Effect of human link length determination on posture reconstruction

Area of Science:

• Biomechanics
• Human Motion Analysis
• Computational Simulation

Background:

• Motion capture data (marker positions, joint centers) are used to validate digital human simulations.
• Posture reconstruction algorithms map Cartesian space to joint space for direct comparison.
• Inaccurate link lengths in simulation models introduce significant errors.

Purpose of the Study:

• To identify the optimal method for calculating link lengths to enhance the accuracy of digital human simulations.
• To compare the effectiveness of different link length calculation methods.

Main Methods:

• Developed and compared three link length calculation methods: experimental-average, trial-specific, and T-pose.
• Benchmarked these methods against the traditional frame-specific approach.
• Evaluated methods based on their ability to match simulation models to experimental data.

Main Results:

• The trial-specific method demonstrated superior accuracy in calculating frame-specific link lengths.
• Accurate link length determination significantly impacts the fidelity of recreated postures in simulations.
• The choice of link length calculation method directly affects simulation-to-experiment correlation.

Conclusions:

• The trial-specific method is recommended for calculating frame-specific link lengths in motion capture-based digital human simulations.
• Minimizing link length errors is critical for reliable simulation validation.
• This study provides a framework for improving the accuracy of digital human models.