Investigation of robotic mapping and obstacle avoidance using stereo vision
Loading...
Date
2019-06
Authors
Kong, Teck Seng
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Robotic Mapping is the generation of map for an unknown environment with no priori map. This can be done by analyzing the information of surrounding such as distance from obstacles. The map generated is useful for path planning in disaster relief and area with hazardous toward human. In this project, stereo vision system is being investigated for the performance on distance measurement for Robotic Mapping and Obstacle Avoidance. Two Universal Serial Bus cameras are being calibrated for stereo vision. Left and Right image pair is acquired and being processed to obtain disparity map. Then, vdisparity and u-disparity map are generated for road extraction and background extraction. The developed system able to measure the obstacles within 3m with error less than 3%. This system is implemented on wheeled robot for autonomous navigation through obstacle avoidance. The robot able to pass through the simple test field with two obstacles. The range information of stereo vision system is used to generate a 2 dimensional occupancy grid map for the test field.