文是一种基于灰度共生矩阵的图像相似性检索算法，首先将图像分成互不重叠的子图像块，根据子图像

块中像素间灰度差别重新设置每个像素的灰度值为0或l，然后将整个图像重新划分成子块，对子块编码，最后借

助共生矩阵提取的不同方向的特征值来检索图像的相似性。实验结果表明本文算法对图像相似性的检索比传统

方法GLCM、CCM更有效，且检索效率较高，更重要的是此算法还可以反应不同方向上图像的相似度。-Man is an image based on gray level co-occurrence matrix of similarity search algorithm, the first image into non-overlapping sub-image block, the sub-image blocks of pixels gray Cibie reset each pixel grayscale value 0 or l, then the image re-divided into sub-blocks, pair block coding, and finally extracted with co-occurrence matrix eigenvalues in different directions to retrieve similar images. Experimental results show that the algorithm for image similarity retrieval method than the traditional GLCM, CCM is more effective and efficient retrieval, more importantly, the algorithm can also respond in different directions on the image similarity.