Hyperspectral Unmixing Based on Constrained Bilinear or Linear-Quadratic Matrix Factorization

被引:8
作者
Benhalouche, Fatima Zohra [1 ,2 ,3 ]
Deville, Yannick [2 ]
Karoui, Moussa Sofiane [1 ,2 ,3 ]
Ouamri, Abdelaziz [3 ]
机构
[1] Ctr Tech Spati, Agence Spatiale Algerienne, Arzew 31200, Algeria
[2] Univ Toulouse, CNRS, Inst Rech Astrophys & Planetol IRAP, UPS,OMP,CNES, F-31400 Toulouse, France
[3] Univ Sci & Technol dOran Mohamed Boudiaf, Oran 31000, Algeria
关键词
hyperspectral imaging; unsupervised bilinear or linear-quadratic spectral unmixing; endmember spectra extraction; bilinear or linear-quadratic matrix factorization; nonnegativity constraints; ENDMEMBER EXTRACTION; COMPONENT ANALYSIS; MIXTURE ANALYSIS; MIXING MODEL; ALGORITHMS; IMAGES; NMF; QUANTIFICATION;
D O I
10.3390/rs13112132
中图分类号
X [环境科学、安全科学];
学科分类号
08 ; 0830 ;
摘要
Unsupervised hyperspectral unmixing methods aim to extract endmember spectra and infer the proportion of each of these spectra in each observed pixel when considering linear mixtures. However, the interaction between sunlight and the Earth's surface is often very complex, so that observed spectra are then composed of nonlinear mixing terms. This nonlinearity is generally bilinear or linear quadratic. In this work, unsupervised hyperspectral unmixing methods, designed for the bilinear and linear-quadratic mixing models, are proposed. These methods are based on bilinear or linear-quadratic matrix factorization with non-negativity constraints. Two types of algorithms are considered. The first ones only use the projection of the gradient, and are therefore linked to an optimal manual choice of their learning rates, which remains the limitation of these algorithms. The second developed algorithms, which overcome the above drawback, employ multiplicative projective update rules with automatically chosen learning rates. In addition, the endmember proportions estimation, with three alternative approaches, constitutes another contribution of this work. Besides, the reduction of the number of manipulated variables in the optimization processes is also an originality of the proposed methods. Experiments based on realistic synthetic hyperspectral data, generated according to the two considered nonlinear mixing models, and also on two real hyperspectral images, are carried out to evaluate the performance of the proposed approaches. The obtained results show that the best proposed approaches yield a much better performance than various tested literature methods.
引用
收藏
页数:29
相关论文
共 69 条
[41]   Hopfield Neural Network Approach for Supervised Nonlinear Spectral Unmixing [J].
Li, Jing ;
Li, Xiaorun ;
Huang, Bormin ;
Zhao, Liaoying .
IEEE GEOSCIENCE AND REMOTE SENSING LETTERS, 2016, 13 (07) :1002-1006
[42]   Blind nonlinear hyperspectral unmixing based on constrained kernel nonnegative matrix factorization [J].
Li, Xiaorun ;
Cui, Jiantao ;
Zhao, Liaoying .
SIGNAL IMAGE AND VIDEO PROCESSING, 2014, 8 (08) :1555-1567
[43]   Kernel-Based Nonlinear Spectral Unmixing with Dictionary Pruning [J].
Li, Zeng ;
Chen, Jie ;
Rahardja, Susanto .
REMOTE SENSING, 2019, 11 (05)
[44]   Pixel Unmixing in Hyperspectral Data by Means of Neural Networks [J].
Licciardi, Giorgio A. ;
Del Frate, Fabio .
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2011, 49 (11) :4163-4172
[45]   Manifold Regularized Sparse NMF for Hyperspectral Unmixing [J].
Lu, Xiaoqiang ;
Wu, Hao ;
Yuan, Yuan ;
Yan, Pingkun ;
Li, Xuelong .
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2013, 51 (05) :2815-2826
[46]   A New Algorithm for Bilinear Spectral Unmixing of Hyperspectral Images Using Particle Swarm Optimization [J].
Luo, Wenfei ;
Gao, Lianru ;
Plaza, Antonio ;
Marinoni, Andrea ;
Yang, Bin ;
Zhong, Liang ;
Gamba, Paolo ;
Zhang, Bing .
IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING, 2016, 9 (12) :5776-5790
[47]   Nonlinear Hyperspectral Unmixing Using Nonlinearity Order Estimation and Polytope Decomposition [J].
Marinoni, Andrea ;
Plaza, Javier ;
Plaza, Antonio ;
Gamba, Paolo .
IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING, 2015, 8 (06) :2644-2654
[48]   Linear-Quadratic Blind Source Separation Using NMF to Unmix Urban Hyperspectral Images [J].
Meganem, Ines ;
Deville, Yannick ;
Hosseini, Shahram ;
Deliot, Philippe ;
Briottet, Xavier .
IEEE TRANSACTIONS ON SIGNAL PROCESSING, 2014, 62 (07) :1822-1833
[49]   Linear-Quadratic Mixing Model for Reflectances in Urban Environments [J].
Meganem, Ines ;
Deliot, Philippe ;
Briottet, Xavier ;
Deville, Yannick ;
Hosseini, Shahram .
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2014, 52 (01) :544-558
[50]   Endmember extraction from highly mixed data using minimum volume constrained nonnegative matrix factorization [J].
Miao, Lidan ;
Qi, Hairong .
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2007, 45 (03) :765-777