基于核方法的高光谱遥感图像解混技术研究
|
摘要
高光谱遥感影像以其光谱分辨率高、成像波段多的优势(0.4μm-2.5μm),丰富了光谱信息同时也提高了光谱解混能力。由于受到遥感器空间分辨力限制和自然界地物的复杂多样性影响,遥感影像中存在大量的混合像元。高光谱解混能利用高光谱遥感提供的上百个波段图像数据,进入像元内部挖掘图像中的亚像元级信息,提高地物分类和目标探测及识别的精度,为微小地物或者异常地物的探测提供了一个研究途径。高光谱遥感混合像元分解方法主要分为两大类：基于线性模型的解混算法和基于非线性模型的解混算法。本论文主要研究基于核方法的高光谱混合像元非线性解混技术及应用。在系统分析传统的光谱解混理论基础上,本文重点研究运用核方法将传统的基于线性光谱混合模型的解混算法推广到非线性特征空间,解决非线性光谱解混问题。
论文的主要工作如下：
(1)正交子空间投影(orthogonal subspace projection, OSP)作为传统的监督解混算法,其突出的特点在于提取端元和丰度时,能消除背景信息的影响,同时也可以抑制图像数据噪声的干扰。该算法是在最小均方误差意义上的一种最优化的压缩投影方法。本文这部分工作主要对其非线性推广的理论进行分析验证。该方法不仅揭示了典型地物光谱之间的高阶特性,同时对噪声影响具有较强的鲁棒性。
(2)针对现有的非负矩阵分解(nonnegative matrix factorization, NMF)算法,本文给出了一种基于核的非负矩阵方法,并就图像本身是否存在纯像元将其分为(pure pixels kernel based nonnegative matrix factorization, pKNMF)算法和(null pixels kernel based nonnegative matrix factorization, npKNMF)算法。该方法对传统的NMF方法进行非线性映射,运用核函数理论,对输入空间数据点积运算进行核转化,解决了寻找具体非线性映射函数的难题。仿真实验结果和实际遥感数据实验均证明了该算法是有效的。
(3)利用高光谱数据端元空间分布的连续性,本文结合端元空间复杂度的思想,提出了基于核方法的空间复杂度非负矩阵分解算法。该方法实现了非线性混合下的像元分解的目的,通过核技巧,对高维特征空间中的映射数据进行高阶特征提取,提高了光谱解混的精度。
Hyperspectral remote sensing images can provide enough information for spectral unmixing, owing to its high spectral resolution and hundreds of spectral channels ranging from 0.4 to 2.5 micrometers. However, the modern spectrometer could not bring us to the same high spatial resolution, so the mixed pixels are widespread in hyperspectral imagery. It can improve the precision of imagery classification and target detection by use of the decomposition of the subpixels in the hundreds of imagery data. It is a good research direction for the detection and analysis of the micro-scale and anomaly objection. There are two main hyperspectral unmixing methods that include linear-based and nonlinear-based algorithms. Recently, many scholars develop kinds of new methods in the area of hyperspectral decomposition. The dissertation mainly studies the kernel-based hyperspectral imagery nonlinear umixing techniques and applications. By systematically analysing the traditional spectral decomposition theory, the thesis focuses on extending the algorithms based on linear spectral model to nonlinear feature space in terms of the kernel functions for resolving the difficulties of hyperspectral nonlinear unmixing problem.The major works and contribution of this dissertation are as follows:
(1) The orthogonal subspcace projection approach is an traditional supervised unmixing method. It can suppress undesired or interfering spectral signatures, and detect the presence of a spectral signature of interest. This operation is an optimal interference suppression process in the least squares sense. In our thesis, we extend it to nonlinear space to exploit the high order features between the spectral data, and get better robust performance.
(2) We extend the traditional nonnegtive matrix factorization method to nonlinear space, and then get a kernel-based nonnegative matrix factorization method that includes pure pixels kernel NMF and null pure pixels kernel NMF. It is a nonlinear version of NMF by using the kernel function to transform the dot product without any knowledge of the nonlinear mapping function. The simulated data and the real hyperspectral imagery experiments show that the methods can provide good performance.
