Color-based Image Retrieval Technology

Content Based Image Retrieval (CBIR) has been a very active research area since 1990's, with the thrust from two major research communities, Database Management and Computer Vision. The traditional approach of Image Retrieval is based on the technology of Database Management Systems (DBMS), with the cost of heavy burden of manual annotation. The proposed Content Based Image Retrieval, however, is a new approach based on Computer Vision, Pattern Recognition which perform the computer-centered image retrieval according to the content of images.As first, this article reviews the theory, applications and its development of Content-Based Image Retrieval, and summarizes the technology and current state of image retrieval, especially the state of content-based image retrieval. Furthermore, this article shows the important technology in content-based image retrieval system (CBIRS) and introduces some famous CBIRS. Secondly, In order to perfect image data and restrain image deform and enhance useful image character, this article introduces classic techniques of pre-processing, there were implemented with Visual C++ 6.0 programming language. For example: image enhancement and adjustment of brightness and contrast. Through pre-processing there will build a database with perfect images. Next, This Article introduces image regular character and matching techniques of content-based image retrieval and color-based image retrieval.Color is the important features of an image. Almost all present CBIR systems use color as an essential content for matching and retrieval. Based on block histogram algorithm for color image retrieval using in CBIR system, The algorithm is improved in constructing a block histogram, merging the histogram distance etc, The performance of color matching pair retrieval algorithm that makes further use of the spatial similarity information between example image and retrieved images is analyzed. In the Windows 2000 environment, the image retrieval system, which was based on the image color characteristic and used the search algorithm of color matching pair, was implemented with Visual C++ 6.0 programming language. Theretrieval precision and recall are increased effectively.The paper holds certain referential value and practical significance in promoting the development of retrieval technique of image database.