The Design And Realize Of Content-based Image Retrieval On Clothes Electronic Business System

With the development of the computer network to expand, application of electroniccommerce based on the Internet in enterprise’s showing the charm and great developmentpotential. Virtual enterprise, Internet banking, online payment, online shopping has becomeindispensable in people’s daily life. More and more companies see the tremendous businessopportunities and sales network advantage, would like to take this opportunity to expand thescale of enterprises, to obtain more profits.According to the characteristics of clothing images, based on the text of the query, try toadd the clothing color, contour and texture search function based on. The use of quantitativemethod of HSV color model, image color feature with the block color histogram method,measure the retrieval of the clothing color using similarity with weight. On the part of edgeoperators are compared and choose the Sobel operator as edge detection operator, invariantmoments as the shape features to HU, using Euclidean distance image edge. In the retrieval ofclothing texture features, research and experiment of GLCM and Gabor filters in two ways,and completed the clothing texture feature retrieval. According to realizing the integratedretrieval methods three features of clothing color, contour and texture. Finally, the whole workprocessing are summarized, and put forward the modification and future research directions.