基于细节信息约束的遥感影像条带噪声去除模型

doi:10.11947/j.AGCS.2024.20230363

摘要/Abstract

摘要：

遥感影像在获取过程中会经常受到条带噪声的污染，降低遥感影像的视觉效果，对影像解译和反演等处理产生不利影响。当前一些主流的基于变分的条带噪声去除方法，虽然可以去除条带噪声，但是往往也会导致影像细节信息的严重丢失。基于上述问题，本文提出了一种基于细节信息约束的遥感影像条带噪声去除模型（DISUTV）。在DISUTV模型中，将所提出的基于双边滤波器与正交子空间投影的细节信息分离算子与单向全变分正则化项、群组稀疏正则化项及单向全变分正则约束项进行了有效结合，并采用交替方向乘子法对其进行求解，用于从条带噪声影像中获取不含有细节信息的高精度条带噪声。利用模拟数据与真实数据对本文方法的条带噪声去除能力、细节信息保持能力及稳健性进行了验证并与现有前沿方法进行了比较。试验结果表明，本文方法在去除条带噪声的同时能更好地保留影像的细节信息，并且呈现出了较好的定性与定量结果。

关键词: 条带噪声提取, 正交子空间投影, 细节信息分离算子, 单向全变分, 群组稀疏

Abstract:

Remote sensing images are often contaminated by stripe noise during the acquisition process, which reduces the visual effect of remote sensing images and has an adverse effect on image interpretation and inversion. Although some mainstream stripe noise removal methods based on variational methods can remove stripe noise, they often lead to serious loss of image detail information. Based on the above problems, this paper proposes a remote sensing image stripe noise removal model DISUTV based on detail information constraint. In the DISUTV model, the proposed detail information separation operator based on bilateral filter and orthogonal subspace projection is effectively combined with one-way total variation regularization term, group sparsity regularization term and one-way total variation regularization constraint term, and the alternating direction multiplier method is used to solve it, which is used to obtain high-precision stripe noise without detail information from stripe noise images. The stripe noise removal ability, detail information retention ability and robustness of the algorithm are verified using simulated data and real data, and compared with existing cutting-edge methods. Experimental results show that the proposed method can better retain the detail information of the image while removing stripe noise, and presents good qualitative and quantitative results.

Key words: stripe noise extraction, orthogonal subspace projection, detail information separation operator, one-way full variational splitting, group sparsity

中图分类号:

P237

王密, 董滕滕, 彭涛, 项韶, 兰穹穹. 基于细节信息约束的遥感影像条带噪声去除模型[J]. 测绘学报, 2024, 53(9): 1799-1816.

Mi WANG, Tengteng DONG, Tao PENG, Shao XIANG, Qiongqiong LAN. Remote sensing image stripe noise removal model based on detail information constraints[J]. Acta Geodaetica et Cartographica Sinica, 2024, 53(9): 1799-1816.

