Visual Guidance of an AGV

Image processing plays an important role in the visual navigation of a mobile robot. A real-time visual tracking system, using an active stereo vision head mounted on a mobile robot, is described in this paper to demonstrate the real-time implementation of a classic stereo method, featuring algorithm simplification for a real application. Disparity information is used directly to guide visual tracking by integrating tracking tightly to its final purpose, to provide control signals for obstacle avoidance by the mobile robot. Using an approach proposed here, the stereo correspondence algorithm can be applied to compute disparity information in real time from stereo images captured from active stereo system with two cameras verging dynamically.