基于局部约束群稀疏的红外图像超分辨率重建

邓承志; 田伟; 陈盼; 汪胜前; 朱华生; 胡赛凤

doi:10.7498/aps.63.044202

摘要

针对红外图像分辨率低、视觉质量差等问题，提出基于局部约束群稀疏模型的红外图像超分辨率重建方法. 考虑到红外图像的纹理自相似性和原子系数的群结构稀疏性，首先建立了基于局部约束的群稀疏表示模型. 然后，在假定低分辨率图像空间和高分辨率图像空间具有相似流形的前提下，联合局部约束群稀疏表示模型和K-SVD（K奇异值分解）方法，训练得到高低分辨率图像对应的群结构字典对. 最后，通过高分辨字典和对应的红外图像群稀疏表示系数重建得到高分辨率的红外图像. 实验结果表明，本文方法具有更好的超分辨率效果，无论是在客观评价指标还是主观视觉效果方面都有明显的提高.

关键词:

Abstract

Aiming at the problems of low-resolution and poor visual quality of infrared images, a locality-constrained group sparsity based infrared image super-resolution algorithm is proposed. Firstly with considering the texture self-similarity of infrared images and group structural sparsity of atom coefficients, a locality-constrained group sparse (LCGS) model is proposed. Secondly, under LCGS and K-singular value decomposition, a pair of group structural dictionaries is learned. The dictionary pair can well capture and preserve the intrinsic geometrical manifold of low and high resolution data. Finally, the high-resolution infrared images are recovered by the high-resolution dictionary and the corresponding low-resolution group sparse coefficients. Experimental results show that the proposed method obtains excellent performance in objective evaluation and subjective visual effect.

Keywords:

作者及机构信息

1.
南昌工程学院信息工程学院, 南昌 330099;

2.
江西科技师范大学通信与电子学院, 南昌 330013

基金项目: 国家自然科学基金（批准号：61162022, 61362036）、江西省自然科学基金（批准号：20132BAB201021）、江西省科技落地计划（批准号：KJLD12098）和江西省教育厅科技项目（批准号：GJJ12632）资助的课题.

Authors and contacts

1.
Department of Information Engineering, Nanchang Institute of Technology, Nanchang 330099, China;

2.
Department of Communication and Electronic, Jiangxi Science and Technology Normal University, Nanchang 330013, China

Funds: Project supported by the National Natural Science Foundation of China (Grant Nos. 61162022, 61362036), the Natural Science Foundation of Jiangxi Province, China (Grant No. 20132BAB201021), the Jiangxi Science and Technology Research Development Project, China (Grant No. KJLD12098), and the Jiangxi Science and Technology Research Project of Education Department, China (Grant No. GJJ12632).

参考文献

[1]	Kang M, Chaudhuri S 2003 IEEE Signal Proc. Mag. 20 1920
[2]	Farsiu S, Robinson D, Elad M, Milanfar P 2004 Int. J. Imag. Syst. Tech. 14 47
[3]	Elad M, Datsenko D 2007 Comput. J. 50 1
[4]	Elad M, Figueiredo M A T, Ma Y 2010 P. IEEE 98 972
[5]	Zhao J J, Tang Z Y, Yang J, Liu E Q, Zhou Y 2011 J. Infrared Millim. Waves 30 156 (in Chinese) [赵佳佳, 唐峥远, 杨杰, 刘尔琦, 周越 2011 红外与毫米波学报 30 156]
[6]	Song C X, Ma K, Qin C, Xiao P 2013 Acta Phys. Sin. 62 040702 (in Chinese) [宋长新, 马克, 秦川, 肖鹏 2013 62 040702]
[7]	Ning F L, He B J, Wei J 2013 Acta Phys. Sin. 62 174212 (in Chinese) [宁方立, 何碧静, 韦娟 2013 62 174212]
[8]	Yang J, Wright J, Huang T 2010 IEEE Trans. Image Process. 19 2861
[9]	Sun Y B, Wei Z H, Xiao L, Zhang Z R 2010 Acta Electron. Sin. 38 2898 (in Chinese) [孙玉宝, 韦志辉, 肖亮, 张峥嵘 2010 电子学报 38 2898]
[10]	Lian Q S, Zhang W 2012 Acta Electron. Sin. 40 920 (in Chinese) [练秋生, 张伟 2012 电子学报 40 920]
[11]	Yang S Y, Wang M, Chen Y G, et al 2012 IEEE Trans. Image Process. 21 4016
[12]	Tang Y, Yuan Y, Yan P K, Li X L 2013 J. Vis. Commun. Image R 24 148
[13]	Huang J Z, Zhang T, Metaxas D 2011 J. Mach. Learn. Res. 12 3371
[14]	Sun H, Zhang Z L, Yu L 2012 Signal Process. 28 759 (in Chinese) [孙洪, 张智林, 余磊 2012 信号处理 28 759]
[15]	Huang J, Zhang T 2010 Ann. Stat. 38 1978
[16]	Liu H C, Li S T, Yin H T 2013 Opt. Commun. 289 45
[17]	Chang H, Yeung D, Xiong Y 2004 Proceedings of IEEE Conference on Computer Vision and Pattern Recognition Washington DC, USA, June 27-July 2, 2004 p275
[18]	Wang J J, Yang J C, Yu K, L F J, Huang T, Gong Y H 2010 Proceedings of IEEE Conference on Computer Vision and Pattern Recognition San Francisco, USA, June 13-18, 2010 p3360
[19]	Majumda A, Ward R K 2009 Can. J. Elect. Comput. E 34 136
[20]	Yang J, Zhang Y 2011 SIAM J. Sci. Comput. 33 250
[21]	Xu Y, Zhong A N, Yang J, Zhang D 2011 Opt. Eng. 50 037202
[22]	Brunet D, Vrscay E R, Wang Z 2012 IEEE Trans. Image Process. 21 1488

