A fuzzy UV-disparity based approach for following a leader mobile robot

Following a leader is a very important task in the autonomous robot navigation. In order to follow a leader, the follower has to detect the leader, to localize it and to estimate its properties. Several techniques have been used to carry out this task; in particular vision based approaches can be used. In this paper, we describe a stereovision based approach for following a leader. The main idea of this paper assumes that the leader is considered as an obstacle. So, the V-disparity obstacles detection approach is combined with the U-disparity based localization and a fuzzy controller in order to execute all the steps of the leader following task. The described approach was successfully implemented on a Pionneer3AT robot.