Vision-Based Leader/Follower Tracking for Nonholonomic Mobile Robots

This paper presents a strategy for a nonholonomic mobile robot to autonomously follow a target based on vision information from an onboard pan camera unit (PCU). Homography-based techniques are used to obtain relative position and orientation information from the monocular camera images. The proposed kinematic controller, based on the Lyapunov method, achieves uniform ultimately bounded (UUB) tracking.