Stanford researchers have patented a unique approach for accurate and precise unencumbered 3D movement analysis for full motion 3D animation.
The system architecture is modular and includes:
1. Multiple synchronized cameras acquisition
2. Foreground/Background segmentation
3. 3D representation reconstruction
4. Tracking of the representation through a model
5. Kinematics extraction and outputting
This invention works with the model generation methods and software described in Stanford docket
S07-254, S07-086, and US Patent 8,180,714.This markerless motion capture system is accurate enough for biomechanical, clinical, sports, animation, video gaming, design, ergonomics, and surveillance applications.
Video of Markerless Motion CaptureRelated Markerless Motion Capture Technologies also available for license: S06-193– US Patent 8,139,067 & S06-193B -software associated with Stanford docket S06-193
S07-254– US Patent 8,180,714 & S07-086 -software associated with Stanford docket S07-254
S08-122– US Patent 8,384,714