This paper proposes an online segmentation scheme for broadcasted American football TV programs. The TV video streams are online partitioned into a series of play-break patterns with a relatively short delay. All the detected plays form an event-based summarization which contains complete information about the original match. The main contributions of the proposed scheme are twofold. First, an online adaptive field-color model is employed to deal with feature variations which may caused by court configurations and time-varying lighting conditions. Second, a temporal pattern based model is constructed to capture intrinsic features of football plays, which considerably increased the performance of the proposed scheme. Effectiveness of our approach is demonstrated by experiments on real recorded TV programs.
展开▼
