An Efficient Scene Detection Using Rough Set-Based Fuzzy Clustering for Film Video

It is important to organize the unstructured video data properly for content-based video analysis and retrieval. In this paper, an efficient film video scene detection method is presented by using a rough set-based shot fuzzy clustering algorithm for film. Equivalence relation theory is applied to group video shots into clusters. Initial shot clustering is performed directly by judging whether equivalence relations are equal, not computing the intersection of equivalence classes as usual. And excessive generation of some small classes is suppressed by secondary clustering on the basis of defining fuzzy similarity between two initial clusters. Afterwards, according to the characterization of film video scene, three types of temporal relationships between two shot clusters are introduced. We introduce a temporally and spatially integrated strategy for parsing shot clusters into semantic scenes. The scheme offers an efficient mean for browsing and effectively retrieving film video