Tracking Players in Sports Games
Team sports (Handball, Basketball)



In presented work, we focus on tracking the players in team sports. Advances in computer and video technology over the past several years enabled us to approach the problem by digitizing the whole handball match, which lasts for an hour at 25 frames per second, using two wide-angle cameras. Acquired data was subsequently used to develop a set of tracking methods, which suit best the nature of the game. The whole match is videotaped, and the video is digitized to M-JPEG format using inexpensive consumer-grade video acquisition hardware. Player tracking and motion analysis are done offline, with the help of human operator, who supervises the tracking process. Some experimental results are presented below. The developed system is used in handball and basketball match analysis.

Handball video demonstration (input data, processing and results - 21 MB, 60 sec, MPEG)  

User interface

The system for tracking players in team sports has been developed to the phase of working prototype, which is routinely used at Faculty of Sports at University of Ljubljana for team sport analysis. To enable routine use, an user interface, which is shown below, was developed. The application runs under the Microsoft Windows operating system.


The results of the tracking algorithm are spatio-temporal trajectories for each of the players. Therefore, the system performs position and velocity measurements, and its accuracy has to be known. Thorough testing of the system resulted in the following accuracy specifications:


Results, provided by our system can be exported in the form of tab-separated text files, which ensures easy integration with other applications, which can be used for further analysis. Alternatively, results can be presented in illustrative and visually attractive manner, for easier analysis by sport experts. Some examples are shown below and represent motion data for the first 10 minutes of the handball match for the 7 players of the winning team. Additionally, our system is able to process the original video data, remove lens distortions and superimpose tracking results on the original video. "Virtual camera" can follow selected players as they move across the court, as illustrated in the last image.



Janez Perš or Stanislav Kovačič