图/表 21

图1

图2

图3

图4

图5

图6

图7

图8

表1

图9

图10

表2

表3

图11

图12

图13

图14

图15

表4

图16

图17

参考文献 33

[1]	SHEN Huanfeng, HUANG Liwen, ZHANG Liangpei, et al. Long-term and fine-scale satellite monitoring of the urban heat island effect by the fusion of multi-temporal and multi-sensor remote sensed data: a 26-year case study of the city of Wuhan in China[J]. Remote Sensing of Environment, 2016, 172:109-125.
[2]	LI Xiaotong, ZHAO Xile, JIANG Taixiang, et al. Low-rank tensor completion via combined non-local self-similarity and low-rank regularization[J]. Neurocomputing, 2019, 367:1-12.
[3]	DING Meng, HUANG Tingzhu, JI Tengyu, et al. Low-rank tensor completion using matrix factorization based on tensor train rank and total variation[J]. Journal of Scientific Computing, 2019, 81(2):941-964.
[4]	ACITO N, DIANI M, CORSINI G. Subspace-based striping noise reduction in hyperspectral images[J]. IEEE Transactions on Geoscience and Remote Sensing, 2011, 49(4):1325-1342.
[5]	KAMIRU L, HAKIM P R, HASBI W. An adaptive stripe noise removal algorithm for optical satellite imagery[C]//Proceedings of 2019 IEEE International Conference on Aerospace Electronics and Remote Sensing Technology. Yogyakarta: IEEE, 2019: 1-7.
[6]	JIANG Hongxu, MIAO Rui, CHEN Jiao, et al. A resource-efficient parallel architecture for infrared image stripe noise removal based on the most stable window[J]. Infrared Physics & Technology, 2019, 97:258-269.
[7]	李文力, 李凯, 彭迪, 等. AGRI遥感图像中条带噪声的分析与去除[J]. 光学学报, 2019, 39(12):1228004.
	LI Wenli, LI Kai, PENG Di, et al. Analysis and removal of stripe noise in AGRI remote-sensing images[J]. Acta Optica Sinica, 2019, 39(12):1228004.
[8]	CUI Hao, JIA Peng, ZHANG Guo, et al. Multiscale intensity propagation to remove multiplicative stripe noise from remote sensing images[J]. IEEE Transactions on Geoscience and Remote Sensing, 2020, 58(4):2308-2323.
[9]	QU Yufu, ZHANG Xuan, WANG Qianyi, et al. Extremely sparse stripe noise removal from non remote-sensing images by straight line detection and neighborhood grayscale weighted replacement[J]. IEEE Access, 2018, 6:76924-76934.
[10]	QU Xiaofei, ZHAO Weiwei, LONG En, et al. Removal of stripes in remote sensing images based on statistics combined with image enhancement[J]. Journal of Geodesy and Geoinformation Science, 2023, 6(1):76-87.
[11]	张霞, 孙伟超, 帅通, 等. 基于小波变换的图像条带噪声去除方法[J]. 遥感技术与应用, 2015, 30(6):1168-1175.
	ZHANG Xia, SUN Weichao, SHUAI Tong, et al. Image destriping method based on wavelet transform[J]. Remote Sensing Technology and Application, 2015, 30(6):1168-1175.
[12]	LI M, NONG S, NIE T, et al. An infrared stripe noise removal method based on multi-scale wavelet transform and multi-nomial sparse representation[J]. Computational Intelligence and Neuroscience, 2022:4044071.
[13]	SHAO Yuyi, SUN Yingwei, ZHAO Mengmeng, et al. Infrared image stripe noise removing using least squares and gradient domain guided filtering[J]. Infrared Physics & Technology, 2021, 119:103968.
[14]	陈劲松, 邵芸, 朱博勤. 中分辨率遥感图像条带噪声的去除[J]. 遥感学报, 2004(3):227-233.
	CHEN Jinsong, SHAO Yun, ZHU Boqin. Removal of strip noise from medium-resolution remote sensing images[J]. Journal of Remote Sensing, 2004(3):227-233.
[15]	FAYYAZ Z, PLATNICK D, FAYYAZ H, et al. Deep unfolding for iterative stripe noise removal[C]//Prcoceedings of 2022 International Joint Conference on Neural Networks. Padua: IEEE, 2022: 1-7.
[16]	GUAN Juntao, LAI Rui, XIONG Ai. Wavelet deep neural network for stripe noise removal[J]. IEEE Access, 2019, 7:44544-44554.
[17]	BOUALI M, LADJAL S. Toward optimal destriping of MODIS data using a unidirectional variational model[J]. IEEE Transactions on Geoscience and Remote Sensing, 2011, 49(8):2924-2935.
[18]	LIU Na, LI Wei, TAO Ran, et al. Wavelet-domain low-rank/group-sparse destriping for hyperspectral imagery[J]. IEEE Transactions on Geoscience and Remote Sensing, 2019, 57(12):10310-10321.
[19]	CHEN Yong, HUANG Tingzhu, ZHAO Xile. Destriping of multispectral remote sensing image using low-rank tensor decomposition[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2018, 11(12):4950-4967.
[20]	WANG Ende, JIANG Ping, LI Xuepeng, et al. Infrared stripe correction algorithm based on wavelet decomposition and total variation-guided filtering[J]. Journal of the European Optical Society-Rapid Publications, 2020, 16:1.
[21]	ZHAO Shuheng, YUAN Qiangqiang, LI Jie, et al. A fast and effective irregular stripe removal method for moon mineralogy mapper (M³)[J]. IEEE Transactions on Geoscience and Remote Sensing, 2022, 60:1-19.
[22]	WANG Chang, ZHANG Yongsheng, WANG Xu, et al. An effective strip noise removal method for remote sensing image[J]. Journal of Geodesy and Geoinformation Science, 2022, 5(4):72-85.
[23]	王昶, 王旭, 纪松. 采用变分法的遥感影像条带噪声去除[J]. 西安交通大学学报, 2019, 53(3):143-149.
	WANG Chang, WANG Xu, JI Song. Stripe noise removal of remote images based on variation[J]. Journal of Xi'an Jiaotong University, 2019, 53(3):143-149.
[24]	陈柳, 陈明举, 吴浩, 等. 一种群稀疏限制的消除条带噪声变分模型[J]. 液晶与显示, 2020, 35(6):604-611.
	CHEN Liu, CHEN Mingju, WU Hao, et al. Variation model with group sparsity constraint for stripe noise removal[J]. Chinese Journal of Liquid Crystals and Displays, 2020, 35(6):604-611.
[25]	李凯, 李文力, 韩昌佩. 基于L1范数优化模型的遥感图像条纹去除方法[J]. 红外与毫米波学报, 2021, 40(2):272-283.
	LI Kai, LI Wenli, HAN Changpei. Remote sensing image stripe removal method based on L1-paradigm optimization model[J]. Journal of Infrared and Millimeter Wave, 2021, 40(2):272-283.
[26]	CHANG Yi, YAN Luxin, WU Tao, et al. Remote sensing image stripe noise removal: from image decomposition perspective[J]. IEEE Transactions on Geoscience and Remote Sensing, 2016, 54(12):7018-7031.
[27]	ZHANG Shengwei, XIANG Wei, XU Baoshu, et al. Stripe noise removal for infrared images using guided filter[C]//Proceedings of 2016 Infrared Technology and Applications, and Robot Sensing and Advanced Control. Beijing: SPIE, 2016: 665-673.
[28]	鞠荟荟, 刘志刚, 姜江军, 等. 基于低通滤波残差图的高光谱条带噪声去除[J]. 光学学报, 2018, 38(12):361-369.
	JU Huihui, LIU Zhigang, JIANG Jiangjun, et al. Hyperspectral stripe noise removal based on low-pass filtered residual map[J]. Journal of Optics, 2018, 38(12):361-369.
[29]	REN Longfei, HONG Danfeng, GAO Lianru, et al. Orthogonal subspace unmixing to address spectral variability for hyperspectral image[J]. IEEE Transactions on Geoscience and Remote Sensing, 2023, 61:1-13.
[30]	REN Longfei, WANG Chengjing, TANG Peipei, et al. A dual symmetric Gauss-Seidel alternating direction method of multipliers for hyperspectral sparse unmixing[J]. Numerical Algorithms, 2021, 87(2):719-754.
[31]	REN Longfei, MA Zheng, BOVOLO F, et al. Hyperspectral unmixing using weighted sparse regression with total variation regularization[J]. International Journal of Remote Sensing, 2022, 43(15/16):6124-6151.
[32]	YANG Jinghua, ZHAO Xile, MA Tianhui, et al. Remote sensing images destriping using unidirectional hybrid total variation and nonconvex low-rank regularization[J]. Journal of Computational and Applied Mathematics, 2020, 363:124-144.
[33]	CHEN Yong, HUANG Tingzhu, ZHAO Xile, et al. Stripe noise removal of remote sensing images by total variation regularization and group sparsity constraint[J]. Remote Sensing, 2017, 9(6):559.

