张量分解和自适应图全变分的高光谱图像去噪

doi:10.19665/j.issn1001-2400.20230412

西安电子科技大学学报 ›› 2024, Vol. 51 ›› Issue (2): 157-169.doi: 10.19665/j.issn1001-2400.20230412

• 计算机科学与技术&网络空间安全 • 上一篇下一篇

张量分解和自适应图全变分的高光谱图像去噪

蔡明娇¹(), 蒋俊正¹^,²(), 蔡万源¹(), 周芳¹^,³()

1.桂林电子科技大学信息与通信学院,广西壮族自治区桂林 541004
2.桂林电子科技大学卫星导航定位与位置服务国家地方联合工程研究中心,广西壮族自治区桂林 541004
3.桂林电子科技大学广西无线宽带通信与信号处理重点实验室,广西壮族自治区桂林 541004

收稿日期:2023-02-23 出版日期:2024-04-20 发布日期:2023-09-06
通讯作者: 周芳(1984—),女,副教授,E-mail:zhoufang1026@guet.edu.cn
作者简介:蔡明娇(1997—),女,桂林电子科技大学硕士研究生,E-mail:21022201003@mails.guet.edu.cn;
蒋俊正(1983—),男,教授,E-mail:jzjiang@guet.edu.cn;
蔡万源(1998—),男,桂林电子科技大学硕士研究生,E-mail:20022201001@mails.guet.edu.cn
基金资助:
国家自然科学基金(62171146);国家自然科学基金(62261014);广西创新驱动发展专项(桂科AA21077008);广西自然科学杰出青年基金(2021GXNSFFA220004);广西科技基地和人才专项(桂科AD21220112);广西无线宽带通信与信号处理重点实验室主任基金(GXKL06220107);桂林电子科技大学研究生教育创新计划(2022YCXS039)

Hyperspectral image denoising based on tensor decomposition and adaptive weight graph total variation

CAI Mingjiao¹(), JIANG Junzheng¹^,²(), CAI Wanyuan¹(), ZHOU Fang¹^,³()

1. School of Information and Communication,Guilin University of Electronic Technology,Guilin 541004,China
2. State and Local Joint Engineering Research Center for Satellite Navigation and Location Service, Guilin University of Electronic Technology,Guilin 541004,China
3. Guangxi Key Laboratory of Wireless Broadband Communication and Signal Processing,Guilin University of Electronic Technology,Guilin 541004,China

Received:2023-02-23 Online:2024-04-20 Published:2023-09-06

摘要/Abstract

摘要：

高光谱图像在采集过程中受到观测条件、成像仪材料属性、传输条件等客观因素的影响,不可避免地会引入各种噪声。这严重降低了高光谱图像的质量以及限制了后续处理的精度。因此,高光谱图像去噪是一个极其重要的预处理步骤。针对高光谱图像去噪问题,提出了低秩张量分解和自适应图全变分的高光谱图像去噪算法。首先,利用低秩张量分解来描述高光谱图像的全局空间和光谱相关性,并使用自适应权重图全变分来刻画高光谱图像空间维度上的分段平滑特性和保留高光谱图像的边缘信息;此外,采用l₁-范数、Frobenius-范数分别刻画包括条纹噪声、脉冲噪声、死线噪声在内的稀疏噪声和高斯噪声。由此高光谱图像去噪问题归结为一个包含低秩张量分解和自适应图全变分的约束优化问题。利用增广拉格朗日乘子法对该优化问题进行交替求解。实验结果表明,所提出的高光谱图像去噪算法与现有的算法相比,能够充分刻画高光谱图像数据的内在结构特性,具有更好的去噪性能。

关键词: 高光谱图像去噪, Tucker分解, 自适应图全变分

Abstract:

