首页 | 本学科首页   官方微博 | 高级检索  
     

基于低秩矩阵恢复的SAR图像相干斑抑制方法
引用本文:李 斌,孙 骁,张水平,窦 浩,明德烈,田金文.基于低秩矩阵恢复的SAR图像相干斑抑制方法[J].现代雷达,2014(5):49-52.
作者姓名:李 斌  孙 骁  张水平  窦 浩  明德烈  田金文
基金项目:国家自然科学基金资助项目
摘    要:针对合成孔径雷达(SAR)图像相干斑噪声的特点,提出了一种基于低秩矩阵恢复的SAR图像相干斑抑制算法。该算法首先对SAR图像进行对数变换,将SAR图像相干斑乘性噪声转化为加性噪声;然后对变换后图像等步长遍历提取图像子参考块,利用局部块匹配技术寻找子参考块的相似块组建相似子集,合并数据集中所有相似子集,构建近似低秩的矩阵;再通过低秩矩阵恢复算法将矩阵分解为低秩矩阵部分和稀疏矩阵部分;最后将低秩矩阵部分逆变换回图像块,基于图像块灰度值对图像的每个像素进行加权重构,生成相干斑抑制后的SAR图像。实验表明,文中所提出的算法能够有效抑制SAR图像中的相干斑噪声,同时很好地保留了边缘细节特征。


SAR Speckle Denoising Based on Low-rank Matrix Recovery
LI Bin,SUN Xiao,ZHANG Shuiping,DOU Hao,MING Delie and TIAN Jinwen.SAR Speckle Denoising Based on Low-rank Matrix Recovery[J].Modern Radar,2014(5):49-52.
Authors:LI Bin  SUN Xiao  ZHANG Shuiping  DOU Hao  MING Delie and TIAN Jinwen
Abstract:A SAR speckle denoising algorithm based on low-rank matrix recovery is proposed in this paper. First, multiplicative speckle is changed into additive noise by logarithmic transformation. Then the image is partitioned into blocks, and a block-matching technique is employed in grouping and constructing the approximately low-rank matrix. The matrix is decomposed into lowrank matrix and sparse matrix using low-rank matrix recovery algorithm. Finally, the image after speckle denoising is computed by weighted reconstruct each pixels based on gray level of the blocks. Experimental results on different SAR images demonstrate that the algorithm proposed can reduce the speckle and preserve edges effectively.
Keywords:SAR  speckle denoising  low-rank matrix recovery  block-matching
点击此处可从《现代雷达》浏览原始摘要信息
点击此处可从《现代雷达》下载全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号