A Human Motion Capture System Based on MEMS Inertial Measurement

Abstract

Motion capture system is usually used to measure the body posture and track the trajectory of motion in three-dimensional space. With the emergence of MEMS inertial sensors that is low power consumption and high integration, the motion capture system based on MEMS inertial measurement has been applied various market prospect. In this paper, we realized a human motion capture system based on inertial sensors, which is low-cost, real-time. The system is mainly composed of sensor measurement unit and pose estimation unit on PC. The sensor unit exchanges data with PC through general WiFi network. A sensor unit includes 9-axis inertial sensors (three-axis gyroscope, three-axis accelerometer and three-axis magnetometer), MCU and network communication module. The sensor units were attached in some parts of the human body to collect the 9-axis information of limbs in real time. The 9-axis information was transmitted to the pose estimation unit through the network communication module. The orientation of single sensor unit was estimated although madgwick algorithm. The pose estimation unit used human skeleton model and vector representation to reconstruct and track the whole-body posture. Finally, qualitative and quantitative tests were carried out. Experiments showed that the system had good dynamic characteristics and high precision.