Abstract: Video segmentation and keyframe extraction are the core issues of Content based Video Retrieval(CBVR).We proposed an improved approach of key-frame extraction for video retrieval, adopting the synthesis method based on histogram-based method and pixel-based method in video shot segmentation. In the method, videos arc firstly segmented into shots according to video content, by our improved histogram-based method, with the use of histogram intersection and nonuniform partitioning and weighting. hhen, the obtained results arc secondly detected by pixel-based method to optimize the results. At last, within each shot, key-frames arc determined with the calculation of image entropy of every frame in HSV color space. Our simulation results prove that the extracted key frames with our method are compact and uniformly distributed.

Key words: Video shot detection, Keyframe extraction, Pixel-based method, Image entropy, Video retrieval