提出了一种基于矢量量化的纹理分类方法，通过矢量量化的方法生成原始图象的符号图象，并生成这个符号图象的共生矩阵，行家和共生矩阵的4个数字特征构成原始图象的纹理特征向量，使用2个特征向量之间的矩离作为匹配的标准，可以实现对纹理图象进行快速匹配分类。

This paper gives a method of texture classification based on vectorquantization. By using vectorquatization to create the symbol images of original images,the cooccurences matrix of these symbol images can be gotten. Four digital features of cooccurences matrix are selected to obtain texture feature vectors of original images. Using the distance between two tex ture feature vectors as a matching standar, guick match and classification of texture images can be realized.