基于联合稀疏表示的复Contourlet域SAR图像与红外图像融合

吴一全; 王志来

doi:10.12000/JR17019

SAR and Infrared Image Fusion in Complex Contourlet Domain Based on Joint Sparse Representation

doi: 10.12000/JR17019

Wu Yiquan^{1,2,3,4,5,6
, ,},
Wang Zhilai^1
,

①.
College of Electronic and Information Engineering, Nanjing University of Aeronautics and Astronautics, Nanjing 211106, China
②.
Jiangsu Key Laboratory of Big Data Analysis Technology, Nanjing University of Information Science & Technology, Nanjing 210044, China
③.
Zhejiang Province Key Laboratory for Signal Processing, Zhejiang University of Technology, Hangzhou 310023, China
④.
Guangxi Key Lab of Multi-Source Information Mining and Security, Guangxi Normal University, Guilin 541004, China
⑤.
Key Laboratory of Geo-Spatial Information Technology, Ministry of Land and Resources, Chengdu University of Technology, Chengdu 610059, China
⑥.
MLR Key Laboratory of Metallogeny and Mineral Assessment Institute of Mineral Resources, Chinese Academy of Geological Sciences, Beijing 100037, China

Funds: The National Natural Science Foundation of China (61573183), The Open Fund of Jiangsu Key Laboratory of Big Data Analysis Technology (KXK1403), The Open Fund of Zhejiang Province Key Laboratory for Signal Processing (ZJKL_6_SP-OP 2014-02), The Open Fund of Guangxi Key Lab of Multi-Source Information Mining and Security (MIMS14-01), The Open Fund of Key Laboratory of Geo-Spatial Information Technology (KLGSIT2015-05), The Open Fund of MLR Key Laboratory of Metallogeny and Mineral Assessment Institute of Mineral Resources (ZS1406)

More Information

Author Bio:
Wu Yiquan (1963–), male, professor, Ph.D. supervisor, Ph.D. degree. He received the doctorate from Nanjing University of Aeronautics and Astronautics in 1998. He is now a professor, Ph.D. supervisor of Nanjing University of Aeronautics and Astronautics. His current research interests include remote sensing image processing and understanding, target detection and identification, visual detection and image measurement, video processing and intelligence analysis, etc. He has published more than 280 papers at home and abroad academic journals. E-mail: nuaaimage@163.com

Wang Zhilai (1992–), male, born in Jiangxi province. He is a graduate student with department of information and communication engineering at college of electronic and information engineering in Nanjing University of Aeronautics and Astronautics. His research interest includes remote sensing image processing and machine vision, etc．E-mail: 1610156025@qq.com

