Research on Human Motion Monitoring Method Based on Multi-Joint Constraint Filter Model

In recent years, with the increasing application of human motion monitoring technology in virtual reality, medical health, human-computer interaction, and other fields, higher requirements have been put forward for measurement accuracy. Aiming at the problem of large accumulative measurement error and multi-joint constraint in the traditional inertial measurement system, a multi-joint constrained filter model human motion monitoring method based on data fusion is proposed. Firstly, an accelerometer and magnetometer are used to construct the initial attitude angle calculation model of the human body at rest. Secondly, combined with gyroscope, accelerometer and magnetometer, the attitude optimal estimation model based on Extended Kalman Filter (EKF) algorithm for human body motion was established. Then, a global coordinate system of human multi-joints is established, and the joint constraint is introduced into the attitude optimal estimation model by using the estimation projection method, and the multi-joint constraint filtering model of human motion monitoring is constructed. Finally, a multi-inertial measurement unit (MIMU) human motion monitoring system was built, and the simulation and experimental analysis were carried out. The experimental results show that the maximum angular measurement error is about 1° after filtering with the motion monitoring method proposed in this paper. Compared with the existing algorithm, the accuracy of the three angle measurements is improved by 38.2%, 39.3% and 67.4%, respectively. Based on this, the human motion monitoring system designed can effectively monitor human motion.