施引文献

[1]	Kang M, Chaudhuri S 2003 IEEE Signal Proc. Mag. 20 1920
[2]	Farsiu S, Robinson D, Elad M, Milanfar P 2004 Int. J. Imag. Syst. Tech. 14 47
[3]	Elad M, Datsenko D 2007 Comput. J. 50 1
[4]	Elad M, Figueiredo M A T, Ma Y 2010 P. IEEE 98 972
[5]	Zhao J J, Tang Z Y, Yang J, Liu E Q, Zhou Y 2011 J. Infrared Millim. Waves 30 156 (in Chinese) [赵佳佳, 唐峥远, 杨杰, 刘尔琦, 周越 2011 红外与毫米波学报 30 156]
[6]	Song C X, Ma K, Qin C, Xiao P 2013 Acta Phys. Sin. 62 040702 (in Chinese) [宋长新, 马克, 秦川, 肖鹏 2013 62 040702]
[7]	Ning F L, He B J, Wei J 2013 Acta Phys. Sin. 62 174212 (in Chinese) [宁方立, 何碧静, 韦娟 2013 62 174212]
[8]	Yang J, Wright J, Huang T 2010 IEEE Trans. Image Process. 19 2861
[9]	Sun Y B, Wei Z H, Xiao L, Zhang Z R 2010 Acta Electron. Sin. 38 2898 (in Chinese) [孙玉宝, 韦志辉, 肖亮, 张峥嵘 2010 电子学报 38 2898]
[10]	Lian Q S, Zhang W 2012 Acta Electron. Sin. 40 920 (in Chinese) [练秋生, 张伟 2012 电子学报 40 920]
[11]	Yang S Y, Wang M, Chen Y G, et al 2012 IEEE Trans. Image Process. 21 4016
[12]	Tang Y, Yuan Y, Yan P K, Li X L 2013 J. Vis. Commun. Image R 24 148
[13]	Huang J Z, Zhang T, Metaxas D 2011 J. Mach. Learn. Res. 12 3371
[14]	Sun H, Zhang Z L, Yu L 2012 Signal Process. 28 759 (in Chinese) [孙洪, 张智林, 余磊 2012 信号处理 28 759]
[15]	Huang J, Zhang T 2010 Ann. Stat. 38 1978
[16]	Liu H C, Li S T, Yin H T 2013 Opt. Commun. 289 45
[17]	Chang H, Yeung D, Xiong Y 2004 Proceedings of IEEE Conference on Computer Vision and Pattern Recognition Washington DC, USA, June 27-July 2, 2004 p275
[18]	Wang J J, Yang J C, Yu K, L F J, Huang T, Gong Y H 2010 Proceedings of IEEE Conference on Computer Vision and Pattern Recognition San Francisco, USA, June 13-18, 2010 p3360
[19]	Majumda A, Ward R K 2009 Can. J. Elect. Comput. E 34 136
[20]	Yang J, Zhang Y 2011 SIAM J. Sci. Comput. 33 250
[21]	Xu Y, Zhong A N, Yang J, Zhang D 2011 Opt. Eng. 50 037202
[22]	Brunet D, Vrscay E R, Wang Z 2012 IEEE Trans. Image Process. 21 1488

