Image Retrieval Based On Color And Texture Features

Content-based Image Rrtrieval(CBIR) is a kind of retrieval,which derictly use content of an image for information search, it is one of most active researches in multimedia retrieval field.In order to analysis the informations included in an image,the CBIR system always ultilize the color,texture,shape and other low level image features.to establish the feature vectors as retrieval index.In present time,the main CBIR method is similiarity search based on multi-dimension feature vector of image.In this paper,we maimly reseach a retrieval method based on both color feature and texture feature of BMP image.In the deformed HSL color space , we extract a number of colors that can representate the original image exactly.Then the texture features are introduced.Using above two features vectors, the CBIR method try to search similar images in image database.First of all ,we discuss the model definition,characteristic and suitable range of several color models from optics,visual psychology and digital image processing.Then a deformed HSL color model conformed to color cluster feature of human's vision is established.In the space, a suitable cluster algorithm is used to extract main colors, then the original image is converted to main image.Secondly,because the avility of expressing image semantic informations with the color feature is too limited, so a texture feature is presented, we discuss the mothed of extracting texture of image.Finally, in this paper , we using a kind of comprehensive image retrieval which fuses color and texture features by linear weights,and discuss the method which the weights are determined. We have completed program design by Visual Basic.Except above contents,BMP image analize, the match algorithm of both color and texture comprehensive feature,the common frame of CBIR system and the methods for fast retrieval are discussed in this paper.The CBIR method in this paper use not only color feature but also texture feature. The experiment gives good results.