Low Speed Mobile Robot Recognition And Tracking System Design Based On Visual Information

Based on summarizing the methods of mobile robot recognition and tracking all over the world, this thesis proposes a low speed mobile robot tracking system using color character. It uses CCD camera and image grabber to capture video information of the aimed robot and the information is transferred from analog to digital, then by running correlative program, the mobile robot recognition and tracking can be realized and the result can be displayed on screen at the same time.As we all know, color image contains more information than gray image, the color character is chosen as the first one for robot recognition and tracking by Threshold segmentation. After analyzing the character of all color space model, the character Hue of HSI space model is selected, it is the most stable and the most obvious character. In order to get more accurate results, an algorithm called two-time search is used for tracking, it is based on the theory of data access in double linked list, data structure. In order to get the accurate position, the processing results of the previous frame are used to forecast the position of the next time. The speed of running digital image processing program is very important for this system, the method to solve the problem is dealing with every three pixels when comparing with the threshold.In this tracking system perspective distortion is the one that more common than any other kind of distortion, so a camera calibration method based on geometric mapping is proposed. By doing this, the information of the aimed robot under the image coordination is transferred to the real world.The software system contains the following two parts: improved static image processing software base on Imgcx and applying software for the whole system. The latter can control the hardware to realize the recognition and tracking of the aimed mobile robot.