Networked human motion capture system based on quaternion navigation

Abstract

In this paper, from the perspective of human ergonomics, we analyze the movement of the joints in the process of human body movements, and we establish a dynamic model according to the human skeleton structure. On this basis, from the rigid body dynamics point of view, combined with the principle of inertial navigation, a body sensor network based on MEMS inertial sensors is built to capture human body motion in real time. On the basis of space trajectory of human body movement and traditional human motion solution strategy, a human motion solution strategy based on particle filter fusion solution is proposed to realize the prediction of human motion analysis. Therefore, we evaluate the performance of the designed system by comparing with the real motion. Finally, in order to verify the human motion data, the motion capture data verification platforms are established. Experimental results show that the proposed joint attitude solution algorithm can achieve a relatively smooth tracking effect and provides a certain reference value.