A Low-Cost Computer Vision System for Real-Time Tennis Analysis

摘要: This paper describes a low-cost vision-based system for real-time tennis game analysis. The elaborates videos captured by four synchronized and calibrated cameras installed at the sides of court in order to accurately localize ball players, track them real-time. From this low-level data mid-level events, like shots, bounces, net, high-level stroke type line calling, are detected. All is made available players both on-court during play or through web device end session. Currently, prototypes undergoing field test three locations Italy. In addition positive comments users, robustness reliability have been demonstrated with specific evaluation tests. Detection rate shots 99.7% while miss detection less than 0.8%. Reliability classification 97.1% in/out 99.5%. On average reaction time calling 152 ms.