Research On Content Based Image Retrieval Technology

With the quick development of the multimedia and network technology, digital images increase rapidly all over the world. How to manage and retrieve images quickly becomes more and more important. Content-based image retrieval (CBIR) technology which uses image visual content to query image information becomes most active one in multimedia retrieval field. CBIR technology extracts image information through analyzing color, texture, shape and other low-layer image feature from image, describes image feature information, matches images and then enables effective, efficient query and access image source.In this paper, the background, significance of CBIR research and status in quo are firstly introduced; and then the key techniques of Content-Based Image Retrieval System (CBIRS) are introduced. Combining with the technologies of image processing, pattern recognition, computer vision and database, this dissertation studies the key issues of CBIR. Some significant conclusions are drawn from this study. The major contents include:The extraction of color, texture and shape descriptor is reported detailedly. These descriptors are dominant color, edge histogram and region shape which are commended by MPEG-7. A method using combined color, texture and shape feature is presented. In the plenty of experimental results compared with which using single feature or two features among them, the experiments give good results. So the integration of multiple image features can describe image content generally and exactly.According to the Discrete Cosine Transform (DCT) of image and its transform coefficients feature, an algorithm based on DCT coefficients is proposed.Study the JPEG image's coding/decoding process and its coefficients on compressed domain, and then develop image retrieval system based on compressed domain color feature and also texture feature.CBIR based on combined color feature and texture feature on compressed domain is proposed. A practical and excellent content based image retrieval system is designed with VC++ and database technology.