Color Tracking Vision System For The Autonomous Robot

As the development of computer technology and camera equipment, the research of robot visual servo system is paid close attention in the recent years .The great progress has been made in the field of robot visual servo system both in theory or applications. It has gradually become a comprehensive independent disciplines came down to technology of the robot, automatic control and image processing.The task of the color tracking vision system discussed in this paper is to extracting the target in the original picture got by camera according to its color characteristics, after the successful identification to the target, two motor under the bottom of the robot is driven to make the robot track the target. The system can be applied to a number of occasions such as intelligent monitoring and robot football. This system is made of two parts: the image segmentation system based on the HSI color space and motion control system based on fuzzy control.The image with RGB format collected by camera and image capture cards is converted to HSI format by image segmentation system, and the target is segmented according to the threshold of hue and the threshold of saturation. The impact caused by the change of illumination is eliminated by using way. The run-length coding technique is used in image fusion process to ensure the real-time of the system. And the coordinate of barycenter and area of target is calculated in the image fusion process which provides the original data for the robot path planning.The input of motion control system is the barycenter and area of target provided by the image segmentation system. Which will be compared to the desired area and the desired barycenter to cause an error .then a closed-loop motion control systems based on the vision is formed .Size and center coordinates and the coordinates of the size and center of gravity of error than the right to assembly error of a -based closed-loop motion control systems, the fuzzy rules of the fuzzy control system is established to imitate tracking habit of human being. Finally, the experiments is carried out on MT-R intelligent robot, the results of the experiments show that the target can be identified by the image segmentation system in a complex environment when illumination conditions is changing, and the robot can be driven to track the target in the real time.