Color And Shape-based Image Retrieval System

With the rapid increase of image application, it is an urgent task thathow to search the image effectively and quickly. At the same time, image iswidely used more and more, and image query in www(Wide World Web) needcontent-based image retrieval urgently.This paper studies the image retrieval methods based on the image colorcontent and image shape content and also designs the system construction ofimage retrieval, meanwhile, it presents the different formulae according to thevariable retrieval feature. Firstly, the paper offers the relevant imagepreprocess methods to extract the image feature accurately and easily.Secondly, the paper studies the retrieval methods based on image colorcontent on the basis of mathematics substation thinking, which can make thecolor feature having the spatiaI comparability. Thirdly, rotation and scale ofimage shape feature extracted from image have been verified by experiment inshape-based image retrieval. Finally, this paper presents various similarformula for various retrieval methods, so it can value the retrieval result byquantification. In the end, the retrieval methods in the paper are tested to beeffective and fast by experiments.