[1]	相萌, 何飘, 王天宇, 袁琳, 邓凯, 刘飞, 邵晓鹏. 计算偏振彩色傅里叶叠层成像: 散射光场偏振特性的复用技术. , 2024, 73(12): 124202. doi: 10.7498/aps.73.20240268
[2]	贺芷椰, 张彦东, 唐春华, 李军利, 李四维, 于斌. 中继透镜分辨率在像素编码曝光成像中对图像重构质量的影响分析. , 2023, 72(2): 024201. doi: 10.7498/aps.72.20221588
[3]	隋怡晖, 郭星奕, 郁钧瑾, Alexander A. Solovev, 他得安, 许凯亮. 生成对抗网络加速超分辨率超声定位显微成像方法研究. , 2022, 71(22): 224301. doi: 10.7498/aps.71.20220954
[4]	郁钧瑾, 郭星奕, 隋怡晖, 宋剑平, 他得安, 梅永丰, 许凯亮. 超分辨率超快超声脊髓微血管成像方法. , 2022, 71(17): 174302. doi: 10.7498/aps.71.20220629
[5]	曾祥昱, 王薇, 刘诚, 单昌功, 谢宇, 胡启后, 孙友文, PolyakovAlexander Viktorovich. 利用地基高分辨率傅里叶变换红外光谱技术探测大气氟氯烃气体CCl₂F₂的时空变化特征. , 2021, 70(20): 200201. doi: 10.7498/aps.70.20210640
[6]	夏正德, 宋娜, 刘宾, 潘晋孝, 闫文敏, 邵子惠. 基于字典学习的稠密光场重建算法. , 2020, 69(6): 064201. doi: 10.7498/aps.69.20191621
[7]	潮兴兵, 潘鲁平, 王子圣, 杨锋涛, 丁剑平. 图像传感器像素化效应对菲涅耳非相干关联全息分辨率的影响. , 2019, 68(6): 064203. doi: 10.7498/aps.68.20181844
[8]	王书, 任益充, 饶瑞中, 苗锡奎. 大气损耗对量子干涉雷达的影响机理. , 2017, 66(15): 150301. doi: 10.7498/aps.66.150301
[9]	龚志双, 王秉中, 王任, 臧锐, 王晓华. 基于光栅结构的远场时间反演亚波长源成像. , 2017, 66(4): 044101. doi: 10.7498/aps.66.044101
[10]	田园, 孙友文, 谢品华, 刘诚, 刘文清, 刘建国, 李昂, 胡仁志, 王薇, 曾议. 地基高分辨率傅里叶变换红外光谱反演环境大气中的CH4浓度变化. , 2015, 64(7): 070704. doi: 10.7498/aps.64.070704
[11]	何林阳, 刘晶红, 李刚. 基于多相组重建的航空图像超分辨率算法. , 2015, 64(11): 114208. doi: 10.7498/aps.64.114208
[12]	范虹, 朱艳春, 王芳梅, 张旭梅. 多分辨率水平集算法的乳腺MR图像分割. , 2014, 63(11): 118701. doi: 10.7498/aps.63.118701
[13]	孙增国. 高分辨率合成孔径雷达图像的Gamma分布下最大后验概率降斑算法. , 2013, 62(18): 180701. doi: 10.7498/aps.62.180701
[14]	宋长新, 马克, 秦川, 肖鹏. 结合稀疏编码和空间约束的红外图像聚类分割研究. , 2013, 62(4): 040702. doi: 10.7498/aps.62.040702
[15]	梁木生, 王秉中, 章志敏, 丁帅, 臧锐. 基于远场时间反演的亚波长天线阵列研究. , 2013, 62(5): 058401. doi: 10.7498/aps.62.058401
[16]	周树波, 袁艳, 苏丽娟. 基于双阈值Huber范数估计的图像正则化超分辨率算法. , 2013, 62(20): 200701. doi: 10.7498/aps.62.200701
[17]	陈英明, 王秉中, 葛广顶. 微波时间反演系统的空间超分辨率机理. , 2012, 61(2): 024101. doi: 10.7498/aps.61.024101
[18]	孙增国, 韩崇昭. 基于拖尾分布的高分辨率合成孔径雷达图像建模. , 2010, 59(2): 998-1008. doi: 10.7498/aps.59.998
[19]	葛广顶, 王秉中, 黄海燕, 郑罡. 时间反演电磁波超分辨率特性. , 2009, 58(12): 8249-8253. doi: 10.7498/aps.58.8249
[20]	张海涛, 巩马理, 赵达尊, 闫平, 崔瑞祯, 贾维溥. 实现超分辨率的微变焦法. , 2001, 50(8): 1486-1491. doi: 10.7498/aps.50.1486

计量

文章访问数: 7216
PDF下载量: 1036
被引次数: 0

姓名
邮箱
手机号码
标题
留言内容
验证码

搜索

留言板