提出一种基于马尔柯夫随机场 （ Ｍ Ｒ Ｆ） 模型和多层前馈神经网络进行纹理分割的方法．利用二阶高斯 Ｍ Ｒ Ｆ模型对图像纹理进行描述，采用最小平方误差方法进行参数估计，将估计参数作为纹理的特征向量，并且利用改进的 Ｂ Ｐ算法对特征进行分类．对纹理图像进行的实验表明，与常规距离方法相比，采用这种方法进行纹理分割能取得更好的效果． A method of texture segmentation based on Markov random field (M R F) model and multi-layer feedforward neural network is proposed. The second-order Gaussian M R F model is used to describe the image texture. The least squares error method is used to estimate the parameters. The estimated parameters are used as the eigenvectors of the texture, and the improved B P algorithm is used to classify the features. Experiments on texture images show that using this method to texture segmentation can achieve better results than the conventional distance method.