A Robust Multiple Object Tracking for Sport Applications
This paper deals with the task of tracking several similar looking athletes during a competition, using only a single camera. Due to the wide range of possible motions, non-rigid shape changes and mutual occlusions the tracking situations can become quite complex. We propose a novel method to use a high dimensional motion model for a particle filter without drastic increase in runtime. Integral histograms are extended to handle also rotated objects. In order to deal effectively with overlapping players we propose a re-weighting and re-sampling method for a particle filter. The paper is primarily designed for beach volleyball but is also evaluated on public available football sequences.
- Mauthner T. and Bischof, H. (2007): A Robust Multiple Object Tracking for Sport Applications, Proceedings Annual Workshop of the Austrian Association for Pattern Recognition
Example videos show tracking results on multi-object sport scenes.
|Download||Results of the proposed global re-sampling scheme during mutual occlusions of baech volleyball players.||Download||Results of the proposed global re-sampling scheme during mutual occlusions of soccer players.|
|Download||Handling of occlusions during tracking, showing the effect of using an background model to avoid particle drifting to the surrounding.||Download||Tracking of athletes during competition, including jumps and rotation.|