Corresponding author: Wu Yiquan nuaaimage@163.com

摘要

摘要: 针对红外图像与SAR图像的灰度差异性大、两者融合图像不太符合人类视觉认知的问题，提出了一种基于联合稀疏表示的复Contourlet域红外图像与SAR图像融合方法。首先对红外图像与SAR图像分别进行复Contourlet分解。然后利用K-奇异值分解(K-Singular Value Decomposition, K-SVD)方法获得两幅源图像低频分量的过完备字典，并根据联合稀疏表示模型生成联合字典，通过正交匹配追踪(Orthogonal Matching Pursuit, OMP)方法求出源图像低频分量在联合字典下的稀疏表示系数，接着采用选择最大化策略对两个低频分量的稀疏表示系数进行选取，随后进行稀疏表示重构获得融合的低频分量；对高频分量结合视觉敏感度系数和能量匹配度两个活跃度准则进行融合，以捕获源图像丰富的细节信息。最后经复Contourlet逆变换获得融合图像。与3种经典融合方法及近年来提出的基于非下采样Contourlet变换(Non-Subsampled Contourlet Transform, NSCT)、基于稀疏表示的融合方法相比，该方法能够有效突出源图像的显著特征，最大程度地继承源图像的信息。
- 图像融合 /
- SAR图像 /
- 红外图像 /
- 复Contourlet变换 /
- 联合稀疏表示
Abstract: To investigate the problems of the large grayscale difference between infrared and Synthetic Aperture Radar (SAR) images and their fusion image not being fit for human visual perception, we propose a fusion method for SAR and infrared images in the complex contourlet domain based on joint sparse representation. First, we perform complex contourlet decomposition of the infrared and SAR images. Then, we employ the K-Singular Value Decomposition (K-SVD) method to obtain an over-complete dictionary of the low-frequency components of the two source images. Using a joint sparse representation model, we then generate a joint dictionary. We obtain the sparse representation coefficients of the low-frequency components of the source images in the joint dictionary by the Orthogonal Matching Pursuit (OMP) method and select them using the selection maximization strategy. We then reconstruct these components to obtain the fused low-frequency components and fuse the high-frequency components using two criteria——the coefficient of visual sensitivity and the degree of energy matching. Finally, we obtain the fusion image by the inverse complex contourlet transform. Compared with the three classical fusion methods and recently presented fusion methods, e.g., that based on the Non-Subsampled Contourlet Transform (NSCT) and another based on sparse representation, the method we propose in this paper can effectively highlight the salient features of the two source images and inherit their information to the greatest extent.
- Image fusion /
- Synthetic Aperture Radar (SAR) image /
- Infrared image /
- Complex contourlet transform /
- Joint sparse representation

HTML全文

Figure 1. Schematic diagram of CCT

下载: 全尺寸图片幻灯片

Figure 2. The procedure of the proposed image fusion method

下载: 全尺寸图片幻灯片

Figure 3. Three groups of infrared images and SAR images

下载: 全尺寸图片幻灯片

Figure 4. Fusion results 1 by six methods

下载: 全尺寸图片幻灯片

Figure 5. Fusion results 2 by six methods

下载: 全尺寸图片幻灯片

Figure 6. Fusion results 3 by six methods

下载: 全尺寸图片幻灯片

Table 1. Quantitative evaluation of six fusion methods

Image group	Fusion method	IE	MI	CC	SF	AG	SD	Time (s)
Group 1	LP method	6.839	11.405	0.687	32.852	22.406	35.499	2.126
	WT method	6.724	11.119	0.869	25.667	17.947	29.863	1.827
	NSCT method	6.821	11.295	0.813	26.881	18.373	33.828	50.332
	DT-CWT method	6.759	11.236	0.719	25.691	17.881	32.614	3.765
	Method in Ref. [11]	6.887	11.411	0.788	24.312	16.955	32.022	67.782
	Proposed method	7.091	11.827	0.829	30.276	20.294	43.965	27.232
Group 2	LP method	6.973	12.070	0.739	29.114	21.005	33.825	1.728
	WT method	7.097	12.181	0.883	25.643	19.298	35.330	0.898
	NSCT method	6.895	11.808	0.729	23.456	17.272	33.062	39.590
	DT-CWT method	6.902	11.848	0.784	22.633	16.745	32.184	3.371
	Method in Ref. [11]	7.140	12.843	0.7833	30.390	22.565	37.019	53.945
	Proposed method	7.305	12.615	0.869	28.648	20.887	44.539	22.733
Group 3	LP method	6.633	12.141	0.834	28.629	18.687	33.825	1.351
	WT method	6.797	12.1241	0.803	25.643	19.298	35.330	0.557
	NSCT method	6.801	11.973	0.864	26.105	17.539	40.052	26.433
	DT-CWT method	6.665	11.723	0.846	23.539	16.745	32.184	1.371
	Method in Ref. [11]	6.796	12.234	0.812	25.518	17.244	37.274	37.274
	Proposed method	6.864	11.8401	0.836	28.689	18.982	40.084	18.613

下载: 导出CSV

参考文献(16)

