images, local binary patterns, histogram of oriented gradients, k-means clustering, and support vector machines. Key frames are extracted and motion history images are generated only for those frames to improve recognition rates. The method is evaluated on a dataset containing videos of different human interactions like beating, caring, pushing and punching.
This document compares the dimensions of a ping pong table and tennis court. A ping pong table measures 5 feet by 9 feet while a tennis court is 27 feet by 78 feet. These dimensions are not proportional as dividing the tennis court measurements by the ping pong table measurements results in ratios of 1:5.4 for width and 1:8.6 for length.
The document provides information on the 2014 Public Health Advisory Council and Board of Health for Snohomish County, Washington. It lists the members of each group and notes that the Snohomish Health District works with these partners to promote public health in the community. The document then summarizes the Health District's strategic plan update and priorities for evolving its programs and services. It also provides some statistics on the services provided and funding challenges faced by the Health District.
Beyond Commodities - Gulf investors and the new AfricaJoannes Mongardini
The document examines Sub-Saharan Africa's growth outside of natural resources and commodities, and the role of Gulf investors. It finds that Africa has proven more resilient to global headwinds than expected, supported by demographic trends, a growing middle class, and economic reforms. East Africa is emerging as the most appealing region for non-commodity Gulf investment, particularly in manufacturing, retail, tourism, and education. Gulf investors have potential options like co-investing with private equity or direct acquisitions. Retail, tourism, and improving logistics networks represent opportunities, but challenges remain in some sectors and countries.
Dr. Ravi R Kasliwal’s personality is truly multifaceted- his list of honors is testimony of this. His passion for developing and applying non-invasive techniques for early detection and prevention of coronary artery diseases is exemplary.
Scaling with Postgres (Highload++ 2010)Robert Treat
The document discusses strategies for scaling PostgreSQL databases. It recommends using PostgreSQL version 8.3 or higher, implementing connection pooling, and setting up replication between a master and slave database. Monitoring and gaining visibility into database metrics and queries is also emphasized as important for capacity planning and performance tuning during scaling. The presenter advocates for a culture where application developers work closely with database administrators on schema design and queries.
Managing Databases In A DevOps EnvironmentRobert Treat
There’s a lot of talk in the devops world about bringing developer concepts to system administration, and discussion the other way about bringing the awareness of operations to developers, but a lot of the conversation leaves out what is often the most critical part of your technology stack: the database. Perhaps that’s because DBA’s have always had to keep one foot in development and one in production, before there was a devops. Or maybe DBA’s just suck at playing well with others. Bottom line; it doesn’t matter. If you are going to store data, you need a plan that both developers and operations people can understand and embrace.
At OmniTI we’ve worked with many of the leaders in the devops movement and we’ve found there are commonalties across these organizations. It’s not so much about the tools, but about the techniques they use to help people break down barriers between different roles and establish a common ownership of technology within their organizations.
Monitoring and visibility, managing schema changes and production pushes, optimization, configuration and backups; there are aspects to data storage that bring about unique challenges. You won’t need to adopt all of these techniques to be successful, but it’s time you had a frank conversation about what it takes to make your database truly “webscale”.
Slides from PGOpen 2011, But this talk was also delivered at Velocity 2011 as well.
This document discusses monitoring in the context of containers and microservices. It provides an overview of the history and evolution of containers from chroot and jails to Docker. It outlines different approaches to monitoring from manual to reactive to proactive. It discusses challenges of monitoring ephemeral infrastructure and the need for automation. It also covers architectural considerations for monitoring containers and microservices as well as specific techniques for collecting events, logs and metrics from containerized applications.
The document discusses improving alert systems by reducing unnecessary alerts. It suggests focusing on alerts that require human action, identifying the business impact of alerts, and documenting how each alert is addressed. Non-actionable alerts should be converted to notices or removed. Organizations should design resilience into their systems to minimize reliance on human intervention during incidents.
Setting up a GeoServer can sometimes be deceptively simple. However, going from proof of concept to production requires a number of steps to be taken in order to optimize the server in terms of availability, performance and scalability. The presentation will show how to get from a basic set up to a battle ready, rock solid installation by showing the ropes an advanced user already mastered.
Salam Z. Rahal is seeking an administrative assistant position utilizing over 15 years of experience. She has held roles with Seedbound, USAgriseeds Co., United Towers Holding Company, KIPCO Asset Management Company, Kuwait Real Estate Investment Consortium, and Kuwaiti Interest For Financial Investment. Rahal has a technical diploma in computer programming and is proficient in Microsoft Office, with strengths in working under pressure and as part of a team.
This magazine document provides information on health, fitness, weight loss, and beauty topics. It promotes detox boot camps, strategies for getting fit on a budget, and ways to lose 10 pounds in 4 weeks through diet and exercise. Additionally, it shares instant beauty tricks and tips for eating and drinking while still losing weight through motivation and lifestyle changes rather than dieting.
Mobile robot controller using novel hybrid system IJECEIAES
Hybrid neuro-fuzzy controller is one of the techniques that is used as a tool to control a mobile robot in unstructured environment. In this paper a novel neuro-fuzzy technique is proposed in order to tackle the problem of mobile robot autonomous navigation in unstructured environment. Obstacle avoidance is an important task in the field of robotics, since the goal of autonomous robot is to reach the destination without collision. The objective is to make the robot move along a collision free trajectory until it reaches its target. The proposed approach uses the artificial neural network instead of the fuzzified engine then the output from it is processed using adaptive inference engine and defuzzification engine. In this approach, the real processing time is reduced that is increase the mobile robot response. The proposed neuro-fuzzy controller is evaluated subjectively and objectively with other approaches and also the processing time is taken in consideration.
Autonomous system to control a mobile robotjournalBEEI
This paper presents an ongoing effort to control a mobile robot in unstructured environment. Obstacle avoidance is an important task in the field of robotics, since the goal of autonomous robot is to reach the destination without collision. Several algorithms have been proposed for obstacle avoidance, having drawbacks and benefits. In this paper, the fuzzy controller is used to tackle the problem of mobile robot autonomous navigation in unstructured environment. The objective is to make the robot move along a collision free trajectory until it reaches its target. The proposed approach uses the fuzzified, adaptive inference engine and defuzzification engine. Also number of linguistic labels is optimized for the input of the mobile robot in order to reduce computational time for real-time applications. The proposed fuzzy controller is evaluated subjectively and objectively with other approaches and also the processing time is taken in consideration.
Optimally Learnt, Neural Network Based Autonomous Mobile Robot Navigation SystemIDES Editor
Neural network based systems have been used in
past years for robot navigation applications because of their
ability to learn human expertise and to utilize this knowledge
to develop autonomous navigation strategies. In this paper,
neural based systems are developed for mobile robot reactive
navigation. The proposed systems transform sensors’ input to
yield wheel velocities. Novel algorithm is proposed for optimal
training of neural network. With a view to ascertain the efficacy
of proposed system; developed neural system’s performance
is compared to other neural and fuzzy based approaches.
Simulation results show effectiveness of proposed system in
all kind of obstacle environments.
This paper reports results of artificial neural network for robot navigation tasks. Machine
learning methods have proven usability in many complex problems concerning mobile robots
control. In particular we deal with the well-known strategy of navigating by “wall-following”.
In this study, probabilistic neural network (PNN) structure was used for robot navigation tasks.
The PNN result was compared with the results of the Logistic Perceptron, Multilayer
Perceptron, Mixture of Experts and Elman neural networks and the results of the previous
studies reported focusing on robot navigation tasks and using same dataset. It was observed the
PNN is the best classification accuracy with 99,635% accuracy using same dataset.
LEARNING OF ROBOT NAVIGATION TASKS BY PROBABILISTIC NEURAL NETWORKcsandit
This paper reports results of artificial neural network for robot navigation tasks. Machine
learning methods have proven usability in many complex problems concerning mobile robots
control. In particular we deal with the well-known strategy of navigating by “wall-following”.
In this study, probabilistic neural network (PNN) structure was used for robot navigation tasks.
The PNN result was compared with the results of the Logistic Perceptron, Multilayer
Perceptron, Mixture of Experts and Elman neural networks and the results of the previous
studies reported focusing on robot navigation tasks and using same dataset. It was observed the
PNN is the best classification accuracy with 99,635% accuracy using same dataset.
LEARNING OF ROBOT NAVIGATION TASKS BY PROBABILISTIC NEURAL NETWORKcscpconf
This paper reports results of artificial neural network for robot navigation tasks. Machine learning methods have proven usability in many complex problems concerning mobile robots
control. In particular we deal with the well-known strategy of navigating by “wall-following”. In this study, probabilistic neural network (PNN) structure was used for robot navigation tasks.
The PNN result was compared with the results of the Logistic Perceptron, Multilayer Perceptron, Mixture of Experts and Elman neural networks and the results of the previous
studies reported focusing on robot navigation tasks and using same dataset. It was observed the PNN is the best classification accuracy with 99,635% accuracy using same dataset.
Smart element aware gate controller for intelligent wheeled robot navigationIJECEIAES
This document presents a modified neuro-controller mechanism for controlling the navigation of an indoor mobile robot. The proposed mechanism uses a modified Elman neural network (MENN) with an effective element aware gate (MEEG) as the neuro-controller. The MEEG controller is able to estimate trajectory and overcome rigid and dynamic barriers intelligently. It was implemented on a Khepera IV mobile robot. Practical results showed the proposed mechanism was more efficient than MENN in providing the shortest distance to reach the goal with maximum velocity, minimizing the error rate by 58.33%. The document describes the system architecture, proposed neuro-controller, training algorithm for the MEEG, and presents analysis of sensor data and practical results.
Fuzzy Logic-Genetic Algorithm-Neural Network for Mobile Robot Navigation: A S...IRJET Journal
This document provides a survey of navigation methods for mobile robots that utilize fuzzy logic, genetic algorithms, and neural networks. It discusses several approaches that have been proposed in the literature, including individual and hybrid methods combining two or more soft computing techniques. The document focuses on fuzzy logic approaches for mobile robot navigation in Section 2, genetic algorithm approaches in Section 3, and neural network approaches in Section 4. Hybrid approaches combining fuzzy logic with genetic algorithms, neural networks, and genetic algorithms with neural networks are covered in Sections 5, 6, and 7 respectively.
