The Development And Design Of Motion Capture System Based On Sensors

This paper mainly discusses the related research and design of the motion capture system based on MEMS sensors. The system is based on unique micro-inertial sensors and wireless transmission systems, has the advantages of Low cost, easy operation, high accuracy, real-time high.This paper discusses the method of the three-dimensional human skeleton model construction and posture calibration, elaborates the process of the system development and design. Main contents and works are included:(1) Studied the working principle and parameter configuration of the MEMS sensors. In order to collceting accurate and real-time motion data, selected the small size and sampling accurately sensors as the equipment of the hunman body motion tracing according the hardware configuration of the PC termination. According to the principle of the sensor hardware, implemented the data transmission module based on wireless transmission, thus effectively avoiding the restrictions of data transmission line in the process of human motion capture.(2) Researched the basic method of modeling the human body for human motion tracking technology, on this basis designed a three-dimensional hunman skeleton model based on the kinetic theory. The model uses a hierarchical tree structure, while adding the human constraints of each joint, thus with it the system can achieve realistic simulation of human motion.(3) Studied the method of posture calibration used in human motion capture, according the system design and related theory, designed a method of posture calibration based on three-dimensional human skeleton model and MEMS sensors.(4) Analyzed the factors of the data acquisition hardware module and the external environment for data collecting, proposed a data fusion algorithm for gyro based on the magnetometer data and accelerometer data. While on the basis of Kalman filter and complementary filter, proposed a direct conversion complementary quaternion filter algorithm.The motion capture system based on MEMS sensors uses sensors to collect the hunman motion data and uses data acquisition module to process the data, then uses the wireless data transmission module to transmit the motion data to the terminal PC . The PC uses the computed data to drive the three-dimensional human skeleton model, and then to achieve the real-time human motion capture.