Player detection and estimation of the number of players is the fundamental task of any sports analysis and monitoring system. It leads to the mapping of player activities, as well as tracking of their movements in sports, is the essential component of such analysis. The analysis leads to consistent and standard monitoring of player performance. It also leads to the investigation of the intensity of game, technical gestures and physiological demands of players, which is vital information for the coaches and trainers. It also helps analyzers in predicting several game movements. In outdoor local fields of sports, these tasks become very challenging due to varying weather and illumination conditions.

The main research areas of this project are to analyse different camera setups for sports analysis and player detection and counting. It is critical for a system to work independently of the size of the court, weather and lighting conditions. In most of the systems, the system is very complex to be employed in local sports fields.

This project focuses on robust counting of players and estimating their positions by testing different camera setups while keeping the installation simple and cost-effective. Four camera setups are defined to perform tests on sports video for variable outdoor phenomena. Our project aims at utilizing these camera setups for robustly detecting and monitoring players in the outdoor sports arenas.
Effektiv start/slut dato01/04/201731/03/2020