Research on Motion Capture System of Motor Skill Based on Computer Vision Technology

Abstract

Compared with the existing motion capture system, the inertial conduction based motion capture method proposed in this project can effectively overcome the shortcomings of the current motion capture, such as limited communication distance, low accuracy of data acquisition, low visualization degree of model display software, narrow application range, etc., and improve the practical application value of the method. A method of motion capture and motion control based on wireless network is proposed. Since most of the skeleton of the human body is rigid, the complete expression of the human movement is realized through the study of 17 important joints, and all the movement information of the athlete can be obtained by placing sensors on the corresponding nodes. Through wireless network transmission technology, the player's action information is transmitted to the 3D modeling software to achieve real-time control of the action. When measuring each sensor, strapdown inertial navigation technology is used to improve the accuracy of acquisition. The 3D modeling software is based on a professional motion capture software, and through the forward and inverse kinematics of its motion, the 3D modeling is highly consistent with the wearable device.