Wearable sensor-based human activity recognition with ensemble learning: a co...IJECEIAES
The spectacular growth of wearable sensors has provided a key contribution to the field of human activity recognition. Due to its effective and versatile usage and application in various fields such as smart homes and medical areas, human activity recognition has always been an appealing research topic in artificial intelligence. From this perspective, there are a lot of existing works that make use of accelerometer and gyroscope sensor data for recognizing human activities. This paper presents a comparative study of ensemble learning methods for human activity recognition. The methods include random forest, adaptive boosting, gradient boosting, extreme gradient boosting, and light gradient boosting machine (LightGBM). Among the ensemble learning methods in comparison, light gradient boosting machine and random forest demonstrate the best performance. The experimental results revealed that light gradient boosting machine yields the highest accuracy of 94.50% on UCI-HAR dataset and 100% on single accelerometer dataset while random forest records the highest accuracy of 93.41% on motion sense dataset.
The document presents a lifelong federated reinforcement learning (LFRL) architecture for navigation in cloud robotic systems. LFRL allows robots to fuse their experience and transfer knowledge so they can effectively use prior knowledge and quickly adapt to new environments. It proposes a knowledge fusion algorithm to upgrade a shared model on the cloud by fusing private models from robots. It also introduces effective transfer learning methods to help robots rapidly adapt to new environments. Experiments show LFRL improves the efficiency of reinforcement learning for robot navigation. A cloud robotic navigation website is also presented to demonstrate LFRL.
Describe the need to multitask in BBC (behavior-based control) syste.pdfeyewaregallery
Describe the need to multitask in BBC (behavior-based control) systems?
Solution
Behavior-based control employs a set of distributed, in-teracting modules, called behaviors that
collectively achieve the desired system-level behavior. To an ex-ternal observer, behaviors are
patterns of the robot’s activity emerging from interactions between the robot and its
environment. To a programmer, behaviors are control modules that cluster sets of constraints in
order to achieve and maintain a goal. Each behavior receives inputs from sensors and/or other
behaviors in the system, and provides outputs to the robot’s actuators or to other behaviors. Thus,
a behavior-based controller is a structured network of interacting behaviors, with no centralized
world representation or focus of control. In-stead, individual behaviors and networks of
behaviors maintain any state information and models.
The basic principles of behavior-based control can be summarized briefly as follows:
• Behaviors are implemented as control laws (some-times similar to those used in control
theory), either in software or hardware, as a processing element or as a procedure.
• Each behavior can take inputs from the robot’s sen-sors (e.g., proximity sensors, range
detectors, contact sensors, camera) and/or from other modules in the
system, and send outputs to the robot’s effectors (e.g., wheels, grippers, arm, speech) and/or to
other modules.
• Many different behaviors may independently re- ceive input from the same sensors and output
action commands to the same actuators.
• Behaviors are encoded to be relatively simple, and are added to the system incrementally.
• Behaviors (or subsets thereof) are executed con- currently, not sequentially, in order to exploit
parallelism and speed of computation, as well as the interaction dynamics among behaviors and
between behaviors and the environment.
The ability to improve performance over time and to reason about the world, in the context of a
chang-ing and dynamic environment, is an important area of research in situated robotics. Unlike
in classical ma-chine learning, where the goal is typically to optimize performance over a long
period of time, in situated learning the aim is to adapt relatively quickly, toward
attaining efficiency in the light of uncertainty. Models from biology are often considered, given
its proper- ties of learning directly from environmental feedback. Variations and adaptations of
machine learning, and in particular reinforcement learning, have been effectively applied to
behavior-based robots, which have demon- strated learning to walk [38.
, navigate and create topological maps, di-vide tasks, behave socially , and even identify
opponents and score goals in robot soc-
cer. Methods from artificial life, evolutionary computation/genetic algorithms, fuzzy logic,
vision and learning, multi-agent systems, and many other research areas continue to be actively
explored and applied to behavior-based robots as their role in ani.
Develop a mobility model for MANETs networks based on fuzzy Logiciosrjce
The study and research in the field of networks MANETs depends alleged understand the protocols
well of the simulation process before they are applied in the real world, so that we create an environment
similar to these networks. The problem of a set of nodes connected with each other wirelessly, this requires the
development of a comprehensive model and full and real emulator for the movement of the contract on behalf of
stochastic models. Many models came to address the problems of random models that restricted the movement
of decade barriers as well as the signals exchanged between them, but these models were not receiving a lot of
light on the movement of the contract, such as direction, speed and path that is going by the node. The main
goal is to get a comprehensive model and simulator for all parts of the environment of the barriers and
obstacles to the movement of the nodes and the mobile signal between them as well as to focus on the movement
transactions for the node of the direction, speed, and best way. . This research aims to provide a realistic
mobility model for MANET networks. It also addresses the problem of imprecision in social relationships and
the location where we apply Fuzzy logic.
This document proposes developing a mobility model for mobile ad hoc networks (MANETs) based on fuzzy logic. It discusses existing mobility models and their limitations in capturing realistic node movement. The proposed model aims to provide a more realistic mobility model for MANETs by incorporating fuzzy logic to address imprecision in social relationships and node locations. It defines mathematical formulas to model social relationships between nodes and calculate the probability of nodes visiting locations based on these relationships and associated weights that vary over time. The model aims to take a more comprehensive approach to mobility modeling in MANETs by considering social, geographical, and temporal factors.
Universal Artificial Intelligence for Intelligent Agents: An Approach to Supe...IOSR Journals
This document proposes a methodology to develop intelligent agents with universal artificial intelligence (UAI) that can operate effectively in new environments. The methodology uses a neuro-fuzzy system combined with a hidden Markov model (HMM) to provide agents with learning capabilities and the ability to make decisions in unknown environments. The neuro-fuzzy system would extract fuzzy rules and membership functions from data to guide an agent. The HMM would generate sequences of sensed states to model dynamic environments. This approach aims to create "super intelligent agents" that can perform human-level tasks in any computable environment without reprogramming. A literature review found that neuro-fuzzy and HMM methods have been successfully used for mobile robot obstacle avoidance and human motion recognition.
AN APPROACH OF IR-BASED SHORT-RANGE CORRESPONDENCE SYSTEMS FOR SWARM ROBOT BA...ijaia
This paper exhibits a short-run correspondence method appropriate for swarm versatile robots application.
Infrared is utilized for transmitting and accepting information and obstruction location. The infrared
correspondence code based swarm signaling is utilized for an independent versatile robot communication
system in this research. A code based signaling system is developed for transmitting information between
different entities of robot. The reflected infrared sign is additionally utilized for separation estimation for
obstruction evasion. Investigation of robot demonstrates the possibility of utilizing infrared signs to get a
solid nearby correspondence between swarm portable robots. This paper exhibits a basic decentralized
control for swarm of self-collecting robots. Every robot in the code based swarm signaling is completely
self-governing and controlled utilizing a conduct based methodology with just infrared-based nearby
detecting and correspondences. The viability of the methodology has been checked with simulation, for a
set of swarm robots.
With the development of robotics and artificial intelligence field unceasingly thorough, path planning for avoid
obstacles as an important field of robot calculation has been widespread concern. This paper analyzes the
current development of robot and path planning algorithm for path planning to avoid obstacles in practice. We
tried to find a good way in mobile robot path planning by using ant colony algorithm, and it also provides some
solving methods.
Robot operating system based autonomous navigation platform with human robot ...TELKOMNIKA JOURNAL
In emerging technologies, indoor service robots are playing a vital role for people who are physically challenged and visually impaired. The service robots are efficient and beneficial for people to overcome the challenges faced during their regular chores. This paper proposes the implementation of autonomous navigation platforms with human-robot interaction which can be used in service robots to avoid the difficulties faced in daily activities. We used the robot operating system (ROS) framework for the implementation of algorithms used in auto navigation, speech processing and recognition, and object detection and recognition. A suitable robot model was designed and tested in the Gazebo environment to evaluate the algorithms. The confusion matrix that was created from 125 different cases points to the decent correctness of the model.
HUMAN ACTION RECOGNITION IN VIDEOS USING STABLE FEATURES sipij
Human action recognition is still a challenging problem and researchers are focusing to investigate this
problem using different techniques. We propose a robust approach for human action recognition. This is
achieved by extracting stable spatio-temporal features in terms of pairwise local binary pattern (P-LBP)
and scale invariant feature transform (SIFT). These features are used to train an MLP neural network
during the training stage, and the action classes are inferred from the test videos during the testing stage.
The proposed features well match the motion of individuals and their consistency, and accuracy is higher
using a challenging dataset. The experimental evaluation is conducted on a benchmark dataset commonly
used for human action recognition. In addition, we show that our approach outperforms individual features
i.e. considering only spatial and only temporal feature.
A novel enhanced algorithm for efficient human trackingIJICTJOURNAL
This paper introduces a novel enhanced algorithm for efficient human tracking based on background subtraction. The algorithm operates in four steps: 1) identifying a fixed background and removing noise, 2) subtracting the background from movable objects, 3) filtering the image to remove shadows and noise, and 4) separating and tracking movable objects using bubble routing. Experimental results found that the proposed algorithm improved motion and trajectory estimation of objects in terms of speed and accuracy compared to previous methods. The algorithm provides a fast and accurate way to track humans which has applications in video surveillance, human action recognition, and human-computer interaction.
Intelligent Robotics Navigation System: Problems, Methods, and Algorithm IJECEIAES
This paper set out to supplement new studies with a brief and comprehensible review of the advanced development in the area of the navigation system, starting from a single robot, multi-robot, and swarm robots from a particular perspective by taking insights from these biological systems. The inspiration is taken from nature by observing the human and the social animal that is believed to be very beneficial for this purpose. The intelligent navigation system is developed based on an individual characteristic or a social animal biological structure. The discussion of this paper will focus on how simple agent’s structure utilizes flexible and potential outcomes in order to navigate in a productive and unorganized surrounding. The combination of the navigation system and biologically inspired approach has attracted considerable attention, which makes it an important research area in the intelligent robotic system. Overall, this paper explores the implementation, which is resulted from the simulation performed by the embodiment of robots operating in real environments.
