Vision Target Detection And Tracking Based On NAO Robot

With the development of science and technology, robots in industrial, military and outer space probe and other fields has been widely used. Humanoid robots as the robots which were most similar with human can replace human repetitive high, dangerous work. Robot platform used for the NAO humanoid robot in this article,based on research of robot vision algorithms, focus on the target detection and tracking algorithms of humanoid robot.In this thesis, based on a number of the related literatures consulted and relevant technology, robot visual detection and tracking algorithms has been studied and implemented. First, in order to reduce the influence that changes in illumination of images captured by the robots, the paper analyzes and compares images that collected under the RGB,HIS,YUV color spaces and chooses the YUV color space which get minimal influence of illumination. Second, the paper segments the acquired images, extracts the desired target image and removes image noise interference. Finally, the contour information and the coordinates information are extracted.After determining vision algorithms, we design visual orientation and crawling experiment of NAO robot to test the robot’s vision algorithms. Firstly, the camera model of NAO robot is analyzes, according to its only use one camera features, internal and external camera parameters were calibrated, and then build a model for monocular distance of NAO robot. Secondly, the different strategies are designed such that the robot can be efficiently and accurately track the target. Motions of the robot during tracking needs are written by Choregraphe software. Finally, the whole framework of NAO robot tracking system is designed, and the NAO robot crawling tracking experiment is designed, after analysising the experimental results, it shows that the target tracking system can track the target accurately.