AR-MoCap: Augmented Reality Motion Tracking System

Motion tracking is critical for applications in sports science, clinical diagnostics, and data acquisition for the AI model. However, traditional methods require costly devices, complicated wearable markers, and controlled environments, which limit their accessibility in reality.

This study proposes a lightweight real-time motion tracking system using a head-mounted display (HMD) with visual overlay in the physical environment. Only with a simple calibration, the full-body motion capturing and point cloud visualisation could be achieved, with a built-in joint angle calculation for analysing the subject's kinematics. The system was evaluated for stability and latency, while the tracking kinematics was evaluated for accuracy, continuity and consistency.

The experimental results showed the reliability of both the system design and the tracking results. Overall, the proposed system offers a portable, convenient, and efficient motion tracking solution. It shows the potential for both the clinical kinematic diagnostics and conventional motion-tracking applications that require flexible, changing viewing perspectives.