A Tracked Mobile Robot with Vision-based Obstacle Avoidance

Abstract

This paper presents VITAR (VIsion based Tracked Au-tonomous Robot), a robotic test bed that consists of a tracked mobile robot equipped with a pan-tilt mounted vision sys-tem, an onboard PC, driver electronics, and a wireless link to a remote PC. A novel vision-based obstacle avoid-ance algorithm is implemented on the robot. The robot uses histograms of images obtained from a monocular and monochrome camera to detect and avoid obstacles in a dy-namic environment. As the generation of histogram is com-putationally inexpensive, the robot is quick to avoid the ob-stacles. Several experiments conducted with different obsta-cle environments validate the effectiveness of the algorithm.