Tracking based Event Detection of Singles Broadcast Tennis Video

For the past few decades automated assistance in sports is one of the active researches in computer technology. In tennis, owing to the increase in the number of videos on the internet and broadcast, there is a need arises to access specific events instead of accessing a full-length video. Skipping the less interesting parts of tennis videos will develop an interest not only for the viewers, but also to the tennis coach who analyses the match, and the cost of videos were reduced too. To attract the users there is a need for an active research on object detection and tracking followed by event recognition arises. In Broadcast Tennis Video (BTV), the main attention is on the moving object as ball is stroke by the players continuously. Based on the motion of player the events will occur. In this paper, especially the research is done to track the moving objects such as ball and player followed by the tracking of moving objects the events are detected and classified using HMM.