During the acquisition process of hyperspectral images,various noises are inevitably introduced due to the influence of objective factors such as observation conditions,material properties of the imager,and transmission conditions,which severely reduces the quality of hyperspectral images and limits the accuracy of subsequent processing.Therefore,denoising of hyperspectral images is an extremely important preprocessing step.For the hyperspectral image denoising problem,a denoising algorithm,which is based on low-rank tensor decomposition and adaptive weight graph total variation regularization named LRTDGTV,is proposed in this paper.Specifically,Low-rank tensor decomposition is used to characterize the global correlation among all bands,and adaptive weight graph total variation regularization is adopted to characterize piecewise smoothness property of hyperspectral images in the spatial domain and preserve the edge information of hyperspectral images.In addition,sparse noise,including stripe noise,impulse noise and deadline noise,and Gaussian noise are characterized by l₁-norm and Frobenius-norm,respectively.Thus,the denoising problem can be formulated into a constrained optimization problem involving low-rank tensor decomposition and adaptive weight graph total variation regularization,which can be solved by employing the augmented Lagrange multiplier(ALM) method.Experimental results show that the proposed hyperspectral image denoising algorithm can fully characterize the inherent structural characteristics of hyperspectral images data and has a better denoising performance than the existing algorithms.

Key words: hyperspectral image denoising, tucker decomposition, adaptive weight graph total variation

中图分类号:

TP751

蔡明娇, 蒋俊正, 蔡万源, 周芳. 张量分解和自适应图全变分的高光谱图像去噪[J]. 西安电子科技大学学报, 2024, 51(2): 157-169.

CAI Mingjiao, JIANG Junzheng, CAI Wanyuan, ZHOU Fang. Hyperspectral image denoising based on tensor decomposition and adaptive weight graph total variation[J]. Journal of Xidian University, 2024, 51(2): 157-169.

