Algorithm Design And Research For The Texture Redundancy-based Video Coding

Video compression has highly developed during past twenty years, for the application of the transmission and processing of video data in almost every field of industry.However, when people are paying more and more attention to family entertainment, not only how to increase the compression quality, but the interaction between person and multimedia data has been a key problem.A new idea about video coding is proposed in this paper. It can not only increases the compression ratio, but also quicken the retrieval of multimedia data.For I-frame coding, existing coding standards only take use of information of itself, and for P-frame and B-frame, inter-frame information is used. However, there is still redundancy among those adjacent frames that are coded as I-frame.The algorithm consists of two important parts: after dividing the image into blocks, the first step is extracting the blocks' texture feature, and the next step is classifying the features sets into clusters. The compressing is implemented by using the center block. The center index block holds important information for the sense of human's eyes; moreover, it is also the important texture information among I-frames. Therefore, it will play a key role to retrieval too.By comparison among the results when using a lot of algorithms in feature extraction, we choose Gabor filter for its good quality in low frequency.And we choose simple Cluster algorithm for the classification. Because we extract 48-dimension feature, simple classification algorithm can meet the system's request.After test of image sequences, the algorithm shows good performance.