Video scenes clustering based on representative shots

Clustering of video data is an important issue in video abstraction, browsing and retrieval. Movie is a kind of complex video with rich content. As for the movie, however, clustering is more complicated than other types of videos like surveillance, sport games, and documentaries. In this paper, we propose a novel shot clustering algorithm combing the editing feature of the movie and the criteria of choosing representative shots. The simulated experiment results demonstrate the effectiveness of our method for the dialogue-dominated movie.