基于深度学习的艺术体操动作精准评分与技术分析系统
By combining multi-view video capture and differential geometry-enhanced human pose reconstruction technology with deep learning, the problems of occlusion, accuracy, and dynamic analysis in rhythmic gymnastics scoring were solved, achieving high-precision motion capture and objective scoring, and improving scoring consistency and training efficiency.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- SHENZHEN UNIV
- Filing Date
- 2025-12-06
- Publication Date
- 2026-07-17
AI Technical Summary
Existing technologies for rhythmic gymnastics scoring suffer from single-view occlusion issues, insufficient accuracy, inadequate dynamic analysis, and a lack of objective quantification in scoring standards, leading to inconsistent scoring and insufficient training guidance.
By employing multi-view video acquisition and differential geometry-enhanced human pose reconstruction technology, combined with deep learning, high-precision motion capture, analysis, and scoring are achieved through multi-view keypoint manifold mapping, Riemannian geometry optimization, and spatiotemporal differential structure.
It significantly improves the accuracy and robustness of human posture reconstruction, enhances dynamic tracking accuracy, and significantly improves the objectivity and accuracy of the scoring system. The scoring results are 91.7% consistent with those of professional referees, providing athletes with timely feedback and effective training guidance.
Smart Images
Figure CN121600598B_ABST