Similar to Reactive Navigation of Autonomous Mobile Robot Using Neuro-Fuzzy System (20)
The Use of Java Swing’s Components to Develop a WidgetWaqas Tariq
Widget is a kind of application provides a single service such as a map, news feed, simple clock, battery-life indicators, etc. This kind of interactive software object has been developed to facilitate user interface (UI) design. A user interface (UI) function may be implemented using different widgets with the same function. In this article, we present the widget as a platform that is generally used in various applications, such as in desktop, web browser, and mobile phone. We also describe a visual menu of Java Swing’s components that will be used to establish widget. It will assume that we have successfully compiled and run a program that uses Swing components.
3D Human Hand Posture Reconstruction Using a Single 2D ImageWaqas Tariq
Passive sensing of the 3D geometric posture of the human hand has been studied extensively over the past decade. However, these research efforts have been hampered by the computational complexity caused by inverse kinematics and 3D reconstruction. In this paper, our objective focuses on 3D hand posture estimation based on a single 2D image with aim of robotic applications. We introduce the human hand model with 27 degrees of freedom (DOFs) and analyze some of its constraints to reduce the DOFs without any significant degradation of performance. A novel algorithm to estimate the 3D hand posture from eight 2D projected feature points is proposed. Experimental results using real images confirm that our algorithm gives good estimates of the 3D hand pose. Keywords: 3D hand posture estimation; Model-based approach; Gesture recognition; human- computer interface; machine vision.
Camera as Mouse and Keyboard for Handicap Person with Troubleshooting Ability...Waqas Tariq
Camera mouse has been widely used for handicap person to interact with computer. The utmost important of the use of camera mouse is must be able to replace all roles of typical mouse and keyboard. It must be able to provide all mouse click events and keyboard functions (include all shortcut keys) when it is used by handicap person. Also, the use of camera mouse must allow users troubleshooting by themselves. Moreover, it must be able to eliminate neck fatigue effect when it is used during long period. In this paper, we propose camera mouse system with timer as left click event and blinking as right click event. Also, we modify original screen keyboard layout by add two additional buttons (button “drag/ drop” is used to do drag and drop of mouse events and another button is used to call task manager (for troubleshooting)) and change behavior of CTRL, ALT, SHIFT, and CAPS LOCK keys in order to provide shortcut keys of keyboard. Also, we develop recovery method which allows users go from camera and then come back again in order to eliminate neck fatigue effect. The experiments which involve several users have been done in our laboratory. The results show that the use of our camera mouse able to allow users do typing, left and right click events, drag and drop events, and troubleshooting without hand. By implement this system, handicap person can use computer more comfortable and reduce the dryness of eyes.
A Proposed Web Accessibility Framework for the Arab DisabledWaqas Tariq
The Web is providing unprecedented access to information and interaction for people with disabilities. This paper presents a Web accessibility framework which offers the ease of the Web accessing for the disabled Arab users and facilitates their lifelong learning as well. The proposed framework system provides the disabled Arab user with an easy means of access using their mother language so they don’t have to overcome the barrier of learning the target-spoken language. This framework is based on analyzing the web page meta-language, extracting its content and reformulating it in a suitable format for the disabled users. The basic objective of this framework is supporting the equal rights of the Arab disabled people for their access to the education and training with non disabled people. Key Words : Arabic Moon code, Arabic Sign Language, Deaf, Deaf-blind, E-learning Interactivity, Moon code, Web accessibility , Web framework , Web System, WWW.
Real Time Blinking Detection Based on Gabor FilterWaqas Tariq
The document proposes a new method for real-time blinking detection based on Gabor filters. It begins by reviewing existing methods and their limitations in dealing with noise, variations in eye shape, and blinking speed. The proposed method uses a Gabor filter to extract the top and bottom arcs of the eye from an image. It then measures the distance between these arcs and compares it to a threshold: a distance below the threshold indicates a closed eye, while a distance above indicates an open eye. The document claims this Gabor filter-based approach is robust to noise, variations in eye shape and blinking speed. It presents experimental results showing the method can accurately detect blinking across different users.
Computer Input with Human Eyes-Only Using Two Purkinje Images Which Works in ...Waqas Tariq
A method for computer input with human eyes-only using two Purkinje images which works in a real time basis without calibration is proposed. Experimental results shows that cornea curvature can be estimated by using two light sources derived Purkinje images so that no calibration for reducing person-to-person difference of cornea curvature. It is found that the proposed system allows usersf movements of 30 degrees in roll direction and 15 degrees in pitch direction utilizing detected face attitude which is derived from the face plane consisting three feature points on the face, two eyes and nose or mouth. Also it is found that the proposed system does work in a real time basis.
Toward a More Robust Usability concept with Perceived Enjoyment in the contex...Waqas Tariq
Mobile multimedia service is relatively new but has quickly dominated people¡¯s lives, especially among young people. To explain this popularity, this study applies and modifies the Technology Acceptance Model (TAM) to propose a research model and conduct an empirical study. The goal of study is to examine the role of Perceived Enjoyment (PE) and what determinants can contribute to PE in the context of using mobile multimedia service. The result indicates that PE is influencing on Perceived Usefulness (PU) and Perceived Ease of Use (PEOU) and directly Behavior Intention (BI). Aesthetics and flow are key determinants to explain Perceived Enjoyment (PE) in mobile multimedia usage.
Collaborative Learning of Organisational KnolwedgeWaqas Tariq
This paper presents recent research into methods used in Australian Indigenous Knowledge sharing and looks at how these can support the creation of suitable collaborative envi- ronments for timely organisational learning. The protocols and practices as used today and in the past by Indigenous communities are presented and discussed in relation to their relevance to a personalised system of knowledge sharing in modern organisational cultures. This research focuses on user models, knowledge acquisition and integration of data for constructivist learning in a networked repository of or- ganisational knowledge. The data collected in the repository is searched to provide collections of up-to-date and relevant material for training in a work environment. The aim is to improve knowledge collection and sharing in a team envi- ronment. This knowledge can then be collated into a story or workflow that represents the present knowledge in the organisation.
Our research aims to propose a global approach for specification, design and verification of context awareness Human Computer Interface (HCI). This is a Model Based Design approach (MBD). This methodology describes the ubiquitous environment by ontologies. OWL is the standard used for this purpose. The specification and modeling of Human-Computer Interaction are based on Petri nets (PN). This raises the question of representation of Petri nets with XML. We use for this purpose, the standard of modeling PNML. In this paper, we propose an extension of this standard for specification, generation and verification of HCI. This extension is a methodological approach for the construction of PNML with Petri nets. The design principle uses the concept of composition of elementary structures of Petri nets as PNML Modular. The objective is to obtain a valid interface through verification of properties of elementary Petri nets represented with PNML.
Development of Sign Signal Translation System Based on Altera’s FPGA DE2 BoardWaqas Tariq
The main aim of this paper is to build a system that is capable of detecting and recognizing the hand gesture in an image captured by using a camera. The system is built based on Altera’s FPGA DE2 board, which contains a Nios II soft core processor. Image processing techniques and a simple but effective algorithm are implemented to achieve this purpose. Image processing techniques are used to smooth the image in order to ease the subsequent processes in translating the hand sign signal. The algorithm is built for translating the numerical hand sign signal and the result are displayed on the seven segment display. Altera’s Quartus II, SOPC Builder and Nios II EDS software are used to construct the system. By using SOPC Builder, the related components on the DE2 board can be interconnected easily and orderly compared to traditional method that requires lengthy source code and time consuming. Quartus II is used to compile and download the design to the DE2 board. Then, under Nios II EDS, C programming language is used to code the hand sign translation algorithm. Being able to recognize the hand sign signal from images can helps human in controlling a robot and other applications which require only a simple set of instructions provided a CMOS sensor is included in the system.
Exploring the Relationship Between Mobile Phone and Senior Citizens: A Malays...Waqas Tariq
There is growing ageing phenomena with the rise of ageing population throughout the world. According to the World Health Organization (2002), the growing ageing population indicates 694 million, or 223% is expected for people aged 60 and over, since 1970 and 2025.The growth is especially significant in some advanced countries such as North America, Japan, Italy, Germany, United Kingdom and so forth. This growing older adult population has significantly impact the social-culture, lifestyle, healthcare system, economy, infrastructure and government policy of a nation. However, there are limited research studies on the perception and usage of a mobile phone and its service for senior citizens in a developing nation like Malaysia. This paper explores the relationship between mobile phones and senior citizens in Malaysia from the perspective of a developing country. We conducted an exploratory study using contextual interviews with 5 senior citizens of how they perceive their mobile phones. This paper reveals 4 interesting themes from this preliminary study, in addition to the findings of the desirable mobile requirements for local senior citizens with respect of health, safety and communication purposes. The findings of this study bring interesting insight to local telecommunication industries as a whole, and will also serve as groundwork for more in-depth study in the future.
Principles of Good Screen Design in WebsitesWaqas Tariq
Visual techniques for proper arrangement of the elements on the user screen have helped the designers to make the screen look good and attractive. Several visual techniques emphasize the arrangement and ordering of the screen elements based on particular criteria for best appearance of the screen. This paper investigates few significant visual techniques in various web user interfaces and showcases the results for better understanding and their presence.
This document discusses the progress of virtual teams in Albania. It provides context on virtual teams and how they differ from traditional teams in their reliance on technology for communication across distances. The document then examines the use of virtual teams in Albania, noting the growing infrastructure and technology usage that enables virtual collaboration. It highlights some virtual team examples in Albanian government and academic projects.
Cognitive Approach Towards the Maintenance of Web-Sites Through Quality Evalu...Waqas Tariq
It is a well established fact that the Web-Applications require frequent maintenance because of cutting– edge business competitions. The authors have worked on quality evaluation of web-site of Indian ecommerce domain. As a result of that work they have made a quality-wise ranking of these sites. According to their work and also the survey done by various other groups Futurebazaar web-site is considered to be one of the best Indian e-shopping sites. In this research paper the authors are assessing the maintenance of the same site by incorporating the problems incurred during this evaluation. This exercise gives a real world maintainability problem of web-sites. This work will give a clear picture of all the quality metrics which are directly or indirectly related with the maintainability of the web-site.
