Real-time Sports Motion Capture via Multi-Camera Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current sport event object monitoring and motion capture systems are unable to provide real-time, non-intrusive tracking and identification of athletes and objects, limiting their application in live sports events and computer games.
Innovation Solution
A system utilizing multiple TV cameras with image processing units for real-time object segmentation, blob analysis, and 3D localization, combined with robotic cameras for manual or automatic identification of players and objects, enabling real-time motion capture and transfer of data for 3D graphical representation in computer games.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If multiple TV cameras with image processing units are deployed for real-time tracking, then real-time motion capture capability is improved, but device complexity increases
Solution Approach 1:
The system divides the tracking task into separate modules: multiple TV cameras capture video feeds, image processing units perform object segmentation and blob analysis, and a central server handles 3D localization and tracking. This segmentation allows real-time processing while distributing system complexity across multiple specialized components rather than requiring a single complex system.
2Ease of operation
If peripheral equipment is used for non-intrusive tracking, then ease of operation is improved, but measurement precision may worsen
Solution Approach 1:
The system merges data from multiple TV cameras viewing the same scene from different angles. By combining these multiple perspectives, the system achieves accurate 3D localization and tracking of objects without requiring intrusive sensors on the objects themselves. The central server integrates information from all cameras to precisely determine object positions, velocities, and trajectories.
3Extent of automation
If robotic cameras are used for automatic identification, then extent of automation is improved, but device complexity increases
Solution Approach 1:
The system implements self-service automatic identification where the image processing units automatically perform object recognition, player identification, and tracking without human intervention. The central server automatically correlates objects across multiple camera views and maintains persistent identification. This automation eliminates the need for manual operation while the modular architecture keeps individual components relatively simple.
Data Source
Figure 1
Figure 2a
Figure 2b
AI summary
Non-intrusive peripheral systems and methods to track, identify various acting entities and capture the full motion of these entities in a sports event. The entities preferably include players belonging to teams. The motion capture of more than one player is implemented in real-time with image processing methods. Captured player body organ or joints location data can be used to generate a three-dimensional display of the real sporting event using computer games graphics.