计算机科学 ›› 2024, Vol. 51 ›› Issue (12): 199-208.doi: 10.11896/jsjkx.231000187

• 计算机图形学&多媒体 • 上一篇    下一篇

联合群稀疏和代表系数双向空间光谱全变分的高光谱图像去噪

司伟纳, 叶军, 姜斌   

  1. 南京邮电大学理学院 南京 210023
  • 收稿日期:2023-10-27 修回日期:2024-03-27 出版日期:2024-12-15 发布日期:2024-12-10
  • 通讯作者: 叶军(yj8422092@163.com)
  • 作者简介:(17201628783@163.com)
  • 基金资助:
    国家自然科学基金(61971234);南京邮电大学校内基金(NY220209)

Hyperspectral Image Denoising Combining Group Sparse and Representative Coefficient Bidirectional Spatial Spectral Total Variation

SI Weina, YE Jun, JIANG Bin   

  1. School of Science, Nanjing University of Posts and Telecommunications, Nanjing 210023, China
  • Received:2023-10-27 Revised:2024-03-27 Online:2024-12-15 Published:2024-12-10
  • About author:SI Weina,born in 1996,postgraduate.Her main research interests include pattern recognition,remote sensing image processing and machine learning.
    YE Jun,born in 1981,Ph.D,associate professor.His main research interests include pattern recognition,machine learning,and image processing.
  • Supported by:
    National Natural Science Foundation of China(61971234) and Intramural Fund of Nanjing University of Posts and Telecommunications(NY220209).

摘要: 高光谱图像去噪是遥感领域的一个基本问题,也是预处理的重要步骤。基于代表系数全变分的去噪方法在高光谱图像(HSI)去噪中有着广泛的应用。代表系数矩阵U继承了干净HSI的先验信息,能够实现全局低秩并降低计算复杂度,但由于一阶全变分的引入,该类方法在去噪过程中产生了很强的阶梯效应并且忽略了不同波段间的共同特征,因此去噪效果很差。针对此问题,提出了一种新的联合群稀疏和代表系数双向空间光谱全变分(RCBGS)的正则化去噪模型。高阶全变分的引入缓解了阶梯效应,并在子空间的差分上引入加权$\ell$2,1范数,充分挖掘不同波段除全局低秩外的共同特征,提高了HSI的内在群稀疏性和整体光滑性。最后,通过交替方向乘子法(ADMM)给出了所提方法的迭代规则,且所提方法的评价指标峰值信噪比相对于对比方法平均提升了8.79%。在模拟和真实数据集上的实验表明,所提方法在视觉质量和定量评估方面都优于相关方法。

关键词: 高光谱图像去噪, 双向变分, 低秩先验, 阶梯效应, 群稀疏

Abstract: Hyperspectral image denoising is a fundamental problem in remote sensing field,which is an important step of preprocessing.Denoising method based on total variation of representative coefficients is widely used in hyperspectral image(HSI) denoising.Representative coefficient matrix U inherits prior information of clean HSI,which can achieve global low rank and reduce computational complexity.However,due to the introduction of first-order total variational,this method produces a strong step effect in the process of denoising and ignores the common features between different bands,so the denoising effect is poor.To solve this problem,a new regularized denoising model of joint group sparse and representative coefficient bidirectional spatial spectral total variational(RCBGS) is proposed.By introducing high-order total variational,the step effect is alleviated,and the weighted $\ell$2,1 norm is introduced into the difference of subspace to fully explore the common features of different bands except global low rank,and improve the intrinsic group sparsity and overall smoothness of HSI.Finally,the iterative rules of the proposed method are given by alternate direction multiplier method(ADMM),and the evaluation index peak signal-to-noise ratio of the proposed method is improved by 8.79% on average compared with the comparison methods.Experiments on simulated and real datasets show that the proposed method outperforms relative methods in both visual quality and quantitative evaluation.

Key words: Hyperspectral image denoising, Bidirectional variation, Low-rank prior, Staircase effect, Group sparse

中图分类号: 