USEFul: A Framework to Mainstream Web Site Usability through Automated Evalua...Waqas Tariq
A paradox has been observed whereby web site usability is proven to be an essential element in a web site, yet at the same time there exist an abundance of web pages with poor usability. This discrepancy is the result of limitations that are currently preventing web developers in the commercial sector from producing usable web sites. In this paper we propose a framework whose objective is to alleviate this problem by automating certain aspects of the usability evaluation process. Mainstreaming comes as a result of automation, therefore enabling a non-expert in the field of usability to conduct the evaluation. This results in reducing the costs associated with such evaluation. Additionally, the framework allows the flexibility of adding, modifying or deleting guidelines without altering the code that references them since the guidelines and the code are two separate components. A comparison of the evaluation results carried out using the framework against published evaluations of web sites carried out by web site usability professionals reveals that the framework is able to automatically identify the majority of usability violations. Due to the consistency with which it evaluates, it identified additional guideline-related violations that were not identified by the human evaluators.
Robot Arm Utilized Having Meal Support System Based on Computer Input by Huma...Waqas Tariq
A robot arm utilized having meal support system based on computer input by human eyes only is proposed. The proposed system is developed for handicap/disabled persons as well as elderly persons and tested with able persons with several shapes and size of eyes under a variety of illumination conditions. The test results with normal persons show the proposed system does work well for selection of the desired foods and for retrieve the foods as appropriate as usersf requirements. It is found that the proposed system is 21% much faster than the manually controlled robotics.
Dynamic Construction of Telugu Speech Corpus for Voice Enabled Text EditorWaqas Tariq
In recent decades speech interactive systems have gained increasing importance. Performance of an ASR system mainly depends on the availability of large corpus of speech. The conventional method of building a large vocabulary speech recognizer for any language uses a top-down approach to speech. This approach requires large speech corpus with sentence or phoneme level transcription of the speech utterances. The transcriptions must also include different speech order so that the recognizer can build models for all the sounds present. But, for Telugu language, because of its complex nature, a very large, well annotated speech database is very difficult to build. It is very difficult, if not impossible, to cover all the words of any Indian language, where each word may have thousands and millions of word forms. A significant part of grammar that is handled by syntax in English (and other similar languages) is handled within morphology in Telugu. Phrases including several words (that is, tokens) in English would be mapped on to a single word in Telugu.Telugu language is phonetic in nature in addition to rich in morphology. That is why the speech technology developed for English cannot be applied to Telugu language. This paper highlights the work carried out in an attempt to build a voice enabled text editor with capability of automatic term suggestion. Main claim of the paper is the recognition enhancement process developed by us for suitability of highly inflecting, rich morphological languages. This method results in increased speech recognition accuracy with very much reduction in corpus size. It also adapts Telugu words to the database dynamically, resulting in growth of the corpus.
An Improved Approach for Word Ambiguity RemovalWaqas Tariq
Word ambiguity removal is a task of removing ambiguity from a word, i.e. correct sense of word is identified from ambiguous sentences. This paper describes a model that uses Part of Speech tagger and three categories for word sense disambiguation (WSD). Human Computer Interaction is very needful to improve interactions between users and computers. For this, the Supervised and Unsupervised methods are combined. The WSD algorithm is used to find the efficient and accurate sense of a word based on domain information. The accuracy of this work is evaluated with the aim of finding best suitable domain of word. Keywords: Human Computer Interaction, Supervised Training, Unsupervised Learning, Word Ambiguity, Word sense disambiguation
Parameters Optimization for Improving ASR Performance in Adverse Real World N...Waqas Tariq
From the existing research it has been observed that many techniques and methodologies are available for performing every step of Automatic Speech Recognition (ASR) system, but the performance (Minimization of Word Error Recognition-WER and Maximization of Word Accuracy Rate- WAR) of the methodology is not dependent on the only technique applied in that method. The research work indicates that, performance mainly depends on the category of the noise, the level of the noise and the variable size of the window, frame, frame overlap etc is considered in the existing methods. The main aim of the work presented in this paper is to use variable size of parameters like window size, frame size and frame overlap percentage to observe the performance of algorithms for various categories of noise with different levels and also train the system for all size of parameters and category of real world noisy environment to improve the performance of the speech recognition system. This paper presents the results of Signal-to-Noise Ratio (SNR) and Accuracy test by applying variable size of parameters. It is observed that, it is really very hard to evaluate test results and decide parameter size for ASR performance improvement for its resultant optimization. Hence, this study further suggests the feasible and optimum parameter size using Fuzzy Inference System (FIS) for enhancing resultant accuracy in adverse real world noisy environmental conditions. This work will be helpful to give discriminative training of ubiquitous ASR system for better Human Computer Interaction (HCI). Keywords: ASR Performance, ASR Parameters Optimization, Multi-Environmental Training, Fuzzy Inference System for ASR, ubiquitous ASR system, Human Computer Interaction (HCI)
Interface on Usability Testing Indonesia Official Tourism WebsiteWaqas Tariq
Ministry of Tourism and Creative Economy of The Republic of Indonesia must meet the wide audience various needs and should reach people from all levels of society around the world to provide Indonesia tourism and travel information. This article will gives the details in the evolution of one important component of Indonesia Official Tourism Website as it has grown in functionality and usefulness over several years of use by a live, unrestricted community. We chose this website to see the website interface design and usability and to popularize Indonesia tourism and travel highlights. The analysis done by looking at the criteria specified for usability testing. Usability testing measures are the ease of use (effectiveness, efficiency, consistency and interface design), easy to learn, errors and syntax which is related to the human computer interaction. The purpose of this article is to test the usability level of the website, analyze the website interface design, and provide suggestions for improvements in Indonesia Official Tourism Website of analysis we have done before.
Beyond Degrees - Empowering the Workforce in the Context of Skills-First.pptxEduSkills OECD
Iván Bornacelly, Policy Analyst at the OECD Centre for Skills, OECD, presents at the webinar 'Tackling job market gaps with a skills-first approach' on 12 June 2024
A Visual Guide to 1 Samuel | A Tale of Two HeartsSteve Thomason
These slides walk through the story of 1 Samuel. Samuel is the last judge of Israel. The people reject God and want a king. Saul is anointed as the first king, but he is not a good king. David, the shepherd boy is anointed and Saul is envious of him. David shows honor while Saul continues to self destruct.
How to Make a Field Mandatory in Odoo 17Celine George
In Odoo, making a field required can be done through both Python code and XML views. When you set the required attribute to True in Python code, it makes the field required across all views where it's used. Conversely, when you set the required attribute in XML views, it makes the field required only in the context of that particular view.
The chapter Lifelines of National Economy in Class 10 Geography focuses on the various modes of transportation and communication that play a vital role in the economic development of a country. These lifelines are crucial for the movement of goods, services, and people, thereby connecting different regions and promoting economic activities.
Temple of Asclepius in Thrace. Excavation resultsKrassimira Luka
The temple and the sanctuary around were dedicated to Asklepios Zmidrenus. This name has been known since 1875 when an inscription dedicated to him was discovered in Rome. The inscription is dated in 227 AD and was left by soldiers originating from the city of Philippopolis (modern Plovdiv).
BÀI TẬP DẠY THÊM TIẾNG ANH LỚP 7 CẢ NĂM FRIENDS PLUS SÁCH CHÂN TRỜI SÁNG TẠO ...
Reactive Navigation of Autonomous Mobile Robot Using Neuro-Fuzzy System
1. M. M. Joshi & M. A. Zaveri
International Journal of Robotics and Automation (IJRA), Volume (2) : Issue (3), 2011 128
Reactive Navigation of Autonomous Mobile Robot Using Neuro-
Fuzzy System
Maulin M. Joshi maulin.joshi@scet.ac.in
Department of Electronics & Communications,
Sarvajanik College of Engineering and Technology
Surat,395001, India
Mukesh A Zaveri mazaveri@coed.svnit.ac.in
Department of Computer Engineering
Sardar vallabhbhai National Institute of Technology
Surat, 395007, India
Abstract
Neuro-fuzzy systems have been used for robot navigation applications because of their ability to
exert human like expertise and to utilize acquired knowledge to develop autonomous navigation
strategies. In this paper, neuro-fuzzy based system is proposed for reactive navigation of a
mobile robot using behavior based control. The proposed algorithm uses discrete sampling based
optimal training of neural network. With a view to ascertain the efficacy of proposed system; the
proposed neuro-fuzzy system’s performance is compared to that of neural and fuzzy based
approaches. Simulation results along with detailed behavior analysis show effectiveness of our
algorithm in all kind of obstacle environments.
Keywords: Reactive Navigation, Mobile Robot, Neural Network, Behavior Analysis, Discrete
Sampling
1. INTRODUCTION
Autonomous robot navigation means the ability of a robot to move purposefully and without
human intervention in environments that have not been specifically engineered for it [1].
Autonomous navigation requires a number of heterogeneous capabilities like ability to reach a
given location in real time to unexpected events, to determine the robot's position; and to adapt to
the changes in the environment. For a mobile robot to navigate automatically and rapidly, an
important factor is to identify and classify mobile robots' currently perceptual environment [1]. The
general theory for mobile robotics navigation is based on a following idea: robot must Sense the
known world, be able to Plan its operations and then Act based on the model.
In spite of impressive advances in the field of autonomous robotics in recent years, it is still the
area of an active research because of uncertainties involved due to unknown environments in
real world scenarios. These uncertainties are due to following reasons [1]: no information or less
information about a prior knowledge of an environment, lack of perceptually acquired information,
limited range, adverse observation conditions, complex and unpredictable dynamics. It is also
required that the behavior of the robot must be reactive to dynamic aspects of the unknown
environments and must be able to generate robust behavior in the face of uncertain sensors,
unpredictable environments and changing scenario.
Many approaches have been proposed to solve the above mentioned challenges for autonomous
robot navigation. Some of the approaches focus on path planning methods [2], few approaches
use potential field [3] in which the robot-motion reaction is determined by the resultant virtual
force. Several other methods have been used like statistical methods, Partially Observable
Markov Decision Process (POMDP) [4] and reinforcement learning schemes [5]. In last few
years, research in the domain is more focused with neural and fuzzy based artificial intelligence
based approaches because of their ability to mimic human expertise.
