Moving object detection and tracking in mobile robot system based on omni-vision and laser rangefinder

Abstract

To meet the requirements of real-time and accurate information processing in robot tracking system, a real-time object detection and tracking method based on Multi-Sensor Information Fusion technique is proposed. Firstly, the real-time environmental data was obtained by using omni-vision and laser rangefinder, so that the moving object was detected; Secondly, the average color of color-similar point set within the neighborhood of a sampling point on the moving target was extracted as color feature, and the color feature was enhanced by using the real-time data obtained from the laser in the direction of object. Finally, the orientation and distance information were calculated between robot and the moving object, and the moving object tracking was achieved by using the robot tracking tactic. Experiments were carried out on the AS-R mobile robot, and the experimental results show that the proposed method can meet the requirements of real time and robust in the robot tracking system. © 2012 Springer-Verlag GmbH.