Recognizing Strokes in Tennis Videos Using Hidden Markov Models

This paper addresses content-based video retrieval with an emphasis on recognizing events in tennis game videos. In particular, we aim at recognizing different classes of tennis strokes using automatic learning capability of Hidden Markov Models. Driven by our domain knowledge, a robust player segmentation algorithm is developed for real video data. Further, we introduce a number of novel features to be extracted for our particular application. Different feature combinations are investigated in order to find the optimal one. The experimental results demonstrate that our method is close to realizing statistics of tennis games automatically using ordinary TV broadcast videos.