纹理图像分割的首要问题是纹理特征的提取与描述 .传统的统计纹理描述方法和结构纹理描述方法各有特点 ,但也各有不足 .本文从结构与统计结合的角度 ,提出了一种基于灰度秩数的结构性统计方法 ,该方法可以较好地抽取物体表面的结构特征 .在用本算法对纹理进行描述的基础上 ,本文采用了一种广义最小体积椭球聚类算法(GMVE)对所抽取的纹理特征进行聚类分割 .算法有效性在实际纹理分割中得到了验证 The main problem of texture image segmentation is the extraction and description of texture features.The traditional statistical texture description methods and structural texture description methods all have their own characteristics but also have their own shortcomings.This paper presents a method based on the combination of structure and statistics This method can well extract the structural features of the surface of the object.On the basis of describing the texture by this algorithm, a generalized minimum volume ellipsoid clustering algorithm (GMVE) The extracted texture features are clustered and the validity of the algorithm is verified in the actual texture segmentation