Synchronised Video, Motion Capture and Force Plate Dataset for Validating Markerless Human Movement Analysis.

Area of Science:

• Biomechanics
• Computer Vision
• Human Motion Analysis

Background:

• Marker-based optical motion capture is the gold standard for human motion analysis.
• Markerless motion capture using computer vision offers a less intrusive alternative.
• Validation of markerless systems requires comprehensive, synchronized datasets.

Purpose of the Study:

• To introduce the BioCV dataset, a novel resource for markerless motion capture system development.
• To provide synchronized multi-modal data for validating computer vision-based motion analysis.
• To enable performance comparison between markerless and marker-based motion capture techniques.

Main Methods:

• Collected synchronized data from 15 healthy participants (7 males, 8 females).
• Utilized multi-camera video, marker-based optical motion capture, and force plates.
• Recorded controlled movements including walking, running, jumping, and hopping.
• Acquired photogrammetry scan data for each participant.

Main Results:

• The BioCV dataset provides a unique combination of synchronized motion capture modalities.
• The dataset enables direct comparison of markerless computer vision techniques against marker-based systems.
• Includes diverse, repeated human motion data for robust system evaluation.

Conclusions:

• The BioCV dataset is a valuable resource for advancing markerless motion capture technology.
• Facilitates the development and validation of computer vision algorithms for human movement analysis.
• Supports research aiming to improve the accuracy and reliability of markerless motion tracking.