一种基于颜色的图像表示及全局相似检索技术

基于内容的图像检索是当图像数据库领域中的一个研究热点，给出了一种描述图像视觉特征的图像表示方法，并据此计算图像之间的全局相似度，首先，通过对彩色空间的分析，提取图像中的颜色不变量，然后在频域内对这样的颜色信息进行分析，对频域分析的结果进行K-L变换，变换后的低维向量即为图像的颜色表示，在此基础上，也讲座他图像的相似度量以及相应的图像检索技术，并给出了实验结果和图像检索性能的评价。

A COLOR-BASED IMAGE REPRESENTATION AND GLOBAL SIMILARITY MATCHING FOR COLOR IMAGES

The retrieval of images based on their similarity to an example image is an important and fascinating area of research. In this paper, an image representation technique to characterize visual appearance for determining global similarity in images is described. Based on the analysis of the color spaces, color invariants are computed from images, which are transformed by using the DFT. Then, the PCA-based dimensionality reduction is used, and the dimensionality-reduced vectors are used as the image representation for similarity matching. A metric of similarity between images is defined, allowing to perform retrieval by using the above representation. Experimental results are reported and presented to demonstrate the effectiveness and efficiency of the proposed approach.