EVENT DETECTION BASED ON NON-BROADCAST SPORTS VIDEO (TA-P6)

Author(s) : Jinjun Wang (Institute for Infocomm Research, Singapore) Changsheng Xu (Institute for Infocomm Research, Singapore) Engsiong Chn (Nanyang Technological University, Singapore) Xinguo Yu (Institute for Infocomm Research, Singapore) Qi Tian (Institute for Infocomm Research, Singapore)

Abstract : Most of current research work on sports video analysis is focusing on the broadcast video. The broadcast video is a post-produced video and has much additional editing information inserted. In this paper, we propose a novel event detection framework in sports game only based on the video taken by the single (main) camera. Compared with event detection from broadcast video, the proposed framework is more challenging. The detected events can be used for automatic replay generation in sports video. A mid-level representation is first created based on low-level audio/visual features. Events related to the replay are then detected from this representation. We experiment our framework and some promising results are obtained.

Menu