Color Texture Image Retrieval Research

With the artificial intelligence technology and the Internet application technology of rapid development,image processing and recognition technology is more and more attention from all walks of life.Color texture image analysis is one of the research focus in the field of image processing and pattern recognition,and content-based image retrieval becomes a research hot spot,the main method is to extract the image of the underlying features such as texture,color,shape,etc.,Using the characteristics of the target image compared with image features in the database,find out the most similar images and display the results,it is concluded that the correct recognition rate.In this paper,the color texture image retrieval research,mainly several aspects has the following:First,this paper introduces the research background of the image retrieval technology,the development status and trend at home and abroad,the current several classic image retrieval system,are introduced USPTex texture database by using color image library in this paper.Second,the study of color texture image retrieval technology,in terms of color feature extraction,have RGB color space,HSV color space and HSI color space,and so on,the study of the commonly used color space module,the color space of quantitative and extracting characteristic and similarity comparison method.Again,in terms of extraction of texture feature,this paper introduces the use of texture feature extraction algorithm such as LBP algorithm,the algorithm of LCCP and CS-LDP algorithms and their respective advantages and disadvantages,etc.and puts forward the use of LBP and LCCP fusion algorithm,the introduction of the main steps and process of lamination fusion algorithm,and compared experimental study in the number of training samples and the block number is different under the condition of these kinds of texture feature extraction algorithm,confirms the superiority of the combination of LBP and LCCP features of the proposed algorithm.Finally,study the integration method of using both color and texture features,color feature extraction is RGB color histogram,texture feature extraction is the choice of combination of LBP and LCCP algorithm,the experiments of recognition rate and retrieval efficiency are carried out,the experimental results is better than that two kinds of features used alone,it is effective to improve the method of image library in this paper,that can achieve the desired experimental results.