The Research Of Video Retrieval Method Based On Shot Discrimination Analysis

With the development of the Internet, multimedia resources are flooding in daily life. Therefore, it is of great importance to describe, store, organize and retrieve multimedia information, which can assist people in finding the resources quickly and conveniently. In recent years, Content-Based Video Retrieval (CBVR) has become a promising field which motivates the interests of researchers worldwide.At first, this paper briefly introduces CBVR technology including its originality, definition, characteristics, research history, current research situation, classification and application fields, etc. Then based on theory and practice, the sub fields including feature extraction, shot boundary detection, key frame representation and similarity matching retrieval step are elaborated. Video shot has different representativeness; some shots can represent the content of videos effectively while others cannot. Traditional method usually treats shots equally, in this way fault retrieval is unavoidable. In order to overcome the disadvantages of this method and ensure the retrieval accuracy and efficiency, this paper proposes the concept of shot discrimination and the calculation method based on shot representativeness. A new video retrieval method is designed based on shot discrimination analysis and a simulation system is accomplished accordingly. Some comparison experiments can prove how accurate and efficient the retrieval performance is.