Determination of rigid body registration marker error from edge error

Area of Science:

• Biomechanics
• Motion Analysis
• Robotics

Background:

• Accurate 3D rigid body motion tracking relies on registration markers, but their position measurements contain errors.
• Previous error propagation studies often assumed marker variation or used static measurements, limiting applicability to dynamic motion.
• Understanding individual marker variation is crucial for theoretical and simulation-based error propagation in motion analysis.

Purpose of the Study:

• To introduce a novel method for determining individual registration marker variation.
• To enable accurate variation assessment irrespective of rigid body position or motion.
• To improve the reliability of error propagation models in 3D motion tracking.

Main Methods:

• Developed a technique utilizing invariant inter-marker distances (edge lengths) to assess marker variation.
• Employed simulations to validate and characterize the new method's performance.
• Applied the technique to experimental motion tracking data for comparison with static measurements.

Main Results:

• The method's predictive accuracy improves with increased edge length and marker count.
• Predictions converge to reference values when edge length is at least four times the maximum vertex variation.
• Under ideal conditions, the confidence interval for predicted variation is within 7% of the maximum variation.

Conclusions:

• The novel method accurately determines individual marker variation during dynamic motion.
• Non-static marker variation significantly exceeds static variation, highlighting the need for dynamic measurement.
• This technique enhances the accuracy of error propagation analysis in 3D rigid body motion tracking.