Automatic video analysis of countermovement jump performance using a single uncalibrated camera

Area of Science:

• Biomechanics
• Sports Science
• Computer Vision

Background:

• Countermovement jump (CMJ) assessments are crucial for sports performance monitoring.
• Traditional methods using force plates are expensive and cumbersome.
• Existing IMU and mobile app solutions require technical expertise and manual input.

Purpose of the Study:

• To develop an automated, camera-based pipeline for CMJ performance analysis.
• To utilize computer vision and biomechanical principles for accurate data extraction.
• To offer a cost-effective and user-friendly alternative to current CMJ assessment tools.

Main Methods:

• Employed human segmentation and pose estimation from single-camera video.
• Applied biomechanical principles (parabolic motion, inverse dynamics) to estimate force-time data.
• Validated the system against a commercial force platform with 77 elite athletes.

Main Results:

• Achieved superior correlations (R > 0.7) and high reliability (%CV < 10%) for most CMJ variables.
• Demonstrated fully automatic analysis of force-time data and performance variables.
• Outperformed IMU-based approaches in accuracy and reliability.

Conclusions:

• The developed camera-based method offers automated, accurate, and reliable CMJ analysis.
• It reduces technical demands and provides more variables than mobile apps.
• Ideal for large-scale studies, enhancing sports performance and health monitoring.