影像	指标	WDLRWGS	UTV	LRTD	TVGF	DIRN	LRHP	本文方法
噪声影像1	I_S	1.325 13	6.077 10	0.986 62	1.840 28	0.881 21	14.370 89	0.993 05
	I_M	0.000 13	0.001 17	0.000 08	0.000 19	0.000 05	0.002 98	0.000 01
	SSIM	0.977 67	0.905 25	0.997 33	0.998 74	0.989 06	0.773 78	0.999 63
噪声影像2	I_S	1.215 80	3.183 34	1.650 13	1.739 21	0.757 49	10.378 90	0.978 89
	I_M	0.000 23	0.001 19	0.000 39	0.000 40	0.000 13	0.005 07	0.000 01
	SSIM	0.979 99	0.907 89	0.998 48	0.998 14	0.986 54	0.923 52	0.999 32
噪声影像3	I_S	2.749 15	4.099 88	2.591 54	1.946 61	0.863 16	15.347 40	0.973 95
	I_M	0.001 50	0.002 49	0.001 43	0.000 77	0.000 22	0.028 61	0.000 03
	SSIM	0.875 91	0.788 50	0.982 69	0.972 57	0.885 28	0.724 37	0.985 58
噪声影像4	I_S	6.609 29	6.718 18	1.673 47	1.813 51	0.842 48	6.222 32	0.953 44
	I_M	0.006 59	0.005 90	0.000 76	0.001 02	0.000 35	0.005 21	0.000 16
	SSIM	0.981 85	0.857 07	0.996 58	0.993 45	0.987 90	0.929 46	0.998 97
噪声影像5	I_S	2.235 93	19.599 77	1.891 68	1.650 29	0.749 28	22.744 06	1.101 17
	I_M	0.001 20	0.016 38	0.000 81	0.000 86	0.000 35	0.048 87	0.000 32
	SSIM	0.974 46	0.880 79	0.995 36	0.981 21	0.967 33	0.815 50	0.995 67

