Study on badminton system with auxiliary training based on Kinect motion capture

Jiang Dong

COMPUTER MODELLING & NEW TECHNOLOGIES 2013 17(5D) 97-100

Yulin University, Shaanxi Yulin, 719000, China

First, the paper studies the algorithm to repair self-occlusion body joint information. Due to the existence of human self-occlusion, the motion capture data is not credible. According to the invariance of the length of human body skeleton and the continuity of human movement, this paper proposes a quick geometric method to repair the skeleton information. The experiments show that the algorithm can real-time repair more than half of the obscured joints information, especially the end of the joint information. Then, study the movement redirection based on constraints of terminal effectors. Different human motions are assigned to another role, it will case feet penetrate the ground, sliding, skin stretching or distortions due to bone size inconsistencies. Base on bone length scale factor of virtual role and Capture body, this thesis proposes a movement redirect method of end constraint effectors. First determine the end constraint of the first frame, and get the three-dimensional coordinate information of the end constraint; secondly, based on the length of the human skeleton invariance redirect the three-dimensional coordinates of the other joints. Third, determine the end constraint of the next frame and get its three-dimensional coordinates; finally, calculate other joints three-dimensional coordinates of other frames. The experimental results show that the algorithm can solve the problem of distortion animation, and get standard capture motion data, which will conducive to data analysis.