Minimum Dispersion Constrained Nonnegative Matrix Factorization to Unmix Hyperspectral Data

Autor: Alexis Huck, Jacques Blanc-Talon, Mireille Guillaume
Přispěvatelé: Institut FRESNEL (FRESNEL), Aix Marseille Université (AMU)-École Centrale de Marseille (ECM)-Centre National de la Recherche Scientifique (CNRS), Centre National de la Recherche Scientifique (CNRS)-École Centrale de Marseille (ECM)-Aix Marseille Université (AMU), Mission pour la Recherche et l'Innovation Scientifique (DGA/MRIS), Ministère de la défense, DGA : délégation générale de l'armement
Jazyk: angličtina
Rok vydání: 2010
Předmět:
Endmember
Projected Gradient
0211 other engineering and technologies
02 engineering and technology
Blind signal separation
Non-negative matrix factorization
Matrix decomposition
Statistics::Machine Learning
[INFO.INFO-TS]Computer Science [cs]/Signal and Image Processing
I.5.4 Signal Processing
0202 electrical engineering
electronic engineering
information engineering

Source separation
Non-negative Matrix Factorization (NMF)
Electrical and Electronic Engineering
021101 geological & geomatics engineering
Mathematics
Pixel
business.industry
Hyperspectral imaging
Pattern recognition
Hyperspectral unmixing
regularization
Computer Science::Computer Vision and Pattern Recognition
Blind source separation
General Earth and Planetary Sciences
A priori and a posteriori
020201 artificial intelligence & image processing
Artificial intelligence
business
Algorithm
[SPI.SIGNAL]Engineering Sciences [physics]/Signal and Image processing
Spectral unmixing
Regularization function
Zdroj: IEEE Transactions on Geoscience and Remote Sensing
IEEE Transactions on Geoscience and Remote Sensing, 2010, PP (99), pp.1-13. ⟨10.1109/TGRS.2009.2038483⟩
IEEE Transactions on Geoscience and Remote Sensing, Institute of Electrical and Electronics Engineers, 2010, PP (99), pp.1-13. ⟨10.1109/TGRS.2009.2038483⟩
IEEE Transactions on Geoscience and Remote Sensing, Institute of Electrical and Electronics Engineers, 2010, 48 (6), pp.2590-2600. ⟨10.1109/TGRS.2009.2038483⟩
IEEE Transactions on Geoscience and Remote Sensing, 2010, 48 (6), pp.2590-2600. ⟨10.1109/TGRS.2009.2038483⟩
ISSN: 0196-2892
DOI: 10.1109/TGRS.2009.2038483⟩
Popis: International audience; This paper considers the problem of unsupervised spectral unmixing for hyperspectral image analysis. Each observed pixel is assumed to be a noisy linear mixture of pure material spectra, namely endmembers. The mixing coefficients, usually called abundances, are constrained to positive and summed to unity. The proposed unmixing approach is based on the Non-negative Matrix Factorization (NMF) framework, which considers the physical constraints of the problem, including the positivity of the endmember spectra and abundances. However, the basic NMF formulation has degenerated solutions and suffers from non-convexity limitations.We consider here a regularization function, called dispersion, which favours the solution such thatthe endmember spectra have minimum variances. Such a solution encourages the recovered spectra to be flat, preserving thepossible spectral singularities (peaks and sharp variations). The regularized criterion is minimized with a Projected Gradient (PG) scheme and we propose a new step-size estimation technique to fasten the PG convergence. The derived algorithm is called MiniDisCo, for Minimum Dispersion Constrained NMF. We experimentally compare MiniDisCo with recently proposed algorithms. It is shown to be particularly robust to the presence of flat spectra, to a possible a priori overestimation of the number of endmembers or if the amount of observed spectral pixels is low. In addition, experiments show that the considered regularization correctly overcomes the degeneracy and nonconvexity problems, leading to satisfactory unmixing accuracy. We include a comparative analysis of a real world scene.
Databáze: OpenAIRE