Home > Published Issues > 2020 > Volume 9, No. 8, August 2020 >

Stereo Image Partitioning based Fuzzy Logic Controller for Real-time Obstacle Detection and Avoidance

Kirti Shankar Sharma and P. V. Manivannan
Department of Mechanical Engineering, Indian Institute of Technology Madras, Chennai, India

Abstract— The research work in this paper deals with the development of a Fuzzy Logic based static and dynamic obstacle avoidance system for navigation of a mobile robot using vision. A low-cost stereo vision system is developed, calibrated and mounted on the differential drive robot, and an initial pair of images is captured. Kanade-Lucas-Tomasi tracker with stereo point cloud is used in this pair of images to initiate the robot’s motion. Once the robot is in motion, the successive stereo pair of images is partitioned. These partitioning helps to detect forthcoming obstacle that appear from the center or left/right corners. The developed Mamdani-fuzzy logic controller then identifies the obstacles using Moore-Neighbor tracing algorithm based boundary detection and subsequently provides a direction of motion to avoid them. A simulation with a stereo camera mounted on mobile robot is carried out in a virtual environment is generated using V-rep and controlled by MATLAB command via API. The algorithm has also been experimentally validated on the Amigobot with Pioneer 3-DX as dynamic obstacle by VICON tracking system in an indoor laboratory environment, and the comparison between the simulation and experimental results are discussed.
Index Terms— fuzzy logic, Kanade-Lucas-Tomasi, differential drive robot, Moore-Neighbor tracing, obstacle avoidance, stereo vision, image partition, boundary detection, VICON tracking system

Cite: Kirti Shankar Sharma and P. V. Manivannan, "Stereo Image Partitioning based Fuzzy Logic Controller for Real-time Obstacle Detection and Avoidance," International Journal of Mechanical Engineering and Robotics Research, Vol. 9, No. 8, pp. 1158-1169, August 2020. DOI: 10.18178/ijmerr.9.8.1158-1169

Copyright © 2020 by the authors. This is an open access article distributed under the Creative Commons Attribution License (CC BY-NC-ND 4.0), which permits use, distribution and reproduction in any medium, provided that the article is properly cited, the use is non-commercial and no modifications or adaptations are made.