The purpose of this study was to evaluate markerless, silhouette-based tracking for different applications in sports science. Data of segment center of gravity locations, joint center locations as well as joint angles were taken into account. To quantify the accuracy of silhouette-based in comparison to marker-based tracking, all mentioned parameters were compared with the correlation coefficient and standard deviation of the differences for three classes of movements: specific joint movements, complex movements, and highly dynamic movements (with racquet). Very strong correlations result for the segment center of gravity locations, the joint center locations as well as for joint angles in the sagittal plane except the elbow joints. Joint angle accuracy impairs in the transversal and the frontal plane with an increasing complexity and speed in the movement patterns. To obtain accurate joint angles separated into the three body planes, however, we recommend to enhance the tracking of segment rotations should be stabilized by additional information (e.g. marker or IMU).
[1]
A. Cappozzo,et al.
Human movement analysis using stereophotogrammetry. Part 3. Soft tissue artifact assessment and compensation.
,
2005,
Gait & posture.
[2]
Thomas P Andriacchi,et al.
The evolution of methods for the capture of human movement leading to markerless motion capture for biomechanical applications
,
2006,
Journal of NeuroEngineering and Rehabilitation.
[3]
Daniel Arthur James.
The Application of Inertial Sensors in Elite Sports Monitoring
,
2006
.
[4]
Emiliano Gambaretto,et al.
Markerless Motion Capture through Visual Hull, Articulated ICP and Subject Specific Model Generation
,
2010,
International Journal of Computer Vision.