【作者】 张朋朋； 王善峰； 公茂果； 张明阳； 武越；

【Author】 Zhang Pengpeng;Wang Shanfeng;Gong Maoguo;Zhang Mingyang;Wu Yue;Key Laboratory of Intelligent Perception and Image Understanding,Ministry of Education,Xidian University;

【机构】 西安电子科技大学；

【摘要】 多时相遥感影像的变化检测是遥感应用的重要方面,已经在国民经济和国防建设领域得到广泛的应用。为了抑制SAR图像中的相干斑噪声,解决变化检测结果图中含有较多噪点的问题,提高变化检测精度,本文首先引入改进的三维块匹配去噪技术对SAR图像进行去噪处理,然后利用模糊C均值聚类将均值比差异图分割成变化类和未变化类,得到初始变化检测二值图,最后根据均值比差异图和初始变化检测二值图建立全连接条件随机场,采用最大后验概率对全连接条件随机场进行标记,得到最终的变化检测结果图。实验结果表明,改进的三维块匹配去噪算法可以有效抑制SAR图像中的相干斑噪声,使得生成的均值比差异图中变化区域和未变化区域具有更高的对比度;同时,全连接条件随机场利用了像素点周围的邻域信息,减少了初始变化检测二值图中的噪声点,进一步提高了变化检测的精度。更多还原

【Abstract】 Change detection of multi-temporal remote sensing images is an important aspect of remote sensing applications and has been widely used in the national economy and national defense construction.In order to suppress the speckle noise in the SAR image and solve the problem of noise in the change map and improve the detection performance.Firstly,the blockmatching 3-D algorithm is introduced to denoise the two-phase SAR images,and then the fuzzy clustering algorithm divides the mean ratio difference map into the changed and unchanged areas to obtain the initial change map.Finally,based on the mean ratio difference map and the initial change map,a fully connected conditional random field is established,and the maximum posterior probability is utilized to solve the full connected condition field to obtain a final change map.Experimental results show that the block-matching 3-D algorithm can effectively suppress the speckle noise in the SAR image,so that the changed areas have higher contrast than the unchanged areas in the difference map.What’s more,the neighborhood information around the pixel used in the full connected conditional random field reduces the noise points in the initial change map so further improving the accuracy of the change detection.更多还原