| Abstract: | Path recognition is an inevitable core technology in the development of tracking robot. In this paper,the path tracking system of tracking robot can be realized by image sensor module based on camera to obtain lane image information,and then extract the path through visual servo. The whole system can be divided into seven modules: micro control unit( MCU) processor module,image acquisition module,debugging module,motor drive module,servo drive module,speed sensor module,and voltage conversion module.In image pre-processing part,there is an introduction of binarization processing and the median filtering to strengthen the image information. About recognition algorithm,three key variables which are changed in the movement state are discussed and there are also many auxiliary algorithms that help to improve the path recognition.The experiment can verify that the whole system can accurately abstract the black guide lines from the white track and make the robot moving fast and stable by following the road parameters and conditions. |
