A high-level semantics extraction model for stored videos

Digital video databases have become more pervasive, and finding video clips in video databases is quickly becoming a major bottleneck. Indexing and annotation can help video systems retrieve appropriate video clips fast. Currently, these techniques are based on computation on image low-level features and manual annotation, whose low success ratio and retrieval speed limit the applications of video databases. In this paper, we consider the semantic interpretation of the the contents as a form of annotation for video clips. We give a self-adaptive high-level description model for application-oriented semantics extraction, which annotates and retrieves video clips flexibly and automatically