Application Of Approximate String Matching In Video Retrieval

With the coming of information age and the rapid development of multimedia technology, multimedia has become an important carrier of information data. As video information comes forth largely, how to search the needed information quickly in a mass of data has attracted the attention of many researchers, the rapid analysis based on video content, video classification and retrieval has become a hot topic in the field.Content-based Video Retrieval is one of the leading methods to get the needed information from the vast amount of video data, which mainly contains shot segmentation, key frame extraction and matching retrieval. Key frame extraction can be carried out in the compressed domain environment, and it also can be carried out in the uncompressed domain environment. In the compressed domain environment, random algorithm is used in the process of key frame extraction, which can improve the operational efficiency of the algorithm, and the experimental results proved to be better.In this paper, how to get the video through the video clip is studied carefully. Based on the research of content-based video retrieval, firstly to extract features from key-frame sequence, secondly to get the characteristic curve from the features, thirdly to get the feature string by characteristic curve, and the last is approximate string matching. The main research work is as follows:a) Feature extraction of key frames. To make the image feature less sensitive for the moving objects in the image, the method to divide the image into n * n sub-blocks is used, and to extract color, texture features of each sub-block as the basic features; For the area closer to the general center of the image, it has a greater impact on eyes, the method using sub-block feature weighting is used based on the visual impact of each sub-block to human eyes; In the calculation of characteristics distance between key frames, the method of sub-block features weighted summation is put into practice.b) Characters of the key-frame sequence features. Make the features of key frames as key points, using the Bessel function to fit the key points to get the characteristic curve of key-frame sequence. Through marking the key points of the characteristic curve, the feature string expressing the formation of characteristic curve can be received.c) Semantic information of feature string. Key-frame feature string contains some semantic information of the video, such as through the feature string to find whether there is abrupt shot, gradient shot, characteristic cyclical changes or other information, and all the information can be used in further video classification.