图/表 14

图1

表1

图2

图3

图4

图5

图6

图7

图8

图9

图10

图11

图12

图13

参考文献 22

[1]	HE W, ZHANG H Y, ZHANG L P, et al. Total-Variation-Regularized Low-Rank Matrix Factorization for Hyperspectral Image Restoration[J]. IEEE Transactions on Geoscience and Remote Sensing, 2016, 54(1):178-188.
[2]	HE C, SUN L, HUANG W, et al. TSLRLN:Tensor Subspace Low-Rank Learning with Non-Local Prior for Hyperspectral Image Mixed Denoising[J]. Signal Processing, 2021, 184(2):108060.
[3]	CHEN Y, HUANG T Z, HE W, et al. Hyperspectral Image Denoising Using Factor Group Sparsity-Regularized Nonconvex Low-Rank Approximation[J]. IEEE Transactions on Geoscience and Remote Sensing, 2022, 60(5515916):1-16.
[4]	曾海金, 蒋家伟, 赵佳佳, 等. L_1-2空谱全变差正则化下的高光谱图像去噪[J]. 光子学报, 2019, 48(10):214-228.
	ZENG Haijin, JIANG Jiawei, ZHAO Jiajia, et al. L_1-2 Spectral-Spatial Total Variation Regularized Hyperspectral Image Denoising[J]. Acta Photonica Sinica, 2019, 48(10):214-228.
[5]	WANG Y, PENG J J, ZHAO Q, et al. Hyperspectral Image Restoration Via Total Variation Regularized Low-Rank Tensor Decomposition[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2018, 11(4):1227-1243.
[6]	ELAD M, AHARON M. Image Denoising Via Sparse and Redundant Representations over Learned Dictionaries[J]. IEEE Transactions on Image Processing, 2006, 15(12):3736-3745. doi: 10.1109/tip.2006.881969 pmid: 17153947
[7]	JIANG H Y, CHENG Z Y, HUO Y, et al. Research on Image Denoising Methods Based on Wavelet Transform and Rolling-Ball Algorithm[C]//2007 International Conference on Wavelet Analysis and Pattern Recognition. Piscataway:IEEE, 2007: 1604-1607.
[8]	DABOV K, FOI A, KATKOVNIK V, et al. Image Denoising by Sparse 3-D Transform-Domain Collaborative Filtering[J]. IEEE Transactions on Image Processing, 2007, 16(8):2080-2095. doi: 10.1109/tip.2007.901238 pmid: 17688213
[9]	ZHANG H Y. Hyperspectral Image Denoising with Cubic Total Variation Model[J]. ISPRS Annals of the Photogrammetry,Remote Sensing and Spatial Information Sciences, 2012,I-7:95-98.
[10]	OTHMAN H, QIAN S E. Noise Reduction of Hyperspectral Imagery Using Hybrid Spatial-Spectral Derivative-Domain Wavelet Shrinkage[J]. IEEE Transactions on Geoscience and Remote Sensing, 2006, 44(2):397-408.
[11]	ZHANG H Y, HE W, ZHANG L P, et al. Hyperspectral Image Restoration Using Low-Rank Matrix Recovery[J]. IEEE Transactions on Geoscience and Remote Sensing, 2014, 52(8):4729-4743.
[12]	CAI W Y, JIANG J Z, OUYANG S. Hyperspectral Image Denoising Using Adaptive Weight Graph Total Variation Regularization and Low-Rank Matrix Recovery[J]. IEEE Geoscience and Remote Sensing Letters, 2022, 19(5509805):1-5.
[13]	陈俊杰, 邓洪高, 马谋, 等. GCN-GRU:一种无线传感器网络故障检测模型[J]. 西安电子科技大学学报, 2022, 49(5):60-67.
	CHEN Junjie, DENG Honggao, MA Mou, et al. GRN-GRU:A Fault Detection Model for Wireless Sensor Networks[J]. Journal of Xidian University, 2022, 49(5):60-67.
[14]	徐莎莎, 周芳, 李杨剑, 等. 一种新的传感器节点分布式定位算法[J]. 西安电子科技大学学报, 2022, 49(2):89-96.
	XU Shasha, ZHOU Fang, LI Yangjian, et al. New Distributed Positioning Algorithm for Sensor Nodes[J]. Journal of Xidian University, 2022, 49(2):89-96.
[15]	JIANG J Z, FENG H R, TAY D B, et al. Theory and Design of Joint Time-Vertex Nonsubsampled Filter Banks[J]. IEEE Transactions on Signal Processing, 2021,69:1968-1982.
[16]	LIU M J, WEI Y. Image Denoising Using Graph-Based Frequency Domain Low-Pass Filtering[C]//2019 IEEE 4th International Conference on Image,Vision and Computing(ICIVC). Piscataway:IEEE, 2019: 118-122.
[17]	ZHANG Y, HAN X Y, ZHANG H, et al. Edge Detection Algorithm of Image Fusion Based on Improved Sobel Operator[C]//2017 IEEE 3rd Information Technology and Mechatronics Engineering Conference(ITOEC). Piscataway:IEEE, 2017: 457-461.
[18]	李健, 吕倩. 高光谱图像噪声分析与降噪模型概述[J]. 现代信息科技, 2019, 3(2):23-24.
	LI Jian, LV Qian. Overview of Noise Analysis and Noise Reduction Models for Hyperspectral Images[J]. Modern Information Technology, 2019, 3(2):23-24.
[19]	BRCK A, TEBOULLE M. Fast Gradient-Based Algorithms for Constrained Total Variation Image Denoising and Deblurring Problems[J]. IEEE Transactions on Image Processing, 2009, 18(11):2419-2434. doi: 10.1109/TIP.2009.2028250 pmid: 19635705
[20]	ZHUANG L N, FU X, NG M K, et al. Hyperspectral Image Denoising Based on Global and Nonlocal Low-Rank Factorizations[J]. IEEE Transactions on Geoscience and Remote Sensing, 2021, 59(12):10438-10454.
[21]	HE W, YAO Q M, LI C, et al. Non-Local Meets Global:An Iterative Paradigm for Hyperspectral Image Restoration[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2022, 44(4):2089-2107.
[22]	RASTI B, ULFARSSON M O, SVEINSSON J R. Hyperspectral Subspace Identification Using SURE[J]. IEEE Geoscience and Remote Sensing Letters, 2015, 12(12):2481-2485.