2. M. M. Joshi & M. A. Zaveri
International Journal of Robotics and Automation (IJRA), Volume (2) : Issue (3), 2011 129
Humans have a remarkable capability to learn and perform a wide variety of physical and mental
tasks via generalization of perceived knowledge. Neural network based approaches are used in
robot navigation applications because neural network learns the humanoid expertise and then
tries to mimic them by implementing in environment which may be similar or even different than
used in its training (i.e. generalization). The attractive potential force attracts the robot toward the
target configuration, while repulsive potential forces push it away from obstacles. The mobile
robot is considered moving under the influence of resultant artificial potential field. The advantage
of neural based approach lies in the learning capacity of the neural network. Performance of
neural based system depends upon the effective training of its adjustable parameters (synaptic
weights and bias parameters). Dahm et al. [6] have introduced a neural field based approach on
robot ARNOLD. The approach was described by non linear competitive dynamical system.
However, kinematics constraints were not considered for activation of set of artificial neurons.
Zalama et. al. [7] have proposed reactive behavioral navigation of mobile robot using competitive
neural network. The authors described various interconnected modules to generate wheel
velocity using neural network. However, in such mechanisms many times learning convergence is
very slow and generalization is not always satisfactory. A neural dynamics based architecture
proposed by Yang and Meng [8]-[ 9] have discussed to reduce the computational complexity by
avoiding learning procedures and also stability has been proven by lyapunov function and
qualitative analysis. However, biologically inspired this neural method did not considered sensor
information fusion and behavior combination. Some of earlier models are not found practical as
they assumed that the whole workspace is definitely known considering only static environment.
Humans’ capability to perform various tasks without any explicit measurements or computations
is mimicked by fuzzy logic by providing formal methodology for representing and implementing
the human expert's heuristic knowledge and perception based actions. Fuzzy logic based many
approaches have been investigated in past years for controlling a mobile robot because of its
capability to make inferences under uncertainty [10]. Artificial potential field approach has been
proposed by Khatib[11] that discussed behavior based control. Saffoitti[12] has proposed fuzzy
based methods for mobile robot navigation. Ismail and Nordin [13] have proposed reactive
navigation by considering two separate fuzzy controllers for velocities and steering angle. In all
these approaches, the purpose was restricted for fundamental and simple control actions. Fuzzy
velocity control of mobile robot has been discussed by Mester [14]. However, only 10 heuristic
fuzzy rules were used in their experiments. These approaches have inherent drawback that much
efforts are needed to adjust tuning parameters and firing in advance. Intelligent navigation
systems for omni directional mobile robots were described by Zavalang et.al.[15], which was
influenced by potential field approach. Ishikawa [16] and Wei li et al.[17] have proposed behavior
fusion for robot navigation in uncertainty using fuzzy logic. Both these approaches need
improvements to handle complex environment. A system integrating techniques like dead–
reckoning, self localization and environment are reported by Lee and Wu in [18]. In their
approach membership functions and fuzzy rules were designed based on genetic algorithm.
However, Genetic algorithm may not the best method for generation of rule base with 25 rules
and priority based selection of heading directions does not take into account the behavior
coordination and this algorithm focuses on direction control without considering velocity control.
An obstacle avoidance approach using fuzzy logic has been proposed by Li and Yang [19]. A
collision-avoidance approach using fuzzy logic is introduced by Lin and Wang [20] where,
different modules e.g. Static-obstacle avoiding module, avoiding moving obstacle module and
directing-toward-target module are created for the robot navigation. However, these modules are
separately inferred and are not as coordinated as human reasoning. In mobile robots reactive
navigation, key problem of local minima is addressed by Zhu and Yang [21] with state memory
strategy; Wang and Liu [22] with minimum risk approach and by Xu and Tso [23] by considering
π radian target switching. O.R.E. Motlagh et.al.[24] proposed virtual target switching strategy to
resolve multiple dead end to improve the performance of earlier methods by considering three
target states and six obstacle states resulting into 18 rules. However, with the limited number of
rules such improvement not always guaranteed in dynamically changing environment with
change in dead end shapes.
3. M. M. Joshi & M. A. Zaveri
International Journal of Robotics and Automation (IJRA), Volume (2) : Issue (3), 2011 130
To improve the performance, some neuro fuzzy methods have been proposed. Song and sheen
[25] have considered heuristic fuzzy- neuro network to reactive navigation of mobile robot. In their
approach, resulted velocity command enabled robot to move in an unknown environment using
Fuzzy Kohonen Clustering Network (FKCN). However, their heuristic approach considered nine
typical obstacle classes to formulate total 16 rules. Wei li et. al.[26] have proposed two level
neuro-fuzzy architecture for behavior based mobile. In that approach, neural training has been
done by four layer standard back propagation network and used only few selected examples to
train neural network. However, in both above approaches; generalization of neural network for
complete input space with limited training examples can not be guaranteed. Marichal et al. [27]
have suggested an-other neuro fuzzy strategy by considering a three-layer neural network with a
competitive learning algorithm for a mobile robot. The approach has been able to extract the
information for fuzzy rules and the membership functions from human guided set of trajectories.
For complex situations, it is difficult to optimally set required trajectories and hence resulted rules
may not work well for generalization. Zhu and yang [28] have proposed five layer neuro-fuzzy
controller considering neural networks to improve the performance of fuzzy network. The
approach includes an algorithm to surpass redundant rules by observing the response of fuzzy
network and removing rules with hamming distance lesser than specified threshold. However,
this obviously requires training to mobile robot in given environment. However, for some critical
operations like mining and under water operations, such training in given environment is never
possible. Approaches without proper generalization will fail to take best decision when mobile
robot needs to take immediate actions without any prior scanning of the given environment.
Heuristic based approaches do not guarantee satisfactory performance for in general, difficult
unknown environment space.
In this paper, we propose two level neuro fuzzy based algorithm that overcomes the shortcoming
of current approaches [25-28] in terms of learning mechanism used. In the proposed system,
environment sense is done by neural network and behavioral control is executed by fuzzy
system. Inputs to the neural network are outputs from multi sensors groups and heading angle.
Output of neural network is reference heading angle that in connection with sensors data serves
as input to fuzzy system. We propose discrete sampling based approach, in which optimal neural
training is achieved by providing effective heterogeneity in training pairs while; retaining
homogeneity in terms of providing different training pairs to the neural network. In our approach,
we have generalized many parameters for robot navigation task like; number of sensors required
for environmental sensing, arrangement of sensors, sensors grouping and quantization and
heading angle inference. These make our approach unique and more generalized compared to
approaches found in literature. Generalization of fuzzy based parameters enables us to select, to
tune parameters as per requirements of given environmental conditions. Behavior based fuzzy
systems used for mobile robot navigation demonstrate reasonably good performance; while
navigating in cluttered and unknown environment
The rest of this paper is organized as follows: the proposed algorithm for neuro-fuzzy based
mobile robot navigation is discussed in Section 2, including range computation from given
obstacles, sensors arrangement, grouping quantization and inference of heading angle. Section 3
descries neuro-fuzzy system for reactive navigation. Section 4 illustrates simulation results and
detailed behavior analysis of neuro-fuzzy based mobile robot navigation. Finally concluding
remarks are given in Section 5.
2. PROPOSED ALGORITHM
In this section, we propose an algorithm for reactive navigation for a mobile robot using neuro-
fuzzy based sys-tem. First, we describe the problem formulation for the motion planning problem.
Let A be the single robot moving in a Euclidian space W, called workspace, represented as RN
,
with N= 2 or 3. Let B1, B2…Bq be the rigid objects distributed in W. The Bi’s are called obstacles.
With assumptions that no kinematics constraints limit the motion of A in W, generate a path T
specifying a sequence of positions and orientations of A avoiding contact with Bi’s, i.e. starting at
the initial position and orientation and terminating at the goal position and orientation.
4. M. M. Joshi & M. A. Zaveri
International Journal of Robotics and Automation (IJRA), Volume (2) : Issue (3), 2011 131
2.1 Mobile Robot Configuration
We consider two dimensional workspace (N=2) for mobile robot as shown in Figure.1. Mobile
robot is having initial and target position coordinates denoted as (xo, yo) and (xt, yt) respectively.
Mobile robot’s current position (calculated and updated at each step) can be denoted as (xcurr
,ycurr). Angle between target and positive y axis is θtr. Robot’s pose (head) with respect to positive
y axis is considered as θhr and θhead is the heading angle between target and robot current
position. Span (S) is the distance between left and right wheel. Vl and Vr are mobile robots left
wheel and right wheel velocities, respectively.
FIGURE 1: Mobile Robot Configuration
The mobile robot has two independently driven co-axial wheels. We consider a mobile robot with
differential drive wheels. Final target positions are known to the robot at all the time. At each step,
current location and orientation are computed. No history of past sensor readings are retained
and thus robot is having pure reactive navigation. Obstacles may be stationary or may be mobile.
2.2 Range Calculation of Mobile Robot From Given Obstacles
Acquisition of precise range information of a mobile robot from each nearby obstacle is one of the
most important tasks for robot navigation. Mobile robot needs to effectively sense surrounding
environment. We have proposed an algorithm to find range information for robot navigation in
presence of moving obstacles in our earlier work [29]. The important point is that because of
presence of moving obstacles, prior geometry information may not help. But, our model acquires
geometry information from sensed signals computed with the help of sensors mounted on robot.
This makes our approach very general and can be used for any scenario. Following steps
explains our algorithm to find out range of obstacle from robot A to Obstacle B:
Target
X-Axis
θhead
Robot
Y-Axis
Robot head
θtr
θhr
5. M. M. Joshi & M. A. Zaveri
International Journal of Robotics and Automation (IJRA), Volume (2) : Issue (3), 2011 132
FIGURE 2: Range Calculation of a Mobile Robot from Obstacles
1) Let total N ultrasonic sensors be placed on robot to sense the surrounding environment as
shown in Fig-ure.2. These sensors are represented as N1,N2… Nk, Nk+1.. NN. Signal of k
th
sensor
(Nk) is represented by an arrow towards the obstacle.
