Research On Content-Based Video Shot Segmentation And Retrieval Technology

With the rapid development of multimedia technology, a large number of video data appears. It becomes more and more necessary to retrieve the required video from mass video data. Against this background Content-Based Video Retrieval (CBVR) technology has aroused widely concern. Content-Based Video Retrieval technology uses computer to automatically analyze and organize the content of video. People can submit sample data or description to find the required video data.This paper studies the key technology of Content-Based Video Retrieval, including shot segmentation, key frame selection and video retrieval. The main contents are as follows:1) Common shot converting forms and shot segmentation algorithms are discussed. An adaptive sliding window double threshold shot boundary algorithm is proposed using the color-texture two dimensional union distribution histogram to calculate the interframe difference. This algorithm is sample and uses both partial and global information of the frame data. Finally the effectiveness of this algorithm is verified through comparing with other algorithms.2) The principle of key frame selection is introduced and some common key frame selection algorithms are discussed. This paper improves a fast and effective key frame selection algorithm to adapt the high activity shots and it is very flexible in deciding the key frame numbers.3) A sample and fast weight adjustment algorithm is proposed based on user feedback. A retrieve interface accepting both image and video segments as input data is designed. User feedback is introduced to return batter results.