A Hidden Markov Model Approach to Parsing MTV Video Shot

In this paper, we present an approach for detecting MTV video shot using Hidden Markov Models (HMMs), in which the color, shape and motion features are utilized. First, the temporal characteristics of different shot transitions are exploited and an HMM is constructed for shot transitions, including cut and gradual transitions. Secondly, a trained HMM are used to recognize the shot transition automatically, it does not suffer from any trouble of threshold selection problem. Experimental results on a set of test MTV videos demonstrate that our approach is validated in the particular domain of MTV videos.