(3) According to the spatial continuity of the spectral abundances, we develop a kernel spatial complexity-based nonnegative matrix factorization by incorporating the spatial complexity into a nonlinear version of NMF method. We realize the nonlinear decomposition of hyperspectral imagery via kernel methods. It can improve the performance of the spectral umixing by exploiting the high order features in the high-dimensional feature space.
论文的主要工作如下：
(1)正交子空间投影(orthogonal subspace projection, OSP)作为传统的监督解混算法,其突出的特点在于提取端元和丰度时,能消除背景信息的影响,同时也可以抑制图像数据噪声的干扰。该算法是在最小均方误差意义上的一种最优化的压缩投影方法。本文这部分工作主要对其非线性推广的理论进行分析验证。该方法不仅揭示了典型地物光谱之间的高阶特性,同时对噪声影响具有较强的鲁棒性。
(2)针对现有的非负矩阵分解(nonnegative matrix factorization, NMF)算法,本文给出了一种基于核的非负矩阵方法,并就图像本身是否存在纯像元将其分为(pure pixels kernel based nonnegative matrix factorization, pKNMF)算法和(null pixels kernel based nonnegative matrix factorization, npKNMF)算法。该方法对传统的NMF方法进行非线性映射,运用核函数理论,对输入空间数据点积运算进行核转化,解决了寻找具体非线性映射函数的难题。仿真实验结果和实际遥感数据实验均证明了该算法是有效的。
(3)利用高光谱数据端元空间分布的连续性,本文结合端元空间复杂度的思想,提出了基于核方法的空间复杂度非负矩阵分解算法。该方法实现了非线性混合下的像元分解的目的,通过核技巧,对高维特征空间中的映射数据进行高阶特征提取,提高了光谱解混的精度。
Hyperspectral remote sensing images can provide enough information for spectral unmixing, owing to its high spectral resolution and hundreds of spectral channels ranging from 0.4 to 2.5 micrometers. However, the modern spectrometer could not bring us to the same high spatial resolution, so the mixed pixels are widespread in hyperspectral imagery. It can improve the precision of imagery classification and target detection by use of the decomposition of the subpixels in the hundreds of imagery data. It is a good research direction for the detection and analysis of the micro-scale and anomaly objection. There are two main hyperspectral unmixing methods that include linear-based and nonlinear-based algorithms. Recently, many scholars develop kinds of new methods in the area of hyperspectral decomposition. The dissertation mainly studies the kernel-based hyperspectral imagery nonlinear umixing techniques and applications. By systematically analysing the traditional spectral decomposition theory, the thesis focuses on extending the algorithms based on linear spectral model to nonlinear feature space in terms of the kernel functions for resolving the difficulties of hyperspectral nonlinear unmixing problem.The major works and contribution of this dissertation are as follows:
(1) The orthogonal subspcace projection approach is an traditional supervised unmixing method. It can suppress undesired or interfering spectral signatures, and detect the presence of a spectral signature of interest. This operation is an optimal interference suppression process in the least squares sense. In our thesis, we extend it to nonlinear space to exploit the high order features between the spectral data, and get better robust performance.
(2) We extend the traditional nonnegtive matrix factorization method to nonlinear space, and then get a kernel-based nonnegative matrix factorization method that includes pure pixels kernel NMF and null pure pixels kernel NMF. It is a nonlinear version of NMF by using the kernel function to transform the dot product without any knowledge of the nonlinear mapping function. The simulated data and the real hyperspectral imagery experiments show that the methods can provide good performance.
(3) According to the spatial continuity of the spectral abundances, we develop a kernel spatial complexity-based nonnegative matrix factorization by incorporating the spatial complexity into a nonlinear version of NMF method. We realize the nonlinear decomposition of hyperspectral imagery via kernel methods. It can improve the performance of the spectral umixing by exploiting the high order features in the high-dimensional feature space.
