Out-of-View Player Location Estimation Using In-Venue Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional player tracking techniques are limited to in-venue tracking and fail to accurately estimate the locations of players not visible in video broadcasts of sporting events due to varying camera angles and focus on main action, which affects statistical analysis.
Innovation Solution
A computing system uses a prediction model, trained on in-venue tracking data and broadcast video data, to estimate the locations of players not visible in broadcast feeds through spline models or nearest neighbor approaches, interpolating or predicting based on known locations at the start and end of possessions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If camera zooms into main action area, then visibility of players near action is improved, but visibility of players at strategic locations is lost
Solution Approach 1:
The system performs preliminary tracking of all players using in-venue cameras before the broadcast camera zooms in. This preliminary action captures the initial positions of all players, including those who will later go out of view. When players become invisible during broadcast, the system refers back to these pre-captured positions and uses prediction models to estimate their current locations, thus preventing information loss.
Solution Approach 2:
The system introduces an intermediary prediction model that acts as a mediator between the visible player positions and the invisible player positions. This model uses the known positions of visible players and game context to infer and estimate the positions of players who are currently out of camera view, effectively bridging the information gap created by camera zooming.
2Measurement precision
If broadcast camera focuses on main action, then detail of main action is improved, but complete player tracking is deteriorated
Solution Approach 1:
The system merges two different tracking approaches: in-venue tracking (which provides complete player visibility) and broadcast video analysis (which provides detailed action views). By combining these two data sources and using the prediction model to reconcile them, the system maintains both the detailed action visibility from broadcast cameras and the complete player location accuracy from in-venue tracking.
Solution Approach 2:
The system implements a feedback mechanism where the prediction model continuously refines player position estimates based on new visible player positions and game events. When players reappear in the broadcast view, their actual positions provide feedback to correct and improve the accuracy of previously estimated positions, ensuring ongoing reliability of player location data.
3Measurement precision
If in-venue tracking is used, then complete player visibility is achieved, but applicability to broadcast analysis is limited
Solution Approach 1:
The system creates a virtual copy of the in-venue tracking capability that works specifically for broadcast analysis. Instead of requiring physical in-venue cameras, the prediction model replicates the functionality of in-venue tracking by estimating player positions from broadcast video data alone, making the system adaptable to broadcast analysis without needing access to venue infrastructure.
Data Source
AI summary
Examples disclosed herein may estimate locations of players not visible in a sporting broadcast video. A prediction model may be generated based on a training data set of in-venue tracking data that includes locations of all players at all times and the corresponding broadcast tracking data that may not necessarily contain the locations of all players at all times. The prediction model may be based on an algorithmic logic (e.g., a spline regression) or machine learning model (e.g., k-nearest neighbor, deep neural network). The generated predicted model may be used to estimate the unknown locations of players in broadcast tracking based on the known locations.