基于颜色和纹理特征图像检索-应用数学专业论文.docxVIP

Content-Based Content-Based Image Retrieval Combining Color and Texture Abstract． very large collections of images are growing ever more common．From stock photo collections to proprietary databases to the Web，these collections are diverse and often poorly indexed；unfortunately，image retrieval systems have not kept pace with the collections they are searching． In this paper，we emphases on four parts Firstly，Color is a feature of the great majority of content—based image retrieval system．However the robustness，effectiveness，and efficiency of its use in image indexing are still open issues．This paper provides a comprehensive survey of the methods for color image indexing and retrieval described in the literature． Secondly，A novel K—Means color clustering algorithm is used，and a new color distance is also defined for this algorithm．The well—known Lab color space is used for K—means clustering．only a and b is selected as L encodes luminance which is not steady．The mean values of color in one cluster are used to represent color feature， which is similarity with human vision，and Can represent the picture dominant color precisely．Experiments shows the proposed method is more effective and efficient than Color Coherence Vector and Color Structure Descriptor． Thirdly，three texture extraction methods：cooccurrence matrix， edge histogram descriptor and Gabor wavelets，are compared．Experiments show that the frequency method as gabor wavelets is better than the other two methods which extract texture feature in space． Fourthly，Image retrieal prototype system is built by combing k-means color clustering and Gabor wavelet．it supports query by example and dominant color retrieval． Ⅱ Content-Based Content-Based Image Retrieval Combining Color and Texture Keywords：k-means clustering，Gabor wavelet，content·based image retrieval 111 中山大学硕士学位论文：基于颜色和纹理特征图像检索第1章 中山大学硕士学位论文：基于颜色和纹理特征图像检索 第1章 引言 1．1论文选题的背景 随着数字图像处理技术的广泛应用和Intemet的普及，涌现出大规模的数字 图像库，有效的管理和查询图像的需求也越来越迫切，图像检索技术引起了图像 处理、模式识别、计算机视觉