引文
[1]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.
[2]童庆禧,电子光学遥感系统的过去、现在、和未来[M].北京：科学出版社,1995.
[3]张良培,张立福.高光谱遥感[M].武汉：武汉大学出版社,2005.
[4]李天宏,杨海宏,赵永平.成像光谱仪遥感现状与展望[J].遥感技术与应用,1997,12(2)：54-5.
[5]童庆禧.遥感科学技术进展[J].地理学报,1994,49：616-624.
[6]D. C. Heinz and C.-I. Chang. Fully constrained least squares linear spectral mixture analysis method for material quantification in hyperspectral imagery [J]. IEEE Transactions on Geoscience and Remote Sensing, Mar.2001,39(3):529-545.
[7]Chang C I, Zhao X L, Althouse M L G, and Pan J. J. Least squares subspace projection approach to mixed pixel classification for hyperspectral images [J]. IEEE Transactions on Geoscience and Remote Sensing,1998,36(3):898-912.
[8]P. Paatero. Least squares formulation of robust non-negative factor analysis [C]. Chemometrics and Intelligent Laboratory Systems,1997,37(1):23-35.
[9]Jutten C, Herault J. Independent component analysis versus PCA [J]. Proceeding of European Signal Processing Conference,1988,287-314.
[10]A. Hyvarinen and E. Oja. Independent component analysis:a tutorial. Neural Networks [J]. 2000,13(4):411-430.
[11]A. J. Bell and T. J. Sejnowski. The independenet components of natural scenes are edge filters [J]. Vision Research,1997,37(23):3327-3338.
[12]A. Hyvarinen and P. O. Hoyer. Emergence of phase and shift invariant features by decomposition of natural images into independent feature subspaces [J]. Neural Computation, 2000,12(7):1705-1720.
[13]王峻峰.基于主分量、独立分量分析的盲信号处理及应用研究[D].武汉：华中科技大学,2005.
[14]J. C. Harsanyi and C.-I. Chang. Hyperspectral image classification and dimension reduction: an orthogonal subspace projection approach [J]. IEEE Transations on Geoscience and Remote Sensing, Jul.1994,32(4):779-785.
[15]J. C. Harsanyi, W. H. Farr and C.-I. Chang. Detection of sub-pixel signatures in hyperspectral image sequences [J]. ASPRS/ACSM Annual Convention and Exposition Technical Papers,1994,1:236-247.
[16]J. M. P. Nascimento and J. M. B. Dias. Vertex component analysis:a fast algorithm to unmix hyperspectral data [J]. IEEE Transations on Geoscience and Remote Sensing, April.2005, 43(4):898-910.
[17]P. Paatero and U. Tapper. Positive matrix factorization:a non-negative factor model with optimal utilization of error estimates of data values [J]. Environmetrics,1994,5:111-126.
[18]D. D. Lee and H. S. Seung. Unsupervised learning by convex and conic coding [J]. In Advances in Neural Information Processing System,1997,515-521.
[19]V. P. Pauca, J. Piper, R. J. Plemmons. Nonnegative matrix factorization for spectral data analysis [J]. Linear Algebra and its Applications, Jul.2006,416(1):29-47.
[20]A. P.-Montano, J. M. Carazo, K. Kochi, D. Lehmann and R. D. P.-Marqui. Nonsmooth nonnegative matrix factorization(nsNMF) [J]. IEEE Transactions on Pattern Anaylsis and Machine Intelligence, Mar.2006,28(3):403-415.
[21]S. Jia, Y. T. Qian. Spectral and Spatial Complexity Based Hyperspectral Unmixing [J]. IEEE Transactions on Geoscience and Remote Sensing,2007,45(12):3867-3879.
[22]Heesung Kwon and Nasser M.Nasrabadi. Hyperspectral target detection using kernel spectral matched filter [J]. Proceeding of the 2004 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops,126-127.
[23]Nasser M.Nasrabadi and Heesung Kwon. Kernel-based subpixel target detection in hyperspectral images [J]. Proceedings of IEEE International Joint Conference on Neural Networks,2004:717-721.
[24]http://spectral.cr.usgs.gov/.
[25]The HYDICE image dataset. http://www.tec.army.mil/Hypercube/.
[26]The AVIRIS LCVF image dataset. http://aviris.ipl.nasa.gov/.
[27]Boardman J W. Automated spectral unmixing of AVIRIS data using convex geometry concepts [J]. In Summaries, Fourth JPL Airborne Geoscience Workshop, JPL Publication 93-26,1:11-14.
[28]Lidan Miao and Hairong Qi. Endmember Extraction From Highly Mixed Data Using Minimum Volume Constrained Nonnegative Matrix Factorization [J]. IEEE Transactions on Geoscience and Remote Sensing,2007,45(3):765-777.
[29]Hapke B. Bidirectional reflectance spectroscopy theory [J]. Journal of Geophysical Research, 86(B4):3039-3045.
[30]J. F. Mustard and C. M. Pieters. Photometric phase functions of common geologic minerals and applications to quantitative analysis of mineral mixture reflectance spectra [J]. Journal of Geophysical Research,94(B10):13619-13634.
[31]P. E. Johnson, M. O. Smith, J. B. Adams. Simple algorithms for remote determination of mineral abundances and particle sizes from reflectance spectra [J]. Journal of Geophysical Research,97(B4):2649-2657.
[32]Kerri J. Guifoyle, L. Althouse, Chein-I Chang. A quantitative and comparative analysis of linear and nonlinear spectral mixture models using radial basis function neural networks [J]. IEEE Transations on Geoscience and Remote Sensing, Aul.2001,39(8):2314-2318.
[33]王强Hyperion高光谱数据进行混合像元分解研究[D].哈尔滨：东北林业大学,2006.
[34]C.-I Chang, Q. Du. Exploration of virtual dimensionality in hyperspectral image analysis [J]. In Algorithms and Technologies for Multispectral, Hyperspectral and Ultralspectral Imagery XII,2006.
[35]C.-I Chang, Q. Du. Estimation of number of spectrally distinct signal sources in hyperspectral imagery [J]. IEEE Transations on Geoscience and Remote Sensing, Mar.2004, 42(3):608-619.
[36]R. E. ROGER. Principal Components transform with simple, automatic noise adjustment [J]. Remote Sensing,1996, vol.17(14):2719-2727.
[37]章孝灿,黄智才,赵元洪.遥感数字图像处理[M].浙江：浙江大学出版社.1997.
[38]G. Swayze. The hydrothermal and structural history of the Cuprite Mining District, south-Western Nevada:An integrated geological and geophysical approach Ph.D. dissertation [D]. Univ. Colorado, Boulder,1997.
[39]G. Swayze, R. Clark, S. Sutley, A. Gallagher. Ground-truthing aviris mineral mapping at Cuprite, Nevada [J]. JPL Airborne Geosciences Workshop, vol.1,1992, pp.47-49.
[40]John Shawe-Taylor, Nello Cristianini. Kernel methods for pattern analysis. [M]北京：机械工业出版社,2006.
[41]厉小润.模式识别的核方法研究[D].杭州：浙江大学,2008.
[42]J. J. Settle, On the relationship between spectral unmixing and subspace projection [J]. IEEE Transactions Geoscience, vol,34.1996,4(34):1045-1046.
[43]张凯,赵辽英.基于子空间分析的高光谱图像目标检测技术研究[D].杭州：杭州电子科技大学,2010.
[44]H. Kwon and N.M. Nasrabadi, Kernel orthogonal subspace projection for hyperspectral signal classification [J]. Geoscience and Remote Sensing, IEEE Transactions on, vol.43, no. 12, pp.2952-2962, Dec.2005.
[45]陶雪涛,王斌,张立明.基于NMF的遥感图像混合像元分解新方法[J].信息与电子工程,2008,6(1)：34-39.
[46]A. Plaza, P. Martinez, R. Perez and J.Plaza. A quantitative and comparative analysis of endmember extraction algorithms from hyperspectral data [J]. IEEE Transactions on Geoscience and Remote Sensing,2004,42(3):650-663.
[47]C. J. Lin. Projected gradient methods for non-negative matrix factorization [R]. Technical Report ISSTECH-95-013, Department of Computer Science, National Taiwan University, 2005.
[48]R. Zdunek and A. Cichocki. Non-negative matrix factorization with quasi-newton optimization [C]. Putkowski L et al. Artificial Intelligence and Soft Computing, Heidelberg, Springer,2006,4029:870-879.
[49]W. Liu, N. Zheng and X. Lu. Nonnegative matrix factorization for visualcoding [C]. In Proceedings of the IEEE International Conference on Acoustics, Speech and Signal Processing(ICASSP'03),2003,3:293-296.
[50]刘维湘,郑南宁,游屈波.非负矩阵分解及其在模式识别中的应用[J].科学通报,2006,51(3)：241-250.
[51]贾森.非监督的高光谱图像解混技术研究[D].杭州：浙江大学,2007.
[52]P. O. Hoyer. Non-negative matrix factorization with sparseness constraints [J]. Journal of Machine Learning Research, Nov.2004,5:1457-1469.
[53]J. V. Stone. Blind source separation using temporal predictability. Neural Computation [J]. Jul. 2001,13(7):1559-1574.
[54]S. Amari, A. Cichocki, and H. H. Yang. A new learning algorithm for blind signal separation [J]. In Adances in Neural Information Processing Systems,8,1996, pages 757-763.
[55]吕亚丽,赵辽英.基于非负矩阵分解的高光谱图像解混技术研究[D].杭州：杭州电子科技大学,2010.
[56]M. Grant, S. Boyd, and Y. Ye. CVX:Matlab software for disciplined convex programming. http://www.stanford.edu/-boyd/c vx/.
[57]L. Molgedey and H. G. Schuster. Separation of a mixture of independent signals using time delayed correlations [J]. Physical Review Letters,1994,72(23):3634-3637.
[58]B. A. Pearlmutter and L. C. Parra. A context sensitive generalization of ICA [J]. In International conference on Neural Information Processing,1996, pages 151-157.
[59]A. Belouchrani, K. A. Meraim, J,-F. Cardoso, E. Moulines. A blind source separation technique based on second order statistics [J]. IEEE Transactions on signal processing, Feb. 1997,45(2):434-444.
[60]J. Besag. On the statistical analysis of dirty pictures [J].Journal of the Royal Statistical Society, Series B,1986,48:259-302.
[61]P. Pajunen. Blind source separation using algorithmic information theory [J]. Neurcomputing, 1988,22:35-48.
[62].J. V. Stone. Blind source separation using temporal predictability [J]. Neural Computation, Jul.2001,13(7):1559-1574.
[63]J. V. Stone. Independent Component Analysis. A Tutorial introduction [M]. MIT Press,2004.
[64]J.-F. Cardoso and B. H. Laheld. Equivariant adaptive source separation [J]. IEEE Transactions on Signal Processing.1996,44(12):3017-3030.
[65]T. M. Cover, J. A. Thomas. Elements of Information theory [M]. John Wiley & Sons,1991.
[66]P. J. Green. Bayesian reconstructions from emission tomography data using a modified algorithm [J].IEEE Transactions on Medical Imaging, March,1990,9(1):84-93.
[67]Daoqiang Zhang, Zhihua Zhou, Songcan Chen. Nonnegative matrix factorization on kernels [J]. PRICAI.
[68]吴波,张良培,李平湘.基于支撑向量回归的高光谱混合像元非线性分解[J].遥感学报,2006,10(3)：312-318.
[69]Nascimento, J.M.P., Dias, J.M.B. Vertex Component Analysis:A Fast Algorithm to Unmix Hyperspectral Data [J]. IEEE Transactions on Geoscience and Remote Sensing,2005,43(4): 898-910.
[2]童庆禧,电子光学遥感系统的过去、现在、和未来[M].北京：科学出版社,1995.
[3]张良培,张立福.高光谱遥感[M].武汉：武汉大学出版社,2005.
[4]李天宏,杨海宏,赵永平.成像光谱仪遥感现状与展望[J].遥感技术与应用,1997,12(2)：54-5.
[5]童庆禧.遥感科学技术进展[J].地理学报,1994,49：616-624.
[6]D. C. Heinz and C.-I. Chang. Fully constrained least squares linear spectral mixture analysis method for material quantification in hyperspectral imagery [J]. IEEE Transactions on Geoscience and Remote Sensing, Mar.2001,39(3):529-545.
[7]Chang C I, Zhao X L, Althouse M L G, and Pan J. J. Least squares subspace projection approach to mixed pixel classification for hyperspectral images [J]. IEEE Transactions on Geoscience and Remote Sensing,1998,36(3):898-912.
[8]P. Paatero. Least squares formulation of robust non-negative factor analysis [C]. Chemometrics and Intelligent Laboratory Systems,1997,37(1):23-35.
[9]Jutten C, Herault J. Independent component analysis versus PCA [J]. Proceeding of European Signal Processing Conference,1988,287-314.
[10]A. Hyvarinen and E. Oja. Independent component analysis:a tutorial. Neural Networks [J]. 2000,13(4):411-430.
[11]A. J. Bell and T. J. Sejnowski. The independenet components of natural scenes are edge filters [J]. Vision Research,1997,37(23):3327-3338.
[12]A. Hyvarinen and P. O. Hoyer. Emergence of phase and shift invariant features by decomposition of natural images into independent feature subspaces [J]. Neural Computation, 2000,12(7):1705-1720.
[13]王峻峰.基于主分量、独立分量分析的盲信号处理及应用研究[D].武汉：华中科技大学,2005.
[14]J. C. Harsanyi and C.-I. Chang. Hyperspectral image classification and dimension reduction: an orthogonal subspace projection approach [J]. IEEE Transations on Geoscience and Remote Sensing, Jul.1994,32(4):779-785.
[15]J. C. Harsanyi, W. H. Farr and C.-I. Chang. Detection of sub-pixel signatures in hyperspectral image sequences [J]. ASPRS/ACSM Annual Convention and Exposition Technical Papers,1994,1:236-247.
[16]J. M. P. Nascimento and J. M. B. Dias. Vertex component analysis:a fast algorithm to unmix hyperspectral data [J]. IEEE Transations on Geoscience and Remote Sensing, April.2005, 43(4):898-910.
[17]P. Paatero and U. Tapper. Positive matrix factorization:a non-negative factor model with optimal utilization of error estimates of data values [J]. Environmetrics,1994,5:111-126.
[18]D. D. Lee and H. S. Seung. Unsupervised learning by convex and conic coding [J]. In Advances in Neural Information Processing System,1997,515-521.
[19]V. P. Pauca, J. Piper, R. J. Plemmons. Nonnegative matrix factorization for spectral data analysis [J]. Linear Algebra and its Applications, Jul.2006,416(1):29-47.
[20]A. P.-Montano, J. M. Carazo, K. Kochi, D. Lehmann and R. D. P.-Marqui. Nonsmooth nonnegative matrix factorization(nsNMF) [J]. IEEE Transactions on Pattern Anaylsis and Machine Intelligence, Mar.2006,28(3):403-415.
[21]S. Jia, Y. T. Qian. Spectral and Spatial Complexity Based Hyperspectral Unmixing [J]. IEEE Transactions on Geoscience and Remote Sensing,2007,45(12):3867-3879.
[22]Heesung Kwon and Nasser M.Nasrabadi. Hyperspectral target detection using kernel spectral matched filter [J]. Proceeding of the 2004 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops,126-127.
[23]Nasser M.Nasrabadi and Heesung Kwon. Kernel-based subpixel target detection in hyperspectral images [J]. Proceedings of IEEE International Joint Conference on Neural Networks,2004:717-721.
[24]http://spectral.cr.usgs.gov/.
[25]The HYDICE image dataset. http://www.tec.army.mil/Hypercube/.
[26]The AVIRIS LCVF image dataset. http://aviris.ipl.nasa.gov/.
[27]Boardman J W. Automated spectral unmixing of AVIRIS data using convex geometry concepts [J]. In Summaries, Fourth JPL Airborne Geoscience Workshop, JPL Publication 93-26,1:11-14.
[28]Lidan Miao and Hairong Qi. Endmember Extraction From Highly Mixed Data Using Minimum Volume Constrained Nonnegative Matrix Factorization [J]. IEEE Transactions on Geoscience and Remote Sensing,2007,45(3):765-777.
[29]Hapke B. Bidirectional reflectance spectroscopy theory [J]. Journal of Geophysical Research, 86(B4):3039-3045.
[30]J. F. Mustard and C. M. Pieters. Photometric phase functions of common geologic minerals and applications to quantitative analysis of mineral mixture reflectance spectra [J]. Journal of Geophysical Research,94(B10):13619-13634.
[31]P. E. Johnson, M. O. Smith, J. B. Adams. Simple algorithms for remote determination of mineral abundances and particle sizes from reflectance spectra [J]. Journal of Geophysical Research,97(B4):2649-2657.
[32]Kerri J. Guifoyle, L. Althouse, Chein-I Chang. A quantitative and comparative analysis of linear and nonlinear spectral mixture models using radial basis function neural networks [J]. IEEE Transations on Geoscience and Remote Sensing, Aul.2001,39(8):2314-2318.
[33]王强Hyperion高光谱数据进行混合像元分解研究[D].哈尔滨：东北林业大学,2006.
[34]C.-I Chang, Q. Du. Exploration of virtual dimensionality in hyperspectral image analysis [J]. In Algorithms and Technologies for Multispectral, Hyperspectral and Ultralspectral Imagery XII,2006.
[35]C.-I Chang, Q. Du. Estimation of number of spectrally distinct signal sources in hyperspectral imagery [J]. IEEE Transations on Geoscience and Remote Sensing, Mar.2004, 42(3):608-619.
[36]R. E. ROGER. Principal Components transform with simple, automatic noise adjustment [J]. Remote Sensing,1996, vol.17(14):2719-2727.
[37]章孝灿,黄智才,赵元洪.遥感数字图像处理[M].浙江：浙江大学出版社.1997.
[38]G. Swayze. The hydrothermal and structural history of the Cuprite Mining District, south-Western Nevada:An integrated geological and geophysical approach Ph.D. dissertation [D]. Univ. Colorado, Boulder,1997.
[39]G. Swayze, R. Clark, S. Sutley, A. Gallagher. Ground-truthing aviris mineral mapping at Cuprite, Nevada [J]. JPL Airborne Geosciences Workshop, vol.1,1992, pp.47-49.
[40]John Shawe-Taylor, Nello Cristianini. Kernel methods for pattern analysis. [M]北京：机械工业出版社,2006.
[41]厉小润.模式识别的核方法研究[D].杭州：浙江大学,2008.
[42]J. J. Settle, On the relationship between spectral unmixing and subspace projection [J]. IEEE Transactions Geoscience, vol,34.1996,4(34):1045-1046.
[43]张凯,赵辽英.基于子空间分析的高光谱图像目标检测技术研究[D].杭州：杭州电子科技大学,2010.
[44]H. Kwon and N.M. Nasrabadi, Kernel orthogonal subspace projection for hyperspectral signal classification [J]. Geoscience and Remote Sensing, IEEE Transactions on, vol.43, no. 12, pp.2952-2962, Dec.2005.
[45]陶雪涛,王斌,张立明.基于NMF的遥感图像混合像元分解新方法[J].信息与电子工程,2008,6(1)：34-39.
