Fuzzy Target Tracking and Obstacle Avoidance of Mobile Robots with a Stereo Vision System

In this paper, a two-level hierarchical intelligent control system is developed on a mobile robot to deal with target tracking and obstacle avoidance tasks. A stereo vision subsystem is first introduced. It can perceive video information in the environment to locate objects, including locating targets and obstacles precisely. This subsystem possesses two CCD cameras mounted on the top of a robot to obtain images from the surroundings. And then, based on the sensed image data in the indoor environment, a robot behavior decision-making subsystem with a fuzzy control strategy is proposed for target tracking and obstacle avoidance. Under this decision-making strategy, the robot can respond correctly to complete the desired task. Finally, site experiments are performed to show that the proposed schemes are feasible and effective. including a single CCD camera [7-8], an omnidirectional vision system [9-10], stereo vision [11-12], and three cameras [13]. There have been many studies done on the target tracking problem [14-21]. A stable target tracking control that considers collision avoidance with obstacle has been proposed [14-15]. In [16], an approach to tracking a person by using stereo vision has been developed. Also, a color signal has been applied to the tracking and matching of objects [17]. The target tracking system is designed to use fuzzy control algorithms [18-24]. In recent years, stereo algorithms have been widely applied because distance information can be gathered. Therefore, we put the stereo vision system on the mobile robot to improve the target tracking and obstacle avoidance. More accurate position information is obtained from the real-time stereo system. In this paper, the two-level hierarchical system is developed to execute target tracking and obstacle avoidance. The first subsystem is the vision system, which perceives the environment information including target and obstacles. Based on the target position in the indoor environment, the target tracking strategy or obstacle avoidance strategy for a mobile robot will be selected. In this subsystem, the distance between the mobile robot and the target or obstacles is considered first. Then, the second subsystem is the intelligent control of a mobile robot.