A Content Based Image Retrieval Tool' Research And Implementing Via Network

With the development of image processing technology and high performance computer coming into being, image processing make great progress in both theory research and application, which has been being applied to more and more field. Image retrieval is a important technology. Traditional retrieval method using such as file name, tile and keywords etc. can't meet the need of application, so retrieval technology based on the visual feature, i.e. content based retrieval technology has been studied by many researchers in the previous several years. Image retrieval involves image processing, image recognition and imd)ge database etc. that has been an inevitable technology at image arrangement and administration in an image database.In this paper, content based theory and method is discussed in detail and a summary is given about elementary procedure and key technology in content based retrieval. By studying theory about color and shape feature extraction and color and shape query theory, a new algorithm which can extract interior shape of image area, is given. At last, a network based image retrieval system using shape and color feature is designed.C/S configuration is adopted in this system, which ensure user to access SQL Server image database by a client. The system includes five modules: image query module, image feature extraction module, similitude match module, image database manage module and image feature display module, which is used separately to input image, extract color and shape feature, image feature match, append or delete image record for image for database, display color histogram and shape feature.