基于颜色和纹理特征的图像检索研究

【作者】 李迎新；

【导师】 张明；

【作者基本信息】 上海海事大学 ， 计算机应用， 2007， 硕士

【摘要】 基于内容的图像检索(CBIR)技术是当前研究的热点问题。它主要是利用图像的视觉特征，如图像的颜色、纹理、形状等特征来进行检索。它突破了传统的基于文本检索技术的局限，直接对图像内容进行分析并抽取特征，然后利用这些内容特征建立索引并进行检索。本文针对基于内容的图像检索的主要方法进行了研究。在颜色特征提取和匹配中，通过研究和分析不同的颜色空间和颜色特征，提出对HSV颜色空间进行合理的量化，提取颜色直方图形成特征矢量用于检索的方法。针对颜色直方图不包含任何空间信息问题，本文给出了一种基于HSV色彩空间累加直方图进行图像检索的方法，并设计了一种具有重叠方式的图像分块方法，即合理地将图像分成若干子快，为图像提出了一定程度的位置信息。本文选取了基于量化的颜色特征、基于共生矩阵的纹理特征进行单特征检索，在这两个特征的基础上，通过构建各特征量的权值，来实现多特征的图像检索。设计了分别基于颜色特征和基于纹理特征的两种图像检索算法。在利用单一特征检索的基础上，提出了一种综合利用上述两个特征共同进行检索的方法。对真实图像数据库的检索实验表明，综合特征检索要比单一特征检索更符合人的视觉感受要求，因而检索效果更好。更多还原

【Abstract】 At present,Content-Based Image Retrieval(CBIR) is becoming a hot research topic.lt is a retrieval technology based on the vision features.such as the color.texture and shape of the image.lt is different from traditional retrieval technology based on the text.This technology directly analyzes image content and extracts image features. These contental features to be built index and be used in retrieval.The main techniques of CBIR are discussed in this thesis. During the study of color features extraction and matching, HSV color model is chosen and divided into small spaces according to the perception of human eyes. The details of extracting HSV-based color histogram are described. In order to deal with the shortcoming of color histogram - it includes no spatial distribution information of the image, this thesis proposes a new content-based color image retrieval method, in which both the color and the spatial relationship of image are taken into account and then designs 12 overlapped sub-regions to get the color accumulative histogram in each region.The new method divides the image into blocks, which contain spatial information. The color feature based quantification,texture feature based Co-occurrence matrix are selected in feature retrieval.In multi-features retrieval,the feedback will be used in adjusting weight and re-used in multi-features retrieval.The methods for image retrieval using color and texture features are first discussed. On the basis of using color and texture features separately, a new method for image retrieval using combined color and texture feature is proposed. Retrienal experiments using real color image database are carried out.The results show that the retrieval results obtained from combined-features fits more closely with human perception than the retrieval results obtained from single-features.更多还原