Indian Pines(145×145×224)
Case	Level	Indexes	LRTV	LRTDTV	GLF	NGmeet	FGSLR	LRTDGTV
1	G=0.075	R_MPSNR	39.493 4	39.248 4	28.838 5	28.235 1	37.563 9	40.068 5
	P=0.15	R_MSSIM	0.984 5	0.990 9	0.959 6	0.963 0	0.964 0	0.993 5
		R_ERGAS	38.888 1	27.560 6	88.330 8	95.288 5	32.521 7	25.000 7
2	G=0.1	R_MPSNR	37.393 6	36.739 5	26.410 3	26.399 3	35.486 9	38.807 6
	P=0.2	R_MSSIM	0.975 6	0.984 1	0.940 6	0.935 6	0.936 8	0.991 0
		R_ERGAS	58.548 6	36.399 2	117.352 9	117.413 1	40.851 4	28.607 9
3	Gaussian	R_MPSNR	36.622 8	37.306 8	31.273 3	29.419 4	36.006 0	38.741 2
	+impulse	R_MSSIM	0.973 4	0.985 9	0.969 6	0.965 5	0.949 2	0.991 3
		R_ERGAS	72.491 7	34.760 7	77.143 5	88.708 0	38.753 4	28.823 4
4	Gaussian	R_MPSNR	36.416 1	37.143 7	29.603 4	28.633 6	35.701 0	38.451 6
	+impulse	R_MSSIM	0.973 2	0.986 1	0.908 6	0.960 6	0.948 0	0.991 0
	+deadlines	R_ERGAS	78.409 4	35.045 4	96.641 6	97.248 9	40.135 0	29.614 0

张量分解和自适应图全变分的高光谱图像去噪

Hyperspectral image denoising based on tensor decomposition and adaptive weight graph total variation

RichHTML

PDF (PC)

赞

可视化

摘要/Abstract

引用本文

使用本文

图/表 14

参考文献 22

相关文章 15

Metrics

本文评价

推荐阅读 0

[1]	韩永赛,马时平,何林远,李承昊,朱明明,张飞. 改进YOLOv3的快速遥感机场区域目标检测[J]. 西安电子科技大学学报, 2021, 48(5): 156-166.
[2]	朱苏雅,杜建超,李云松,汪小鹏. 采用U-Net卷积网络的桥梁裂缝检测方法[J]. 西安电子科技大学学报, 2019, 46(4): 35-42.
[3]	杨建功,汪西莉,刘侍刚. 融合谱-空域信息的DBM高光谱图像分类方法[J]. 西安电子科技大学学报, 2019, 46(3): 109-115.
[4]	孙宸, 成立业. 空间感知矩阵学习的极化SAR图像分类[J]. 西安电子科技大学学报, 2018, 45(6): 92-98.
[5]	马荣贵;崔华;薛世焦;郭璐;袁超. 改进蚁群算法的多约束质量最优路径选择[J]. 西安电子科技大学学报, 2016, 43(3): 185-189.
[6]	赵宝玮;相里斌;黄旻;吕群波;张桂峰. 视差约束的改进Hough变换多光谱影像匹配方法[J]. J4, 2015, 42(6): 94-98.
[7]	张泽均;水鹏朗. G⁰分布中链码网格SAR图像分割算法[J]. J4, 2015, 42(5): 48-54.
[8]	崔华;严玍伻. 融合卡方累积分布函数和PDE的曲波阈值去噪法[J]. J4, 2014, 41(4): 108-110+172.
[9]	赵宝玮;相里斌;吕群波;张桂峰;曾晓茹;郭文记. FPGA和多DSP系统的并行RX探测算法[J]. J4, 2014, 41(3): 152-156.
[10]	颜学颖;焦李成;王凌霞. 利用Brushlet变换进行SAR图像变化检测[J]. J4, 2013, 40(6): 67-73.
[11]	白键;冯象初. 去除乘性噪声的积分微分方程模型[J]. J4, 2013, 40(3): 132-138.
[12]	白键;冯象初. 非凸泛函的图像分解[J]. J4, 2013, 40(2): 67-71+137.
[13]	白键冯象初. 非凸泛函的图像分解[J]. J4, 2013, 40(2): 67-71.
[14]	辛芳芳;焦李成;王桂婷;万红林. 利用小波域HMC模型进行遥感图像变化检测[J]. J4, 2012, 39(3): 43-49.
[15]	李权合;毕笃彦;何林远. 退化过程模拟模型及其在图像增强中的应用[J]. J4, 2011, 38(6): 185-192.