Player tracking in team sports presents significant challenges due to frequent occlusions and the difficulty of re-identifying athletes who wear identical uniforms. A promising solution to mitigate these issues involves the use of top-view cameras, which offer a vantage point where occlusions are minimized. However, in indoor environments such as volleyball courts, structural height limitations and the need to cover a wide playing area often necessitate the use of multiple top-view cameras. In this work, we address the problem of tracking volleyball players with three multi-optics cameras from the top view. Our approach involves four main stages: (1) camera calibration and geometric adjustment, (2) motion-based player detection and evaluation, (3) multi-player tracking, and (4) ball tracking. Experimental results show that the proposed pipeline improves robustness against occlusions and maintains reliable tracking performance across multiple players. These findings demonstrate the feasibility of top-view multi-camera setups for sports analytics and highlight their potential to enhance both player performance assessment and tactical analysis in volleyball and related team sports.
Tracking Volleyball Players from a Multi-Camera Top-View Setup / Bisagno, N., Martinelli, G., Conci, N.. - (2025), pp. 60-65. (2025 IEEE International Workshop on Sport, Technology and Research, STAR 2025 ita 2025) [10.1109/STAR66750.2025.11264783].
Tracking Volleyball Players from a Multi-Camera Top-View Setup
Bisagno N.;Martinelli G.;Conci N.
2025-01-01
Abstract
Player tracking in team sports presents significant challenges due to frequent occlusions and the difficulty of re-identifying athletes who wear identical uniforms. A promising solution to mitigate these issues involves the use of top-view cameras, which offer a vantage point where occlusions are minimized. However, in indoor environments such as volleyball courts, structural height limitations and the need to cover a wide playing area often necessitate the use of multiple top-view cameras. In this work, we address the problem of tracking volleyball players with three multi-optics cameras from the top view. Our approach involves four main stages: (1) camera calibration and geometric adjustment, (2) motion-based player detection and evaluation, (3) multi-player tracking, and (4) ball tracking. Experimental results show that the proposed pipeline improves robustness against occlusions and maintains reliable tracking performance across multiple players. These findings demonstrate the feasibility of top-view multi-camera setups for sports analytics and highlight their potential to enhance both player performance assessment and tactical analysis in volleyball and related team sports.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione



