Please wait a minute...
img

官方微信

遥感技术与应用  2013, Vol. 28 Issue (6): 1013-1019    
图像与数据处理     
基于训练字典的压缩感知光谱稀疏化方法
汪琪1,2,李传荣1,马灵玲1,唐伶俐1,李剑剑1,2
(1.中国科学院光电研究院定量遥感信息重点实验室,北京 100094;2.中国科学院大学,北京 100049)
Compressive Sensing Spectral Sparsification Method based on Training Dictionary
Wang Qi1,2,Li Chuanrong1,Ma Lingling1,Tang Lingli1,Li Jianjian1,2
(1.Key Laboratory of Quantitative Remote Sensing Information Technology,Academy of Opto\|Electronics,Chinese Academy of Sciences,Beijing 100094,China;
2.University of Chinese Academy of Sciences,Beijing 100049,China)
 全文: PDF(2811 KB)  
摘要:

压缩感知理论利用目标的稀疏特性,能从极少的测量值中重构出目标图像,已成为突破奈奎斯特采样定理,实现超分辨成像的一个极具潜力的研究方向,其应用于对地观测遥感成像的一个核心问题在于面对复杂的地物场景,如何探求有效的稀疏化表达方法。对于具有超高数据量的高光谱成像而言,充分利用波段间丰富的冗余光谱信息,研究有效的光谱稀疏化表达方法更加具有实用价值。首先介绍了压缩感知光谱成像以及光谱稀疏化表达的基本原理,然后利用来自ASTER光谱库的多种类型地物光谱数据构建了一种基于K-SVD方法的训练字典,将其与DCT基、小波基分别作为稀疏基,对于几种典型地物目标进行仿真重构,结果表明:所构建的稀疏字典在采样数较少的情况下明显优于DCT基和小波基,在20%的低采样率时即可近乎完美地重构光谱曲线。

关键词: 压缩感知高光谱稀疏基字典训练    
Abstract:

Using object’s sparsity,compressive sensing theory is able to breakthrough Nyquist sampling theorem and reconstruct object image from very few observations,which has become a very potential research field to achieve super-resolution image.One of the crucial problems of applying compressive sensing to earth observation remote sensing image is how to pursue effective sparsification method under the complexity of ground scene.It will be more practical for hyperspectral imaging with a large amount of data to search for effective spectral sparsification method using rich redundant spectral information in bands.This paper introduces the basic principle of compressive sensing spectral imaging and spectral sparsification,and trains a redundant dictionary based on K\|SVD method using multiple types of ground objects’ spectral data from ASTER spectral library,reconstructs several typical objects using DCT basis,wavelet basis and training dictionary via simulation experiment and compares the reconstruction results.The result shows using training dictionary as sparse basis performs is better at low sampling rate than DCT basis and wavelet basis,spectral curves can be reconstructed almost perfectly at sampling rate as low as 20%.Our research indicates the great potential for sparse dictionary applied in compressive sensing spectral imaging.

Key words: Compressive sensing    Hyperspectral imaging    Sparse basis    Dictionary training
收稿日期: 2012-11-29 出版日期: 2014-02-28
:  TP 79  
基金资助:

中国科学院光电研究院创新项目“压缩感知遥感关联成像机理与应用研究”(Y12401A01Y)。

通讯作者: 马灵玲(1982-),女,安徽宿州人,博士,副研究员,主要从事遥感信号与信息处理。Email:llma@aoe.ac.cn。    
作者简介: 汪琪(1990-),男,安徽潜山人,硕士研究生,主要从事压缩感知关联成像方面的研究。Email:wangqi10@mails.gucas.ac.cn。
服务  
把本文推荐给朋友
加入引用管理器
E-mail Alert
RSS
作者相关文章  
汪琪
李传荣
马灵玲
唐伶俐
李剑剑

引用本文:

汪琪,李传荣,马灵玲,唐伶俐,李剑剑. 基于训练字典的压缩感知光谱稀疏化方法[J]. 遥感技术与应用, 2013, 28(6): 1013-1019.

Wang Qi,Li Chuanrong,Ma Lingling,Tang Lingli,Li Jianjian. Compressive Sensing Spectral Sparsification Method based on Training Dictionary. Remote Sensing Technology and Application, 2013, 28(6): 1013-1019.

链接本文:

http://www.rsta.ac.cn/CN/        http://www.rsta.ac.cn/CN/Y2013/V28/I6/1013

[1]Xiang Libin,Wang Zhonghou,Liu Xuebin,et al.Spatially Modulated Fourier Transform Hyperspectral Imager for HJ-1A Satellite[J].Spacecraft Engineering,2009,(6):43-49.[ 相里斌,王忠厚,刘学斌,等.环境减灾-1A卫星空间调制型干涉光谱成像仪技术[J].航天器工程,2009,(6):43-49.]

[2]Chrien T G,Green R O,Eastwood M L.Accuracy of the Spectral and Radiometric Laboratory Calibration of the Airborne Visible/Infrared Imaging Spectrometer(AVIRIS)[C]//Proceedings of the International Society for Optical Engineering,Imaging Spectroscopy of the Terrestrial Environment,1990,37-49.[3]Tong Qingxi,Zhang Bing,Zheng Lanfen.Hyperspectral Remote Sensing[M].Beijing:Higher Education Press,2006.[童庆禧,张兵,郑兰芬.高光谱遥感——原理、技术与应用[M].北京:高等教育出版社,2006.]

[4]Donoho D L.Compressed Sensing[J].IEEE Transactions on Information Theory,2006,52(4):1289-1306.

[5]Candes E J,Wakin M B.An Introduction to Compressive Sampling[J].IEEE Signal Processing Magazine,2008,25(2):21-30.

[6]Han Shensheng.Intensity Correlation Imaging Technology for Remote Sensing[J].Spacecraft Recovery and Remote Sensing,2011,(5):44-51.[韩申生.强度关联遥感成像技术[J].航天返回与遥感,2011,(5):44-51.]

[7]Duarte M F,Davenport M A,Takbar D,et al.Single-pixel Imaging via Compressive Sampling:Building Simpler,Smaller,and Less-expensive Digital Cameras[J].IEEE Signal Processing Magazine,2008,25(2):83-91.

[8]Wagadarikar A,John R,Willett R,et al.Single Disperser Design for Coded Aperture Snapshot Spectral Imaging[J].Applied Optics,2008,47(10):B44-B51.

[9]Russell T A,McMackin L,Bridge B,et al.Compressive Hyperspectral Sensor for LWIR Gas Detection[C]//Proceedings of the International Society for Optical Engineering,Compressive Sensing,2012.

[10]Lu Minghai,Shen Xia,Han Shensheng.Ghost Imaging via Compressive Sampling based on Digital Micromirror Device[J].Acta Optica Sinica,2011,(7):105-110.[陆明海,沈夏,韩申生.基于数字微镜器件的压缩感知关联成像研究[J].光学学报,2011,(7):105-110.]

[11]Candes E J,Romberg J,Tao T.Robust Uncertainty Principles:Exact Signal Reconstruction from Highly Incomplete Frequency Information[J].IEEE Transactions on Information Theory,2006,52(2):489-509.

[12]Li Chengbo.An Effcient Algorithm for Total Variation Regularization with Applications to the Single Pixel Camera and Compressive Sensing[D].Rice University,Houston,2009.

[13]Li Xiaobo.Research on Measurement Matrix based on Compressive Sensing[D].Beijing:Beijing Jiaotong University,2010.[李小波.基于压缩感知的测量矩阵研究[D].北京:北京交通大学,2010.]

[14]Zhou Canmei.Research on Siginal Reconstruction Algorithms based on Compressive Sensing[D].Beijing:Beijing Jiaotong University,2010.[周灿梅.基于压缩感知的信号重建算法研究[D].北京:北京交通大学,2010.]

[15]Mallat S G,Zhang Z.Matching Pursuits with Time-frequency Dictionaries[J].IEEE Transactions on Signal Processing,1993,41(12):3397-3415.

[16]Aharon M,Elad M,Bruckstein A.K-SVD:An Algorithm for Designing Overcomplete Dictionaries for Sparse Representation[J].IEEE Transactions on Signal Processing,2006,54(11):4311-4322.

[17]Yu F,Xi J,Zhao L,et al.Sparse Presentation of Underdetermined Blind Source Separation based on Compressed Sensing and K-SVD[J].Journal of Southeast University.Natural Science Edition,2011,41(6):1127-1131.

[19]Li Min,Cheng Jian,Le Xiang,et al.Super-resolution based on Sparse Dictionary Coding[J].Journal of Software,2012,(5):1315-1324.[李民,程建,乐翔,等.稀疏字典编码的超分辨率重建[J].软件学报,2012,(5):1315-1324.]

[20]Gesen Z,Shuhong J,Xiaoli X,et al.Compressed Sensing and Reconstruction with Bernoulli Matrices[J].IEEE International Conference on Information and Automation(ICIA),2010:455-460.

[21]Mohimani G H,Babaie-Zadeh M,Jutten C.Fast Sparse Representation based on Smoothed L0 Norm[C]//Proceedings of the 7th International Conference on Independent Component Analysis and Signal Separation,London,UK,2007:389-396.

[1] 陈伟民,张凌,宋冬梅,王斌,丁亚雄,许明明,崔建勇. 基于AdaBoost改进随机森林的高光谱图像地物分类方法研究[J]. 遥感技术与应用, 2018, 33(4): 612-620.
[2] 苏阳,祁元,王建华,徐菲楠,张金龙. 基于航空高光谱影像的额济纳绿洲土地覆被提取[J]. 遥感技术与应用, 2018, 33(2): 202-211.
[3] 秦振涛,杨茹,张靖,杨武年. 基于聚类结构自适应稀疏表示的高光谱遥感图像修复研究[J]. 遥感技术与应用, 2018, 33(2): 212-215.
[4] 郭宇柏,卓莉,陶海燕,曹晶晶,王芳. 基于空谱初始化的非负矩阵光谱混合像元盲分解[J]. 遥感技术与应用, 2018, 33(2): 216-226.
[5] 刘爱林,郭宝平,李岩山 . 基于离散粒子群算法的凸多模态高光谱图像端元提取研究[J]. 遥感技术与应用, 2018, 33(2): 227-232.
[6] 吴兴,张霞,孙雪剑,张立福,戚文超. SPARK卫星高光谱数据辐射质量评价[J]. 遥感技术与应用, 2018, 33(2): 233-240.
[7] 宋婷婷,付秀丽,陈玉,魏永明,王钦军,程先锋. 云南个旧矿区土壤锌污染遥感反演研究[J]. 遥感技术与应用, 2018, 33(1): 88-95.
[8] 周子勇. 运用经验模态分解和压缩感知方法进行遥感影像超分辨率重建[J]. 遥感技术与应用, 2018, 33(1): 96-102.
[9] 刘慧珺,苏红军,赵-波. 基于改进萤火虫算法的高光谱遥感多特征优化方法[J]. 遥感技术与应用, 2018, 33(1): 110-118.
[10] 李伟娜,韦玮,张怀清,刘华. 基于多角度高光谱数据的高寒沼泽湿地植被生物量估算[J]. 遥感技术与应用, 2017, 32(5): 809-817.
[11] 肖昊,王杰. 基于IDL和MATLAB混合编程的两种光谱混合分析方法比较[J]. 遥感技术与应用, 2017, 32(5): 858-865.
[12] 唐超,邵龙义. 高光谱遥感地物目标识别算法及其在岩性特征提取中的应用[J]. 遥感技术与应用, 2017, 32(4): 691-697.
[13] 李恒凯,欧彬,刘雨婷,邱玉宝. 基于混合像元分解的高光谱影像柑橘识别方法[J]. 遥感技术与应用, 2017, 32(4): 743-750.
[14] 苏红军,赵波. 基于共形几何代数的高光谱遥感波段选择方法[J]. 遥感技术与应用, 2017, 32(3): 539-545.
[15] 史飞飞,高小红,杨灵玉,何林华,贾伟. 基于HJ-1A高光谱遥感数据的湟水流域典型农作物分类研究[J]. 遥感技术与应用, 2017, 32(2): 206-217.