Based Relevance Feedback Image Retrieval Techniques And Realization

Along with the development of computer and communication technology, the multimedia appears to the trend of information increasing rapidly. Comparing with the traditional _text information, Image has the character of large information and it's difficult to describe exactly. In this way, one side, The image information are increasing at exponent level, one other side, people have difficult to find what they need from vast Images Information, accordingly, it forms the phase that looks for a needle in a bottle of hay. So, how to manage efficiently and retrieve the need of image is the problem crying for settling for people in Information Age.Content-based image retrieval(CBIR) and relevance feedback technology,is the one of top research domains, which is the solution of image retrieval problem, but Retrieval precision and retrieval by semantic is yet the problem requiring to research, they are something required to research that image content description for content retrieval and data model, high dimension image feature vector 'calculation, the structure and index of image database, etc.Contraposing to the actuality and developing trend of content-based image retrieval research, .we lucubrate on CBIR, relevance feedback arithmetic and semantics of image extract, simultaneously research associated problems. Via analysis of the hiberarchy of image retrieval and the essential conception of relevance feedback, image data model and model of content-based & relevance feedback image retrieval were, put forward;we design the content-based retrieval arithmetic and relevance feedback arithmetic, including the arithmetic of feedback and sematics extract & retrieval;we have made a system: FiShip which realize content-based and relevance feedback image retrieval.