郑州大学学报（理学版）2017，Vol.49Issue(2)：72-77,6.DOI:10.13705/j.issn.1671-6841.2016320
基于NSST变换域WNNM和KAD算法的SAR图像去噪
SAR Image Denoising Based on NSST with WNNM and KAD
摘要
Abstract
The SAR image denoising based on non-subsample shearlet transform with weighted nuclear norm minimization and kernel anisotropic diffusion was presented to minimize the effect of speckle noise in synthetic aperture radar.Firstly, the image global noise variance was estimated in advance.Secondly, multiplicative speckle was changed into additive noise by logarithmic transformation.Thirdly the SAR image was decomposed by no`n-subsample shearlet transform;the high frequency component were processed by kernel anisotropic diffusion;and low frequency component was processed by WNNM algorithm.Finally, the reconstructed image was reconstructed by NSST algorithm.An efficient implementation of this algorithm was presented in full detail.Also the comparison of this improved algorithm with the NSST and WNNM approach were given.The experimental results showed that the peak signal to noise ratio objective indicators had significantly improved, the local structure of the image was better preserved, and the good subjective visual effect was produced.关键词
合成孔径雷达图像去噪/非下采样剪切波变换/加权核范数最小化/核各向异性扩散Key words
synthetic aperture radar/non-subsample shearlet transform/weighted nuclear norm minimization/kernel anisotropic diffusion分类
信息技术与安全科学引用本文复制引用
赵杰,王配配..基于NSST变换域WNNM和KAD算法的SAR图像去噪[J].郑州大学学报（理学版）,2017,49(2):72-77,6.基金项目
国家自然科学基金项目(61572063,61401308) （61572063,61401308）
河北省自然科学基金项目(F2016201187) （F2016201187）
河北大学自然科学研究计划项目(2014-303) （2014-303）
河北大学研究生创新项目(X2015085). （X2015085）
