Research Of Image Retrieval Technology Based On Pattern Image Production

With the development of computer science and technology, the application of Internet and the rapid increment of multimedia database, we have a lot of digital images and video information. In order to manage and retrieve the information, the CBIR (Content-Based Image Retrieval) has emerged to be one of the hot research areas in image domain.Content-Based Image Retrieval technology, by the machine automatically extracts the visual features of images contents color, texture, shape and mutual relations. And the retrieval process is the images in the database and sample queries image match in the similar features space, to search similar images. This method uses automatic image feature extraction and retrieval methods, thereby greatly increasing the retrieval efficiency.Because there is existed serious difference between low-level image features and human understanding to image, the former cannot describe image content exactly. That is to say, there is a "semantic gap" between low-level features and image semantics. This leads to the semantic image retrieval and classification. The approach combines the semantic information of images with visual features, among at retrieving or classifying images. The Semantic image retrieval has recently attracted many researchers and become an interesting issue in multimedia information retrieval. It is difficult to extract semantics, represent and apply it for the complexity of image semantics. Thus there is a challenging problem.This paper proposes a new method to solve the issue of "semantic gap". A formal image description language was defined to connect the user's abstract description to low-level features of images. According to the user's description and amendments, a pattern image generates. Then this image was as a model to research in the image retrieval systems. Retrieval efficiency can be improved.