What Players do with the Ball: A Physically Constrained Interaction Modeling

TL;DR

This paper introduces a robust, physics-constrained model for tracking balls in team sports videos, improving accuracy over existing methods especially in challenging low-resolution scenarios.

Contribution

It presents a novel mixed integer programming approach that models ball-player interactions with physical constraints, enhancing tracking robustness and accuracy.

Findings

Outperforms state-of-the-art methods on volleyball, basketball, and soccer datasets.

Provides a generic framework applicable to various team sports.

Demonstrates improved tracking accuracy in low-resolution and occluded conditions.

Abstract

Tracking the ball is critical for video-based analysis of team sports. However, it is difficult, especially in low-resolution images, due to the small size of the ball, its speed that creates motion blur, and its often being occluded by players. In this paper, 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.