[1]	Chen Lei, Yang Feng-bao, Wang Zhi-she, et al. Mixed fusion algorithm of SAR and visible images with feature level and pixel[J]. Opto-Electronic Engineering, 2014, 41(3): 55–60.
[2]	Zeng Xian-wei, Fang Yang-wang, Wu You-li, et al.. A new guidance law based on information fusion and optimal control of structure stochastic jump system[C]. Proceedings of 2007 IEEE International Conference on Automation and Logistics, Jinan, China, 2007: 624–627.
[3]	Ye Chun-qi, Wang Bao-shu, and Miao Qi-guang. Fusion algorithm of SAR and panchromatic images based on region segmentation in NSCT domain[J]. Systems Engineering and Electronics, 2010, 32(3): 609–613.
[4]	Xu Xing, Li Ying, Sun Jin-qiu, et al. An algorithm for image fusion based on curvelet transform[J]. Journal of Northwestern Polytechnical University, 2008, 26(3): 395–398.
[5]	Shi Zhi, Zhang Zhuo, and Yue Yan-gang. Adaptive image fusion algorithm based on shearlet transform[J]. Acta Photonica Sinica, 2013, 42(1): 115–120. DOI: 10.3788/gzxb
[6]	Liu Jian, Lei Ying-jie, Xing Ya-qiong, et al. Fusion technique for SAR and gray visible image based on hidden Markov model in non-subsample shearlet transform domain[J]. Control and Decision, 2016, 31(3): 453–457.
[7]	Chen Di-peng and Li Qi. The use of complex contourlet transform on fusion scheme[C]. Proceedings of World Academy of Science, Engineering and Technology, Prague, Czech Republic, 2005: 342–347.
[8]	Wu Yi-quan, Wan Hong, and Ye Zhi-long. Fabric defect image noise reduction based on complex contourlet transform and anisotropic diffusion[J]. CAAI Transactions on Intelligent Systems, 2013, 8(3): 214–219.
[9]	Wei Qi, Bioucas-Dias J, Dobigeon N, et al. Hyperspectral and multispectral image fusion based on a sparse representation[J]. IEEE Transactions on Geoscience and Remote Sensing, 2015, 53(7): 3658–3668. DOI: 10.1109/TGRS.2014.2381272
[10]	Yu Nan-nan, Qiu Tian-shuang, Bi Feng, et al. Image features extraction and fusion based on joint sparse representation[J]. IEEE Journal of Selected Topics in Signal Processing, 2011, 5(5): 1074–1082. DOI: 10.1109/JSTSP.2011.2112332
[11]	Wang Jun, Peng Jin-ye, Feng Xiao-yi, et al. Image fusion with nonsubsampled contourlet transform and sparse representation[J]. Journal of Electronic Imaging, 2013, 22(4): 043019. DOI: 10.1117/1.JEI.22.4.043019
[12]	Duarte M F, Sarvotham S, Baron D, et al.. Distributed compressed sensing of jointly sparse signals[C]. Proceedings of Conference Record of the Thirty-Ninth Asilomar Conference on Signals, Systems and Computers Asilomar, Pacific Grove, CA, USA, 2005: 1537–1541.
[13]	Aharon M, Elad M, and Bruckstein A. rmK-SVD: An algorithm for designing overcomplete dictionaries for sparse representation[J]. IEEE Transactions on Signal Processing, 2006, 54(11): 4311–4322. DOI: 10.1109/TSP.2006.881199
[14]	Mallat S G and Zhang Zhi-feng. Matching pursuits with time-frequency dictionaries[J]. IEEE Transactions on Signal Processing, 1993, 41(12): 3397–3415. DOI: 10.1109/78.258082
[15]	Kong Wei-wei and Lei Ying-jie. Technique for image fusion based on NSST domain and human visual characteristics[J]. Journal of Harbin Engineering University, 2013, 34(6): 777–782.
[16]	Fan Xin-nan, Zhang Ji, Li Min, et al. A multi-sensor image fusion algorithm based on local feature difference[J]. Journal of Optoelectronics·Laser, 2014, 25(10): 2025–2032.