Video Key Frame Extraction Algorithm Based on Possibilistic C-Patterns Clustering

A new method is proposed to extract video key frames, based on a possibilistic C-patterns (PCP) clustering algorithm. By the method, the dominant colors and the subordinate colors are used to represent the features of video frames. Consequently, the key frames can be extracted directly without shot partition. The algorithm adapts well to the visual contents. Its efficiency and effectiveness are demonstrated with the experimental results on various kinds of videos.