2) Let, (x1, y1) and (x2, y2) are two points on robot to represent k
th
sensor direction. The ray
emerging from sensor mounted on mobile robot to obstacle can be considered in terms of
parametric equation form of straight line as:
x= x1+ ( x2- x1) Dk
y= y1+ (y2- y1) Dk (1)
Where, Dk is a real value that denotes the distance of a mobile robot from obstacle. In order to
ensure that robot looks only in forward direction and the maximum range of ultrasonic sensor is
set to Dmax,
0 < Dk< Dmax (2)
3) Small line segment on an obstacle will be represented by points (x3, y3) and (x4, y4). This line
segment will intersect with ray emitted by the sensors on robot. Particularly this line segment
being very small can be considered as straight line segment. This assumption will allow us to
calculate range for any shape obstacle in our algorithm. Consider (x3, y3) and (x4, y4) be two
points representing one line segment on the ith
obstacle and described by parametric equation
form of straight line as:
x = x3+ (x4- x3) Sij
y = y3+ (y4- y3) Sij (3)
Where, Sij - a real value presenting line segment of i
th
obstacle’s j
th
side. To ensure that a
particular ray emitted by sensor mounted on the robot hits the line segment (side of the obstacle);
the value of Sij should be between 0 and 1, i.e.
0 <= Sij <= 1 (4)
4) Solution of equations (1) and (3) will give us the distance DK, i.e. distance between robot’s kth
sensor to the i
th
obstacle’s j
th
side:
6. M. M. Joshi & M. A. Zaveri
International Journal of Robotics and Automation (IJRA), Volume (2) : Issue (3), 2011 133
((y -y )*(x -x )-(y -y )*(x -x ))4 3 1 3 1 3 4 3
((y - y )*(x -x )-(y -y )*(x -x ))2 1 4 3 4 3 2 1
DK = (5)
5) Computation of the value of DK is to be carried out for each of total N sensors.
For example, rectangle shaped n obstacles will have 4*n edges. For total N sensors, there will N*
(4* n) size matrix computed at each step.
Sensors Grouping
In our algorithm, we consider robot fitted with N ultrasonic sensors in the front. If the front (head)
of the robot is at 0 degrees (w.r.t. +y axis), then the sensors are located between -90 to +90
degrees each being separated by θs =180/N degrees as shown in figure. 3(a).
FIGURE 3(a): Arrangement of ultrasonic sensors FIGURE 3(b): grouping of sensors
Sensors are grouped and final values are quantized before sending into the intelligent network. A
sensor grouping will result into reduction of computational cost. In our algorithm, we sense
unknown environment with N Sensors to extract more information about surroundings. At the
same time, we resize sensors into M groups (M< N) before giving as input to intelligent system to
reduce computational complexities still retaining the essence of more information. As the final
value for each of M group, minimum value among the corresponding sensors readings are taken
and then fed to intelligent system module. For figure.3 (b) Considering d(i)–ultrasonic data for ith
sensor; distances to the obstacles may defined as below:
Left_obs = min{d (i)} where, i= 1,2…x.
Front_obs = min{d (i)} where, i= x+1,x+2 …y.
Right_obs = min{d (i)} where, i= y+1, y+2,.. N (6)
2.3 Quantization of Sensors Values
In our approach, we perform quantization to provide discrete samples for neural training.
Quantization formula for groups (Xi) where, i=1, 2...M (M<=N) is as follows:
Xi = 1 for 0< di <=D1,
2 for D1 < di <=D2,
3 for D2 < di <=D3,
……………. . .
Z for di >DZ. (7)
7. M. M. Joshi & M. A. Zaveri
International Journal of Robotics and Automation (IJRA), Volume (2) : Issue (3), 2011 134
Where, di is the minimum sensor value of the ith group and D1, D2 … DZ are threshold values for
quantization.
2.4 Defining Heading Angle
We define heading angle (θhead) as follows:
• If θhead < p then θhead = α,
• If p <= θhead <= q then θhead =β,
• If θhead > q then θhead =γ (8)
Once surrounding environment sensing is completed; set of information is available for planning.
Next step is to train intelligent system with these set of information. As stated earlier, neuro-fuzzy
systems have abilities to learn and then perform intelligent task based on learning. Next
subsection describes neuro-fuzzy based system.
3. NEURO-FUZZY SYSTEM FOR REACTIVE NAVIGATION
Neural and Fuzzy based hybrid systems have been used in many applications in order to take
advantage of individual systems. This motivates us to use combined neuro-fuzzy system for
reactive navigation of a mobile robot in the presence of obstacles. We propose two stage, hybrid
neuro-fuzzy system in which information from an environment (Sense) is obtained by neural
networks while; more correct decisions (Act) are performed using a fuzzy system. As far as
environment understanding (Sense) is concerned; neural network will be more effective
candidate; as it gives computationally cost effective solutions than fuzzy system [30]. On the
other had, we require tight control to exert final wheel velocity where fuzzy system would be
better choice because of its functional mapping ability [31]. Our proposed framework provides an
optimum learning of neural networks via discrete sampling that overcomes the problems faced by
existing neuro fuzzy approaches based on experimental and heuristic bases training [25-28].
We consider two stages neuro-fuzzy based hybrid architecture as shown in Figure.4. In our
proposed neuro-fussy system, the inputs from the sensors are fed to neural network which forms
first stage of the proposed sys-tem and it is cascaded with a fuzzy system to generate final
control action. First stage neural network has four inputs. Out of four inputs, three inputs are the
distance information from the left, front and right obstacles pre-sent in robot’s perceptual
environment. The fourth input is the heading angle. As an output, neural network generates
Reference Heading Angle (RHA); an inferred angle than original head angle. During this process ,
as heading angle inference is already been processed by neural network and bettered with the
support of local sensory data, resulted reference heading angle imparts better information to the
subsequent fuzzy system than an actual heading angle. Heading angle is one of critical
parameters and should be inferred correctly for reactive robot navigation. The neural network
does this task and provides a reference heading angle as an output. In the second stage, fuzzy
logic processes this information and drives the output wheels of mobile robot. Outputs of the
system are left and right wheel velocities. Input sensory information’s cardinality for the neural
and fuzzy networks can be shared or can be set to higher value for neural network to take
maximum advantage of its learning capabilities.
8. M. M. Joshi & M. A. Zaveri
International Journal of Robotics and Automation (IJRA), Volume (2) : Issue (3), 2011 135
FIGURE 4: Two stage Neuro-Fuzzy System
3.5 Training Using Neural Network
In our framework, neural network has M inputs (one per each of sensor group) derived from
grouping of N sensors and giving the distance information about robot’s perceptual environment.
One more input given to the neural network is the heading angle. Neural network processes
these inputs and generates a reference heading angle. Neural networks have got remarkable
generalization capabilities, once trained properly. Training of intelligent system is crucial for
successful navigation of mobile vehicle. Generally, it is difficult to train such system as the input
space may contain infinitely many possibilities and mobile robot needs to learn effectively for
successful navigation. Many times mobile robot needs to execute operations in hazardous
environments like fire or space missions where, online training is not feasible. Off line training is
only possibility in such cases. Mobile robot needs to sense environment in real time and also to
make precise decision based on learning.
Various training approaches reported in literature are of following categories: a) generating
training sequences by experimental set up as in [17] and b) heuristic approach based on expert
rules [25]. In the first approach, the system learns by setting the different environmental set ups.
i.e. different start, end (target) positions, different obstacles positions etc. In this case, the number
of training patterns resulted for different input conditions may not be evenly distributed. Some of
the input patterns may appear more number of times, while some may appear lesser or even may
not appear. Training may not be considered optimum as; for some inputs patterns are not learnt
while some are over learnt. In case of second alternative (training by expert rules [8]), training is
per-formed using fewer number of input patterns. This type of training may save training time,
may give good performance in some cases but, they may not perform well in all kind of
environmental conditions. This is because of the fact that selection of training pairs is for a
particular task and they do not represent entire space uniformly.
In this paper, we propose mobile robot’s training based on discrete uniform sampling that
overcomes the problems with above mentioned methods. The proposed algorithm not only takes
samples from the entire sample space (to provide heterogeneity), also takes equal number of
sample data from all possible input space (to pro-vide homogeneity). In the proposed algorithm,
9. M. M. Joshi & M. A. Zaveri
International Journal of Robotics and Automation (IJRA), Volume (2) : Issue (3), 2011 136
actual sensor readings are considered to be quantized in to n linguistic values. Uniform sampling
of these quantized values will enable us a) to consider entire space of input region and; b) will
enable us to generate optimum number of training pairs required for training. In the proposed
approach, we train the network as follows:
1. Let input cardinality (number of inputs i.e sensors plus heading angle ) of the neural
networks equal to M+1 and each input takes Z linguistic values (e.g. near, medium, far-
as discussed in earlier section). Then we can generate total Z M+1
training patterns.
2. Output values of each of these input patterns are decided based on experimentation or
by expert rules.
3. Neural network is trained accordingly to training pairs generated and the performance of
the network can be verified using a proper evaluating function e.g. MSE (mean square
error).
4. If any correction is required; make adjustment to step 2 and then repeat steps.
3.5 Fuzzy System (FS)
Fuzzy logic provides a formal methodology for representing and implementing the human expert's
heuristic knowledge and perception based actions. We utilize the fuzzy system as shown in
Figure.4. Out of total four in-puts, three inputs are the distance information from the left, front and
right obstacles present in robot’s perceptual environment. The fourth input is the reference
heading angle. Outputs of the network are Left and right wheel velocities. Fuzzy system needs to
define the membership functions for input and output variables. These membership functions for
input and output variables are defined in Table 1 and Table 2 respectively. Linguistic values near,
med (medium) and far are chosen to fuzzify left_obs, front_obs and right_obs. We define
linguistic values slow, med (medium) and fast to show output parameters left and right velocities.
_____________________________________________________________________________
Order Linguistic Membership Corresponding
Values Function fuzzy numbers
_____________________________________________________________________________
1 Near Trapezoidal [0.01, 0.01, 1.5, 2.0]
2 Medium Triangular [1.5, 2.0, 2.5]
3 Far Trapezoidal [2.0, 2.5, 4.0, 4.5]
TABLE 1: TRANSFORMATION RULES FOR FUZZY INPUTS
_____________________________________________________________________________
Order Linguistic Membership Corresponding
Values Function fuzzy numbers
_____________________________________________________________________________
1 Slow Trapezoidal [0.01, 0.01, 1.5, 2.0]
2 Medium Triangular [1.5, 2.0, 2.5]
3 Fast Trapezoidal [2.0, 2.5, 4.0, 4.5]
_____________________________________________________________________________
TABLE 2: TRANSFORMATION RULES FOR FUZZY OUTPUTS
10. M. M. Joshi & M. A. Zaveri
International Journal of Robotics and Automation (IJRA), Volume (2) : Issue (3), 2011 137
3.5 Behaviors Fusion based on Fuzzy Reasoning
Mobile robot moves in a given environment from start position to the end position. In order to
avoid obstacles in its path, reactive navigation is performed in response to the sensor data
perception. In order to coordinate different type of behaviors, various methods are available: i)
priorities based data fusion ii) inhibiting strategy and iii) behavior based fuzzy reasoning. In
priority based fusion, certain rules are always given priorities compared to others which may not
be true always. In second case, when multiple rules are fired simultaneously, few rules are
dominating and hence other rules are inhibited at the particular stage. In both the cases, enough
attention may not be given to some rules which in turn may become critical after some period.
