Hyperspectral Image Denoising Using Improved Low-Rank and Sparsity Constraints

被引:0
作者
Zhong, Chongxiao [1 ]
Zhang, Junping
Guo, Qingle [1 ]
机构
[1] Harbin Inst Technol, Inst Image & Informat Technol, Harbin, Heilongjiang, Peoples R China
来源
EARTH OBSERVING SYSTEMS XXIII | 2018年 / 10764卷
关键词
Hyperspectral image (HSI) denoising regularization; low-rank matrix reconstruction; sparse representation; iterative reweighting; ALGORITHM;
D O I
10.1117/12.2320148
中图分类号
TP7 [遥感技术];
学科分类号
081102 ; 0816 ; 081602 ; 083002 ; 1404 ;
摘要
Hyperspectral imaging has been widely applied in many fields due to the advantage of high spectral resolution. However, consisting of acquisition, transmission, reception and display a hyperspectral imaging system may be disturbed in each part and thus leads to degradations that limit the precision of subsequent processing. It is therefore an important preprocessing step to remove the noise of acquired image data as much as possible. In this paper. we propose a novel regularization method for hyperspectral image denoising. Firstly, the low-rank and sparsity constraints are jointly used to establish the regularization model. For each spectral band, the low-rank constraint is for exploiting inter-column/-row correlations, while the sparsity constraint aims to exploit intra-column correlations. Secondly, reweighed l(1), norm strategy, which solves a sequence of weighted l(1), norm optimization problems and updates the weights with the solution of the last iteration, is introduced to approximate l(0) norm to achieve improved priori performance of the two constraints. Lastly, we apply the alternating direction method (ADM) under the augmented Lagrangian multiplier (ALM) framework to solve the model efficiently. Both low-rank and sparsity priors arc reweighted at each iteration to promote low-rankness and sparseness of the solution. The demising effect of our method is tested on real hyperspectral image data with different noise level. The experiments demonstrate the practicality of our proposed method.
引用
收藏
页数:7
相关论文
共 5 条
  • [1] A SINGULAR VALUE THRESHOLDING ALGORITHM FOR MATRIX COMPLETION
    Cai, Jian-Feng
    Candes, Emmanuel J.
    Shen, Zuowei
    [J]. SIAM JOURNAL ON OPTIMIZATION, 2010, 20 (04) : 1956 - 1982
  • [2] Lin Z, 2009, AUGMENTED LAGRANGE M, V9
  • [3] Reweighted Low-Rank Matrix Recovery and its Application in Image Restoration
    Peng, Yigang
    Suo, Jinli
    Dai, Qionghai
    Xu, Wenli
    [J]. IEEE TRANSACTIONS ON CYBERNETICS, 2014, 44 (12) : 2418 - 2430
  • [4] Hyperspectral Image Denoising Using First Order Spectral Roughness Penalty in Wavelet Domain
    Rasti, Behnood
    Sveinsson, Johannes R.
    Ulfarsson, Magnus Orn
    Benediktsson, Jon Atli
    [J]. IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING, 2014, 7 (06) : 2458 - 2467
  • [5] Toh KC, 2010, PAC J OPTIM, V6, P615