论文部分内容阅读
提出了一种基于同步自回归（ＳＡＲ） 模型和模糊信息原理进行纹理分割的方法。利用二阶ＳＡＲ 模型对图像纹理进行描述，用最小平方误差方法对模型参数进行估计，在对模型参数分析的基础上，将估计的参数进行改进后作为纹理的特征向量用于纹理图像的分类与分割。由于实际图像带有许多的模糊信息，组成纹理的基元和基元之间的关系也具有很大的模糊性，文中根据模糊信息原理，分析了纹理图像的模糊特性，给出了一种基于模糊贴近度的纹理分割方法。实验结果表明，与常规的距离方法相比，用文中的方法进行图像纹理分割能取得更好的效果。
A method of texture segmentation based on synchronous auto-regressive (SAR) model and fuzzy information principle is proposed. The second-order SAR model is used to describe the image texture, and the least square error method is used to estimate the model parameters. Based on the analysis of the model parameters, the estimated parameters are improved and used as texture eigenvectors to classify the texture images segmentation. Because the actual image has a lot of fuzzy information, the relationship between the primitives and the primitives that make up the texture also has a great fuzziness. According to the principle of fuzzy information, the fuzzy characteristics of the texture image are analyzed. Texture Segmentation with Fuzzy Closeness. Experimental results show that, compared with the conventional distance method, texture segmentation using the method in this paper can achieve better results.