[46]A. Plaza, P. Martinez, R. Perez and J.Plaza. A quantitative and comparative analysis of endmember extraction algorithms from hyperspectral data [J]. IEEE Transactions on Geoscience and Remote Sensing,2004,42(3):650-663.
[47]C. J. Lin. Projected gradient methods for non-negative matrix factorization [R]. Technical Report ISSTECH-95-013, Department of Computer Science, National Taiwan University, 2005.
[48]R. Zdunek and A. Cichocki. Non-negative matrix factorization with quasi-newton optimization [C]. Putkowski L et al. Artificial Intelligence and Soft Computing, Heidelberg, Springer,2006,4029:870-879.
[49]W. Liu, N. Zheng and X. Lu. Nonnegative matrix factorization for visualcoding [C]. In Proceedings of the IEEE International Conference on Acoustics, Speech and Signal Processing(ICASSP'03),2003,3:293-296.
[50]刘维湘,郑南宁,游屈波.非负矩阵分解及其在模式识别中的应用[J].科学通报,2006,51(3)：241-250.
[51]贾森.非监督的高光谱图像解混技术研究[D].杭州：浙江大学,2007.
[52]P. O. Hoyer. Non-negative matrix factorization with sparseness constraints [J]. Journal of Machine Learning Research, Nov.2004,5:1457-1469.
[53]J. V. Stone. Blind source separation using temporal predictability. Neural Computation [J]. Jul. 2001,13(7):1559-1574.
[54]S. Amari, A. Cichocki, and H. H. Yang. A new learning algorithm for blind signal separation [J]. In Adances in Neural Information Processing Systems,8,1996, pages 757-763.
[55]吕亚丽,赵辽英.基于非负矩阵分解的高光谱图像解混技术研究[D].杭州：杭州电子科技大学,2010.
[56]M. Grant, S. Boyd, and Y. Ye. CVX:Matlab software for disciplined convex programming. http://www.stanford.edu/-boyd/c vx/.
[57]L. Molgedey and H. G. Schuster. Separation of a mixture of independent signals using time delayed correlations [J]. Physical Review Letters,1994,72(23):3634-3637.
[58]B. A. Pearlmutter and L. C. Parra. A context sensitive generalization of ICA [J]. In International conference on Neural Information Processing,1996, pages 151-157.
[59]A. Belouchrani, K. A. Meraim, J,-F. Cardoso, E. Moulines. A blind source separation technique based on second order statistics [J]. IEEE Transactions on signal processing, Feb. 1997,45(2):434-444.
[60]J. Besag. On the statistical analysis of dirty pictures [J].Journal of the Royal Statistical Society, Series B,1986,48:259-302.
[61]P. Pajunen. Blind source separation using algorithmic information theory [J]. Neurcomputing, 1988,22:35-48.
[62].J. V. Stone. Blind source separation using temporal predictability [J]. Neural Computation, Jul.2001,13(7):1559-1574.
[63]J. V. Stone. Independent Component Analysis. A Tutorial introduction [M]. MIT Press,2004.
[64]J.-F. Cardoso and B. H. Laheld. Equivariant adaptive source separation [J]. IEEE Transactions on Signal Processing.1996,44(12):3017-3030.
[65]T. M. Cover, J. A. Thomas. Elements of Information theory [M]. John Wiley & Sons,1991.
[66]P. J. Green. Bayesian reconstructions from emission tomography data using a modified algorithm [J].IEEE Transactions on Medical Imaging, March,1990,9(1):84-93.
[67]Daoqiang Zhang, Zhihua Zhou, Songcan Chen. Nonnegative matrix factorization on kernels [J]. PRICAI.
[68]吴波,张良培,李平湘.基于支撑向量回归的高光谱混合像元非线性分解[J].遥感学报,2006,10(3)：312-318.
[69]Nascimento, J.M.P., Dias, J.M.B. Vertex Component Analysis:A Fast Algorithm to Unmix Hyperspectral Data [J]. IEEE Transactions on Geoscience and Remote Sensing,2005,43(4): 898-910.