影像	指标	WDLRWGS	UTV	LRTD	TVGF	DIRN	LRHP	本文方法
测试影像1	I_S	1.325 13	6.077 10	0.986 62	1.840 28	0.881 21	14.370 89	0.993 05
	I_M	0.000 13	0.001 17	0.000 09	0.000 19	0.000 05	0.002 99	0.000 01
	SSIM	0.977 67	0.905 25	0.997 33	0.998 74	0.989 06	0.773 78	0.999 63
测试影像2	I_S	1.136 28	4.365 11	0.918 53	1.328 23	0.701 87	9.796 84	0.980 72
	I_M	0.000 15	0.001 12	0.000 11	0.000 11	0.000 19	0.002 99	0.000 01
	SSIM	0.976 84	0.890 30	0.995 98	0.998 68	0.961 72	0.773 92	0.999 31
测试影像3	I_S	0.973 98	4.225 19	0.709 79	1.018 74	0.817 20	6.185 95	0.989 26
	I_M	0.000 11	0.001 85	0.000 05	0.000 04	0.000 30	0.002 98	0.000 02
	SSIM	0.973 98	0.883 26	0.994 00	0.997 32	0.933 51	0.773 83	0.998 15
测试影像4	I_S	1.491 04	2.413 47	1.108 92	1.481 22	0.724 29	4.399 20	0.952 36
	I_M	0.000 62	0.001 40	0.000 42	0.000 50	0.000 46	0.003 00	0.000 06
	SSIM	0.973 07	0.891 89	0.991 35	0.990 16	0.905 62	0.773 74	0.992 53
测试影像5	I_S	1.064 84	2.566 63	0.797 23	1.097 03	0.729 26	3.992 92	0.940 76
	I_M	0.000 27	0.001 61	0.000 20	0.000 19	0.000 53	0.003 00	0.000 07
	SSIM	0.970 85	0.897 71	0.989 06	0.989 22	0.892 17	0.773 49	0.990 71

影像	指标	噪声影像	WDLRWGS	UTV	LRTD	TVGF	DIRN	LRHP	本文方法
真实噪声影像1	δ	138.855 94	147.801 22	137.872 99	137.164 33	138.678 79	142.138 95	106.009 90	139.651 91
	PSNR	56.996 71	66.197 22	56.099 15	69.571 66	65.311 47	62.086 04	57.844 93	69.900 19
	SSIM	0.854 24	0.979 09	0.945 11	0.993 16	0.989 97	0.932 50	0.843 57	0.993 47
	H	6.604 80	6.560 11	6.431 75	6.555 04	6.575 33	6.564 55	6.427 18	6.593 57
	NR	0	0.920 62	1.413 41	1.124 26	1.390 46	1.115 73	1.339 22	1.391 13
真实噪声影像2	δ	323.049 52	328.230 63	316.036 46	316.548 27	323.549 80	319.592 59	265.940 65	323.565 86
	PSNR	63.470 20	65.751 399	63.119 75	65.701 25	67.668 15	65.892 48	63.351 93	74.613 66
	SSIM	0.995 52	0.977 38	0.973 21	0.976 60	0.999 50	0.976 89	0.900 98	0.999 68
	H	6.370 75	6.372 71	6.336 83	6.372 033	6.372 56	6.365 75	6.200 99	6.374 60
	NR	0	0.939 95	1.202 94	0.958 07	1.096 46	1.041 65	1.072 50	1.116 32