  • TP751
[1]HOU Y,ZHU W,WANG E.Hyperspectral mineral target detection based on density peak[J].Intelligent Automation and Soft Computing,2019,25(4):805-814.
[2]FAUVEL M,TARABALKA Y,BENEDIKTSSON J A,et al.Advances in spectral-spatial classification of hyperspectral images[J].Proceedings of the IEEE,2023,101(3):652-675.
[3]TIWARI K C,ARORA M K,SINGH D.An assessment of independent component analysis for detection of military targets from hyperspectral images[J].International Journal of Applied Earth Observation and Geoinformation,2011,13(5):730-740.
[4]SUN L,ZHAN T,WU Z,et al.A Novel 3D Anisotropic Total Variation Regularized Low Rank Method for Hyperspectral Image Mixed Denoising[J].ISPRS International Journal of Geo-Information,2018,7(10):412-425.
[5]DABOV K,FOI A,KATKOVNIK V,et al.Image Denoising bySparse 3-D Transform-Domain Collaborative Filtering[J].IEEE Transactions on Image Processing,2007,16(8):2080-2095.
[6]AHARON M,ELAD M,BRUCKSTEIN A.K-SVD:An algo-rithm for designing overcomplete dictionaries for sparse representation[J].IEEE Transactions on Signal Processing,2006,54(11):4311-4322.
[7]GU S,ZHANG L,ZUO W,et al.Weighted Nuclear Norm Minimization with Application to Image Denoising[J].IEEE Confe-rence on Computer Vision and Pattern Recognition,2014,32:2862-2869.
[8]HE W,ZHANG H,ZHANG L,et al.Total-Variation-Regula-rized Low-Rank Matrix Factorization for Hyperspectral Image Restoration[J].IEEE Transactions on Geoscience and Remote Sensing,2016,54(1):178-188.
[9]HE W,ZHANG H,SHEN H,et al.Hyperspectral Image De-noising Using Local Low-Rank Matrix Recovery and Global Spatial-Spectral Total Variation[J].IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing,2018,11(3):713-729.
[10]PENG J,WANG Y,ZHANG H,et al.Exact Decomposition of Joint Low Rankness and Local Smoothness Plus Sparse Matrices[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2023,45(5):5766-5781.
[11]CAO X,FU X,XU C,et al.Deep Spatial-Spectral Global Reasoning Network for Hyperspectral Image Denoising[J].IEEE Transactions on Geoscience and Remote Sensing,2021,60:1-14.
[12]ZHENG Y,HUANG T,ZHAO X,et al.Double-Factor Regula-rized Low-Rank Tensor Factorization for Mixed Noise Removal in Hyperspectral Image[J].IEEE Transactions on Geoscience and Remote Sensing,2020,58(12):8450-8464.
[13]CHEN Y,HE W,YOKOYA N,et al.Hyperspectral image restoration using weighted group sparsity-regularized low-rank tensor decomposition[J].IEEE Transactions on Cybernetics,2020,50(8):3556-3570.
[14]PENG J,WANG H,CAO X,et al.Fast Noise Removal in Hyperspectral Images via Representative Coefficient Total Variation[J].IEEE Transactions on Geoscience and Remote Sen-sing,2022,60:1-17.
[15]CHANG Y,YAN L,FANG H,et al.Anisotropic Spectral-Spatial Total Variation Model for Multispectral Remote Sensing Image Destriping[J].IEEE Transactions on Image Processing,2015,24(6):1852-1866.
[16]AGGARWAL H,MAJUMDAR A.Hyperspectral Image De-noising Using Spatio-Spectral Total Variation[J].IEEE Geo-science and Remote Sensing Letters,2016,13(3):442-446.
[17]YANG J,ZHAO X,MEI J,et al.Total variation and high-order total variation adaptive model for restoring blurred images with Cauchy noise[J].Computers and Mathematics with Applications,2019,77(5):1255-1272.
[18]ZHU J,LV H,HAO B,et al.Non-Convex High Order Total Variation With Overlapping Group Sparsity Denoising Model Under Cauchy Noise[J].IEEE Access,2021,9:49901-49911.
[19]CAI J,HE W,ZHANG H.Anisotropic Spatial-Spectral TotalVariation Regularized Double Low-Rank Approximation for HSI Denoising and Destriping[J].IEEE Transactions on Geoscience and Remote Sensing,2022,60:1-19.
[20]CHOKSI R,GENNIP Y.Deblurring of One Dimensional BarCodes via Total Variation Energy Minimization[J].SIAM Journal of Imaging Science,2009,3(4):735-764.
[21]ZHANG H,CAI J,HE W,et al.Double low-rank matrix decomposition for hyperspectral image denoising and destriping[J].IEEE Transactions on Geoscience and Remote Sensing,2021,60:1-19.
[22]PENG J,XIE Q,ZHAO Q,et al.Enhanced 3DTV Regularization and Its Applications on HSI Denoising and Compressed Sensing[J].IEEE Transactions on Image Processing,2020,29:7889-7903.
[23]HUYNH-THU Q,GHANBARI M.Scope of validity of PSNR in image/video quality assessment[J].Electronics Letters,2008,44(13):800-801.
[24]WANG Z,BOVIK A C,SHEIKH H R,et al.Image quality assessment:from error visibility to structural similarity[J].IEEE Transaction Image Processing,2004,13(4):600-612.
[25]SUN L,JEON B,ZHENG Y,et al.A Novel Weighted Cross Total Variation Method for Hyperspectral Image Mixed Denoising[J].IEEE Access,2017,5:27172-27188.
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!