In our work, we have used behavior based fuzzy reasoning in which all fired rules are given due
weight age according to their firing level. For our proposed method, the following behaviors are
realized: Target Steer, Obstacle Avoidance and Edge following. It is very difficult to acquire
precise information about dynamic environments through ultrasonic sensors. A set of fuzzy logic
rules to describe various behaviors are defined for the proposed system. Table 3 gives few
samples of our defined fuzzy rules. These fuzzy rules show that the robot mainly adjusts its
motion direction and quickly moves to the target if there are no obstacles around the robot. When
the acquired information from the ultrasonic sensors shows that there are no obstacles to the left,
front or right of robot, its main reactive behavior is target steer. When the acquired information
from the ultrasonic sensors shows that there exist obstacles nearby robot; it must try to change its
path in order to avoid those obstacles (i.e. Obstacle Avoidance behavior). When the robot is
moving to a specified target inside a room or escaping from a U-shaped obstacle, it must reflect
Edge Following behavior.
If input then output
Rule
no..
Fuzzy
Behaviour
Left
Obs.
Front
Obs
Right
Obs.
Head
ang
Left Vel Right Vel
1 Target Steer Far Far Far Negative Low Fast
2 Target Steer Far Far Far Zero Fast Fast
3 Target Steer Far Far Far Positive Fast Low
4 Obstacle
Avoidance
Near Near Far Negative Fast Low
5 Edge
Following
Far Far Near Positive Med Med
TABLE 3: Fuzzy If-then rules
4. SIMULATION RESULTS
In this section, we demonstrate the effectiveness, robustness and comparison of various systems
for robot navigation using single stage neural network, single stage fuzzy system and our
proposed hybrid neuro-fuzzy system. We have considered mobile robot having differential drive
mechanism with span of mobile robot 50 cm. Total 9 ultrasonic sensors (N) are used for the study
after comparing the results for 5, 9 and 15 no. of sensors and their effectiveness. These sensors
are equally separated by θs = π/8 and detect the distance of obstacle along the radial direction up
to 300 cm. The wheels can have a maximum velocity up to 30 cm/s. Input dimensions to the
neural, fuzzy and neuro-fuzzy system are set to four. Dmin is set to 100 cm and Dmed is set to
200cm. In order to define heading angle (θhead), we have set the values of p, q, α, β, γ to -π/8,
+π/8, 1, 2, and 3 respectively. For our simulation we use two layers feed forward back
propagation network (FF- BPN) for mapping the input quantized values to the output. Batch mode
of training is used for neural network.
For neuro-fuzzy system, we have trained first stage neural network by considering 4 inputs as
described earlier and each input takes 3 linguistic values (near (1), medium (2), far (3)). Hence,
11. M. M. Joshi & M. A. Zaveri
International Journal of Robotics and Automation (IJRA), Volume (2) : Issue (3), 2011 138
total 81 training pairs are generated and used for optimal training of neural network. For the
second stage of our neuro-fuzzy architecture, i.e., fuzzy system, the fuzzy rules are generated
using 3 linguistic values and 4 inputs. Total 3 groups are formed in order to give them as inputs to
fuzzy system module. We use the minimum value among the corresponding sensors’ readings as
the final value for that group of sensors which is fed to the system module. Left, front and right
obstacles are equally important inputs to the fuzzy systems. The fourth input to the fuzzy network
is RHA which is output of neural network stage. The membership function values are fine tuned
by simulating the navigation in many different setups and correcting the errors over number of
experiments. For fuzzy reasoning Min – Max (Min- for the implication and Max- for aggregation)
are used. De-fuzzification is done using centroid method. Using fuzzy reasoning, various
behaviors are weighted to determine final control variables i.e. left and right velocities.
As stated earlier, we compare it with single stage neural network and single stage fuzzy based
systems. Figure.5 shows the path comparison of a mobile robot between single stage neural [30]
and fuzzy approaches [31] while; figure.6 depicts the mobile robot path comparison between
neural and proposed neuro-fuzzy based systems. These results suggest that, in the case of
second stage (driving stage), fuzzy systems are preferred. This is because of the fact that the
neural network’s output in the unexplored regions of inputs is not predictable and error at each
stage gets accumulated and hence, do not give good and stable path.
FIGURE 5: Comparison of Robot navigation: Neural
& Fuzzy system
FIGURE 6: Comparison of Robot navigation: Neural
and Neuro Fuzzy system
Figure.7 illustrates robot navigation with fuzzy system [31] while; figure. 8 shows robot navigation
with proposed neuro-fuzzy system. Comparing the results, it is found that in figure.7 robot
eventually strikes with the obstacle located to the left bottom corner while with the same scenario;
the Neuro- fuzzy system avoids the same obstacle successfully. It is because of the fact that in
the case of a single stage fuzzy system, one of the inputs (i.e. heading angle) contradicts to the
perception by the other inputs while; in the case of neuro-fuzzy system (as shown in figure.8) the
RHA, which is an inferred heading angle, has been proved very useful input to the fuzzy system.
The use of neural network as first stage in neuro–fuzzy system architecture provides better
inference of an environment using the sensed input values. The simulation results highlight the
fact that adding the neural stage enhances environmental sensing capacity of the fuzzy system.
The same fact is observed from the outputs of various experiments performed in different
environmental conditions.
12. M. M. Joshi & M. A. Zaveri
International Journal of Robotics and Automation (IJRA), Volume (2) : Issue (3), 2011 139
FIGURE 7: Robot navigation with single stage fuzzy
system
FIGURE 8: Robot navigation with two stage Neuro-
Fuzzy system
FIGURE 9: Neuro-fuzzy based mobile robot navigation
Next subsection presents the detailed behavior analysis of proposed neuro-fuzzy based systems
that highlights the effectiveness of our proposed system in given environment.
A Various Fuzzy Based Behaviors and Heading Angle
Consider mobile robot navigation for the case shown in figure.9. Mobile robot starts its journey
from position “START” to the final position “F”. For a given scenario, mobile robot follows path
from START-> A -> B -> C -> E -> F as shown in figure.9. Figure.10 shows mobile robot’s
heading angle during its journey. Head-ing angle is the difference between target and head of the
13. M. M. Joshi & M. A. Zaveri
International Journal of Robotics and Automation (IJRA), Volume (2) : Issue (3), 2011 140
mobile robot and provides information about current head orientation. Initially, robot performs
“Target steer” behavior and reaches to position “A” with “ZERO” heading angle, where it takes
right turn which is a result of “avoid obstacle” behavior; and heading angle changes quickly to -90.
There after, the robot follows the wall, i.e., the wall following behavior, and it reaches to “B”. At
the same time, heading angle slightly varies from -90 to -100 degrees. At this point “B”, it takes
right turn again (head angle (equal to α) changed to -180) and following the wall and reaches to
“C” by decreasing heading angle (equal to β) further to -200 degrees. Mobile robot finds the end
of the wall and perceives potential attraction by the target and takes left turn by avoiding contact
with obstacle and reaches to “D”. During the same, the heading angle (equal to γ) starts
increasing to -45 degrees. From position “D” to “E” it continues its quest to reach target following
wall (heading angle (decreases slightly from -45 to -50 degree (δ)), finds opening at “E”. Finally
mobile robot reaches from “E” to “F” with “target steer” behavior (first reducing head angle to near
zero and then with almost zero head reaches target “F”).
FIGURE 10: Mobile robot’s heading angle
FIGURE 11: Left and right velocity control over time for mobile robot navigation
Speed Control of Mobile Robot
In mobile robot navigation, speed control analysis gives information about and robot’s left and
right velocities over the time. Figure.11 shows left and right wheel speed control. In differential
drive mechanism, to take right turn; robot increases its left velocity and decreases right velocity
and vice versa. As shown in figure11, from “START” position to point “A”, left and right velocities
14. M. M. Joshi & M. A. Zaveri
International Journal of Robotics and Automation (IJRA), Volume (2) : Issue (3), 2011 141
are found same. From “A” to “B” position (first, mobile robot takes right turn and then follows
straight line). Hence, initially left velocity is higher than right velocity and then both almost being
same till “B”. At “B” robot has left velocity higher again (avoid obstacle- right turn) and then with
almost same left and right velocities (wall following) reaches “C”. Journey from “C” to “D” is with
right velocity values higher than left velocity values (avoid obstacle /attraction potential - left turn).
Mo-bile robot moves straight with same left and right velocity values (wall following) from point “D”
to “E”. At last, at point “E”, right velocity values are more than left velocity values (right turn) and
finally, it settles to point “F”.
Target Distance and Distance Traveled
Figure.12 shows mobile robot’s target distances and distance traveled over the time period for a
case discussed in figure. 9. Start and target (F) coordinates for mobile robots are (10, 2) and (10,
18) respectively. For mobile robot From “START” to “A”, target distance is linearly decreased
(START->A). When mobile robot finds an obstacle in between and trying to move out of the
same, target distance is gradually increased again till “D” (A->B->C->D). Then onwards, target
distance is again linearly decreased on path from D -> E -> F. Second graph in figure.12 shows
the total distance traveled a curve with a linear rise from start to finish point. For the given
situation, in presence of given obstacle scenario; mobile robot travels total 39.65 meters
compared to 16 meters if it had traveled without any obstacles from START to F position (i.e.
distance between (10, 2) and (10, 18) ) as shown in figure.9.
