Human Motion Capture Algorithm Based on Inertial Sensors

Abstract

On the basis of inertial navigation, we conducted a comprehensive analysis of the human body kinematics principle. From the direction of two characteristic parameters, namely, displacement and movement angle, we calculated the attitude of a node during the human motion capture process by combining complementary and Kalman filters. Then, we evaluated the performance of the proposed attitude strategy by selecting different platforms as the validation object. Results show that the proposed strategy for the real-time tracking of the human motion process has higher accuracy than the traditional strategy.