What Players do with the Ball: A Physically Constrained Interaction Modeling
Tracking the ball is critical for video-based analysis of team sports. However, it is difficult, especially in lowresolution images, due to the small size of the ball, its speed that creates motion blur, and its often being occluded by players. We propose a generic and principled approach to modeling the interaction between the ball and the players while also imposing appropriate physical constraints on the ball’s trajectory. We show that our approach, formulated in terms of a Mixed Integer Program, is more robust and more accurate than several state-of-the-art approaches on real-life volleyball, basketball, and soccer sequences.
Results
We show our ball tracking results on volleyball, basketball and soccer sequences as follows.
Volley1 dataset
Embed of video is only possible from Mediaspace, SwitchTube, Vimeo or Youtube
Volley2 dataset
Embed of video is only possible from Mediaspace, SwitchTube, Vimeo or Youtube
Basket1 dataset
Embed of video is only possible from Mediaspace, SwitchTube, Vimeo or Youtube
Basket2 dataset
Embed of video is only possible from Mediaspace, SwitchTube, Vimeo or Youtube
APIDIS dataset
Embed of video is only possible from Mediaspace, SwitchTube, Vimeo or Youtube
ISSIA dataset
Embed of video is only possible from Mediaspace, SwitchTube, Vimeo or Youtube
Play the Basketball Roulette!
Try your luck at guessing where the ball is here and see if you can beat the algorithm!