FIGURE 12: mobile robot’s target distance and distance traveled
15. M. M. Joshi & M. A. Zaveri
International Journal of Robotics and Automation (IJRA), Volume (2) : Issue (3), 2011 142
FIGURE 13(a): Typical scenario for obstacle avoidance: case01
FIGURE 13(b): Typical scenario for obstacle avoidance: case02
FIGURE 13(c): Typical scenario for obstacle avoidance : case03
Importance of Adding Input Neural Stage
Figure.13 (a-c) shows three different cases for mobile robot navigation which highlights the
importance of neural network stage. In first case, current positions of robot, obstacle and target
are as shown in Figure.13 (a). Perceptions from mobile robot sensors suggests that left and front
obstacles are near, right obstacle is far and head angle is negative. These inputs suggest that
mobile robot should move left (because head angle is negative) while; to the left and front there
are obstacles at near distances. Here, two senor input groups distances left and front obstacles,
contradict to another input, namely, head angle. Due to this fact, a single stage fuzzy system may
not be able to take best decision. When neural network is used along with fuzzy system (i.e.
neuro-fuzzy), it is observed that the output of neural stage, Reference heading angle (RHA),
becomes positive; compared to earlier input, i.e., heading angle which was negative. When this
inferred input (RHA) is given to fuzzy system for inference; now instead of contradicting with the
rest of inputs it supports the inference by ultrasonic sensors. Hence, this enables fuzzy system to
exert better control action.
In second case (as shown in figure.13 (b)), front obstacle is removed keeping other conditions
same as in case one; first stage neural network corrects head angle to zero (indicating go straight
i.e. wall following behavior) from negative (go left i.e target steer). As a third case (as shown in
figure..13 (c)) both the obstacles are re-moved keeping robot, target positions and other
conditions are kept unchanged. Here, first stage neural network output suggests to continue
earlier perception (i.e. same as input - reference heading angle inference is negative). In
summary, neural network suggests different behaviors like avoid obstacle (move to right), wall
16. M. M. Joshi & M. A. Zaveri
International Journal of Robotics and Automation (IJRA), Volume (2) : Issue (3), 2011 143
following (go straight) and target steer (continue left) in all three cases respectively. Inferences
made by the neural stage, when in turn given as input to the subsequent fuzzy system, it
strengthens fuzzy system’s local perception provided by the local sensors. Resulted neuro-fuzzy
system performs better than single stage neural or single stage fuzzy systems. These results
highlight the importance of adding a neural stage before the fuzzy control stage in the proposed
system.
5. CONCLUSIONS
In this paper, an approach for robot navigation using neuro-fuzzy based system is discussed. The
mobile robot performs reactive navigation which is very useful for real time, dynamic environment
rather than looking for an optimal path as performed by path planning techniques. Fuzzy system
architecture for behavior based control of robot navigation gives better performance compared to
neural based systems. Neural network’s output in the case of unexplored regions of inputs is not
predictable and error at each stage is accumulated. As a result it does not lead to good and
stable navigation path. The performance of mobile robot navigation system is improved by
cascading the neural network and fuzzy system. The simulation results show that RHA provides
better inference compared to original heading angle. The behavior based analysis of mobile robot
navigation using the proposed neuro-fuzzy system demonstrates the excellent performance in
complex and unknown environments. Simulation results for dynamic, complex and cluttered
environment of mobile robot navigation space with neuro-fuzzy based architecture demonstrate
good performance compared to most recent comparable approaches. This is because of our
generalization of most of the parameters likes number of sensors, threshold values to measure
distances and heading angles, optimum training using discrete sampling based approach for
neural system training.
6. REFERENCES
[1]G. Dudek and M. Jenkin. Computational Principles of Mobile Robotics, Cambridge university
press, 2000, pp. 01-40.
[2]C.M. Clark & S. M. Rock. “Motion Planning for Multiple Mobile Robots using Dynamic
Networks”, Proc. of IEEE international Conference on Robotics and Automation, 2003,
pp.4222-4227.
[3]A. Haddad, M. Khatib, S. Lacroix and R. Chatila. “Reactive Navigation in Outdoor Environment
using Potential Fields”, in proc. of IEEE international Conference on Robotics and
Automation, 1998, pp. 1232-1237.
[4]G.Theocharous and S. Mahadevan. “Approximate planning with hierarchical partially
observable Markov decision process models for robot navigation”, IEEE International
Conference on Robotics and Automation, 2002. Proceedings. ICRA'02. pp 1347– 1352.
[5]S.M.Chun, D.Y. Huang, C.H. Chou and C.C. Hsieh. “A reinforcement-learning approach to
robot navigation”, IEEE International Conference on Networking, Sensing and Control,
2004, pp. 665 – 669.
[6]P. Dahm, C. Bruckhoff and F. Joublin. "A neural field approach for robot motion control," IEEE
International Conference on Systems, Man, and Cybernetics, 1998, pp.3460-3465
[7]E. Zalama, J. Gomez, M.Paul and J.R. Peran. "Adaptive behavior navigation of a mobile
robot," IEEE Transactions on Systems, Man and Cybernetics, Part A: Systems and
Humans, vol.32, pp.160-169, Jan 2002.
[8]S.X. Yang and M. Meng. "Neural network approaches to dynamic collision-free trajectory
generation," IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics,
vol. 31(3), pp.302-318, Jun 2001.
17. M. M. Joshi & M. A. Zaveri
International Journal of Robotics and Automation (IJRA), Volume (2) : Issue (3), 2011 144
[9]S.X. Yang and M. Meng. "Real-time collision-free motion planning of a mobile robot using a
Neural Dynamics-based approach," IEEE Transactions on Neural Networks, vol.14 (6), pp.
1541- 1552, Nov. 2003.
[10]J.C. Latombe. Robot Motion Planning, kluwer academic publishers- 1991.
[11]M. Khatib, "Real-time obstacle avoidance for manipulators and automobile robots", Int. J. of
Robotics Research, vol.05 (1), 1986.
[12]Saffiotti. ”The uses of fuzzy logic in autonomous robot navigation”, International journal on
Soft Computing”, vol. 1, pp. 180-197, 1997.
[13]I.I.Ismail and M.F. Nordin. "Reactive navigation of autonomous guided vehicle using fuzzy
logic," Student Conference on Research and Development, 2002, pp. 153- 156.
[14]G. Mester, “Obstacle Avoidance and Velocity Control of Mobile Robots”, proceedings of 6th
international symposium on intelligent Systems and Interpretation, Sep. 2008, pp.1-5.
[15]P.G. Zavlanga, S.G.Tzafestas , K. Althoefer , “Fuzzy Obstacle Avoidance and Navigation for
Omni directional Mobile Robots”, ESIT, 2000.
[16]S. Ishikawa. "A method of indoor mobile robot navigation by using fuzzy control," Proceedings
of Intelligent Robots and Systems IROS '91. vol.2, 3-5, 1991, pp.1013-1018.
[17]W. Li. “Fuzzy Logic based Perception-Action Behavior Control of a Mobile Robot in Uncertain
Environments,” IEEE International Conference on AI, 1994, pp. 231-235.
[18]T. Lee and C. Wu. "Fuzzy Motion Planning of Mobile Robots in Unknown Environments,”
presented at Journal of Intelligent and Robotic Systems, pp.177-191, 2003.
[19]H. Li and S. X. Yang. “A behavior-based mobile robot with a visual landmark recognition
system,” IEEE Trans. Mechatronics, vol. 8, no. 3, pp. 390–400, Sep. 2003.1.
[20]C.H. Lin and L.L. Wang. “Intelligent collision avoidance by fuzzy logic control,” Robotics and
Autonomous Systems, Volume 20, pp 61-83, 1997.
[21]A. Zhu and S.X. Yang. "A fuzzy logic approach to reactive navigation of behavior-based
mobile robots," IEEE International Conference on Robotics and Automation, 2004. vol.5,
no., pp. 5045- 5050.
[22]M.Wang and J.N.K.Liu. “Fuzzy logic based robot path planning in unknown environments”, in
Proceeding of International Conference on Machine Learning and Cybernetics,
Vol.2,2005,pp.813–818. 1.
[23]W.L. Xu, S.K.Tso and Y.H. Fung. "Sensor-based reactive navigation of a mobile robot
through local target switching," Proceedings of 8th International Conference on Advanced
Robotics, ICAR '97, 1997, pp.361-366.
[24]E.O.Motlagh, T.S.Hong and N.Ismail. “Development of a new minimum avoidance system for
a behavior-based mobile robot,” Proceedings of international journal on Fuzzy Sets and
Systems, Vol. 160,issue 13,pp.19129-1946,July 2009.
[25]K.T. Song. and L.H. Sheen. “Heuristic Fuzzy–neuro network and its application to reactive
navigation of a mobile robot,” International journal on Fuzzy Sets and systems Vol. 110, pp.
331-340, 2000.
18. M. M. Joshi & M. A. Zaveri
International Journal of Robotics and Automation (IJRA), Volume (2) : Issue (3), 2011 145
[26]W. Li, M Chenya and F.M. Wahl. “A Neuro- Fuzzy system architecture for the behavior
based control of a mobile in unknown environment,” International journal on Fuzzy Sets
and systems, Vol. 87, pp.133-140, 1997.
[27] G. N. Marichal, L. Acosta, L. Moreno, J. A. M´endez, J. J. Rodrigo and M. Sigut. “Obstacle
avoidance for a mobile robot: A neuro-fuzzy approach,” International journal on Fuzzy Sets
and Systems, vol. 124, no. 2, pp. 171–179, Dec. 2001.
[28] Zhu and S.X. Yang. "Neuro fuzzy-Based Approach to Mobile Robot Navigation in Unknown
Environments," IEEE Transactions on Systems, Man, and Cybernetics, Part C: Applications
and Reviews, vol.37, no.4, pp.610-621, July 2007.
[29] M.M.Joshi and M.A. Zaveri. “Neuro-Fuzzy Based Autonomous Mobile Robot Navigation”,
IEEE 11th International Conference on Control, Automation, Robotics and Vision, ICARCV
2010 , Singapore, Dec 2010.
[30] M.M.Joshi and M.A. Zaveri. “Optimally learnt, neural network based autonomous mobile
robot navigation system”, International Conference on Advances in Electrical & Electronics
, AEE 2010 , 20-23 December 2010.
[31] M.M.Joshi and M.A. Zaveri. “Fuzzy Based Autonomous Mobile Robot Navigation”, IEEE
India Conference INDICON2009, Dec. 2009.