基于异质性测量和非局部平均的斑点噪声抑制

发布时间：2018-01-10 07:15

本文关键词：基于异质性测量和非局部平均的斑点噪声抑制　出处：《计算机应用研究》2014年08期 　论文类型：期刊论文

【摘要】：传统的欧氏距离不能鲁棒地度量含有相干斑噪声的SAR图像块之间的相似性。针对这一问题,将SAR图像的异质性测量方法与传统的欧氏距离结合,产生了一种新的SAR图像相似性测度,并在此基础上,提出了一种新的用于斑点噪声抑制的非局部平均滤波算法。该算法首先计算相似性窗口之间的欧氏距离和搜索窗口的变差系数;然后利用搜索窗口的变差系数自动调整退化参数h,并在欧氏距离和调整后的退化参数的基础上计算新的SAR图像相似性测度;最后利用新的相似性测度对待处理像素点进行非局部平均恢复。对仿真与实际SAR图像的斑点噪声抑制实验表明,新算法能有效去除斑点噪声和保留边缘纹理等细节区域,视觉效果很好;且与现有的非局部抑斑算法相比,计算复杂度大大降低。
[Abstract]:The traditional Euclidean distance can not measure the similarity between the blocks of SAR images with speckle noise. In order to solve this problem, the heterogeneity measurement method of SAR images is combined with the traditional Euclidean distance. A new similarity measure of SAR image is presented and based on it. A new non-local average filter algorithm for speckle noise suppression is proposed. The Euclidean distance between similarity windows and the coefficient of variation of the search window are calculated. Then the variation coefficient of the search window is used to automatically adjust the degradation parameter h, and a new similarity measure of SAR image is calculated on the basis of Euclidean distance and the adjusted degradation parameter. Finally, a new similarity measure is used to deal with the non-local average recovery of the processed pixels. The experimental results of speckle noise suppression in simulated and actual SAR images show that the proposed method is effective. The new algorithm can effectively remove speckle noise and preserve edge texture and other detail areas, and the visual effect is very good. Compared with the existing non-local speckle suppression algorithm, the computational complexity is greatly reduced.
【作者单位】：中南民族大学电子信息工程学院;中南民族大学智能无线通信湖北省重点实验室;
【基金】：国家自然科学基金资助项目(61201448) 湖北省自然科学基金资助项目(2011CHB043,2012FFA113) 中南民族大学中央高校基本科研业务费专项资金自科一般项目(CZY10001)
【分类号】：TN957.52
【正文快照】： 0引言合成孔径雷达(SAR)是一种主动式微波遥感器,由于具有全天时、全天候成像、高分辨率和强穿透力等优点,被广泛应用到军事和民用的各领域。然而,由于SAR系统的相干成像特点,SAR图像不可避免地受到相干斑噪声的污染。相干斑噪声的存在严重影响了SAR图像的视觉效果及其自动解

【参考文献】