基于截断核范数的遥感影像混合噪声去除方法  

Mixed Noise Removal Method for Remote Sensing Images Based on Truncated Kernel Norm

作　　者：刘婉芸 贾世会[1,2] 李高西 迟晓妮 LIU Wan-yun;JIA Shi-hui;LI Gao-xi;CHI Xiao-ni(College of Science,Wuhan University of Science and Technology,Wuhan Hubei 430081,China;Hubei Province,Key Laboratory of Systems Science in Metallurgical Process,Wuhan Hubei 430081,China;College of Mathematics and Statistics,Chongqing Normal University,Chongqing 400067,China;College of Mathematics and Computing Science,Guilin University of Electronic Technology,Guilin Guangxi 541004,China)

机构地区：[1]武汉科技大学理学院,湖北武汉430081 [2]武汉科技大学冶金工业工程系统科学湖北省重点实验室,湖北武汉430081 [3]重庆工商大学数学与统计学院,重庆400067 [4]桂林电子科技大学数学与计算机科学学院,广西桂林541004

出　　处：《计算机仿真》2025年第3期249-253,422,共6页Computer Simulation

基　　金：国家自然科学基金(12361064);国家自然科学基金资助项目(11901068,11861026);广西自然科学基金(2021GXNSFAA 220034);湖北省教育厅科学研究计划资助青年项目(Q20211111);湖北省冶金工业过程系统科学重点实验室开放基金(Y201905);重庆市自然科学基金面上项目(CSTB2022NSCQ-MSX0606)。

摘　　要：针对遥感影像中主要含有高斯噪声和椒盐噪声的特点,提出一种基于截断核范数的鲁棒主成分分析去噪模型。该模型基于广义鲁棒主成分分析模型去除混合噪声,利用截断核范数凸近似秩函数,保证了秩函数的准确逼近,并运用随机排序的交替方向乘子法对新模型进行求解,数值实验结果显示,新的模型对混合噪声污染的图片不仅能够有效的分离出低秩部分、稀疏大噪声部分和稠密小噪声部分,而且图像去噪效果更佳,客观标准上新模型的PSNR值和ERR也优于TRPCA和GRPCA模型。Aiming at the characteristics of Gaussian noise and salt noise in remote sensing images,a robust principal component analysis denoising model based on the truncated nuclear norm is proposed.Based on the generalized robust principal component analysis model to remove the mixed noise,and the truncated nuclear norm convex approximate rank function is used to ensure the low-rank property of the matrix,the new model is solved by the alternating direction multiplier method of random ordering.Numerical experimental results show that the new model can not only effectively separate low-rank parts,sparse loud noise parts and dense small noise parts from images with mixed noise pollution,but also has a better image denoising effect.The PSNR and ERR values of the new model are better than those of TRPCA and GRPCA models on objective criteria.

关 键 词：遥感影像 混合噪声 截断核范数 随机排序 交替方向乘子法 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]