Research On Retrieving Images Based On The Color Characteirstic Of Images

In recent years,with the rapid development of Internet,the multimedia informationsuch as digital images has increased dramatically. How to effectively organize, manageand retrieve large image data becomes the urgent problem to be solved. Content BasedImage Retrieval (CBIR) technology has become a research hotspot at present.Althoughsome achievements have obtained, many theoretical and technical difficulties have tobe a breakthrough, the application also has some problems to be solved.In this paper the image color feature is mainly studied. First of all, the choice ofcolor model, color quantization, color histogram method construction and similaritymatching algorithm were analyzed in detail. The existing division method and colorquantizing method was discovered some problems, which to some extent affected theimage retrieval efficiency and accuracy. Therefore, based on the picture photographycomposition method and human visual perception characteristics research, a newmethod for image retrieval which based on the visual focus area histogram wasproposed.Secondly, because of the lack of the" lace border" image retrieval methods,this paper puts forward a new image regional division method of the color entropybased on the rectangle ring of the image.Through simulation test and comparative analysis, the image retrieval methodbased on the visual focus area histogram has been confirmed that it has high precision,strong robustness and small calculation quantity characteristics. To the " lace border"images, the image retrieval method based on rectangular ring color entropy has goodretrieval performance. Because these two methods are all for the purposes of the image’scolor characteristics and have the absence of other underlying characteristics, whichhave an impact on retrieval performance, future research will focus on this matter.