Motion+: Robust 3D Motion Tracking using Depth Estimation in a Single Camera Setup

The accurate conversion of translations in camera coordinates of a 3D motion tracking marker into a displacement vector is critical for applications in biomechanics, robotics, and virtual reality. Current methods often suffer from significant inaccuracies due to calibration errors, occlusions, and sensor noise. These inaccuracies compromise the reliability of motion analysis. This paper analyzes existing approaches, identifies limitations, and proposes a novel hybrid algorithm that leverages depth mapping, calibration constants, and quaternion rotation to enhance accuracy and robustness. Simulation-based validations demonstrate the efficacy of the proposed method in delivering reliable 3D displacement vectors across multiple conditions.