Binocular Vision-based Mobile Robot To Track Moving Objects Research

From the 80's, many countries had plans to research the technology of mobile robot. It involves in image processing, computer vision, sensor technology, artificial intelligence, automation, parallel computer processing technology, mechanism and so on. It reflects the newest achievement of information science and the artificial intelligence.The issue of mobile robot based on binocular vision to track moving object is mainly studied in this thesis, the tracking object is another mobile robot. Firstly, Kalman filter is used to forecast the position and the speed of the tracking object, then a scene on the frame is selected and the image based on the forecasted position is cut. The results of image processing show that better image processing results can be obtained when the methods of median filter and threshold are applied to smooth and segment the images respectively. Furthermore, five features which are easy to recognize the object successfully is chosen, and the spatial position of the object is calculated according to the conversion among the different coordinates (camera coordinate, image coordinate and the world coordinate). Finally, control strategy is designed for the mobile robot, and the moving object is tracked exactly.All the methods of image processing-are realized by VC language.