一种基于去除干扰基团 Markov 随机场的SAR 图像分割算法

摘要

This article discusses the relationship between the Markov random field model and image . With aid of the iterated conditional model (ICM ) algorithm , the labels of the image were updated , and then combined with region growing thoughts for segmentation of SAR image post-processing , a SAR image segmentation method is given based on removed the interference clique Markov random field . The numerical experiments have been compared with the traditional Markov segmentation algorithm in terms of segmentation time , iteration numbers , and convergence energy . The results showing that this method has higher segmentation accuracy , more smooth target boundaries and better segmentation effects .%研究了马尔科夫随机场模型与图像的关系。利用条件迭代模型算法逐点更新图像标记，并结合区域生长思想，提出了一种基于去除干扰基团 Markov 随机场的 SAR 图像分割方法。数值试验通过与传统 Markov 分割算法从分割时间、迭代次数和收敛能量进行对比分析，结果表明，该方法具有更高的分割正确率，分割目标边界更清晰平滑，分割效果也更好。