Evaluating markerless biomechanical analysis in a real-world pole vault competition setting

Area of Science:

• Biomechanics
• Sports Science
• Motion Capture Technology

Background:

• Marker-based motion capture is the standard for elite pole vault biomechanical analysis.
• Practical constraints in live competition limit the application of traditional methods.
• Markerless systems offer a potential alternative for real-world sports analysis.

Purpose of the Study:

• To evaluate a purpose-trained markerless motion capture system for elite pole vault biomechanics.
• To compare the markerless system's results against manual digitization under practical conditions.
• To assess the system's capability with a fixed four-camera setup.

Main Methods:

• Collected data from eight elite pole vaulters during competition using a four-camera setup.
• Recorded the final run-up and take-off phases at 100 Hz.
• Analyzed data using both manual digitization and the SIMI Nemo Markerless system.

Main Results:

• Strong agreement between markerless and manual methods for spatial and center of mass (CM) variables (mean bias 0.3%, random error 3.8%).
• High accuracy for CM height and velocity at pole plant and take-off (RMSE ~0.1 m/s).
• Excellent waveform agreement (Coefficient of Multiple Determination >0.90), though joint angles showed less precision (RMSE 5°-10°).

Conclusions:

• A four-camera markerless system can capture essential elite pole vault performance indicators with high accuracy.
• The system demonstrates strong potential for scientific and applied use in real-world sports environments.
• Further refinement is needed for precise joint angle analysis in complex movements.