Detection of Tennis Court Lines for Sport Video Categorization

Digital video data are stored and offered by many public Web services such as Internet video collections, TV shows archives, Internet video-on-demand systems, personal video archives, and others. New methods and technologies of video indexing and retrieval in the Web are developed. Content-based indexing of TV sports news is based on the automatic segmentation, then recognition and classification of scenes reporting the sport events in a given discipline. Automatic classification of sports in TV sports news is one of the basic process in video indexing. There are many different strategies how to recognize a sport discipline. It may be achieved by player scenes analyses leading to the detection of playing fields, of superimposed text like player or team names, recognition of player faces or sport objects, detection of player and audience emotions, and also to the detection of lines typical for a given playing fields and for a given sport discipline. The paper proposes a framework of the automatic line detection of a tennis court for the selection of tennis shots from TV sports news. Categorization of sport videos is based on the minimum set of lines for the detection of a tennis court. The framework has been verified and tested in the Automatic Video Indexer AVI.