Classification of summarized videos using hidden markov models on compressed chromaticity signatures

Tools for efficiently summarizing and classifying video sequences are indispensable to assist in the synthesis and analysis of digital video. In this paper, we present a method for effective classification of different types of videos that uses the output of a concise video summarization technique that forms a list of keyframes. The summarization is produced by a method recently presented, in which we generate a universal basis on which to project a video frame feature that effectively reduces any video to the same lighting conditions. Each frame is represented by a compressed chromaticity signature. A multi-stage hierarchical clustering method efficiently summarizes any video. Here, we classify TV programs using a trained hidden Markov model, using the keyframe plus temporal features generated in the summaries.