Unmixing of Hyperspectral Images using Bayesian Non-negative Matrix Factorization with Volume Prior

Authors:
Morten Arngren;Mikkel N. Schmidt;Jan Larsen
Affiliations:
DTU Informatics, Technical University of Denmark, Lyngby, Denmark 2800 and FOSS Analytical A/S, Hillerød, Denmark 3400;DTU Informatics, Technical University of Denmark, Lyngby, Denmark 2800;DTU Informatics, Technical University of Denmark, Lyngby, Denmark 2800
Venue:
Journal of Signal Processing Systems
Year:
2011

Citing 9
Cited 1

Nonnegative matrix factorization with Gaussian process priors

Computational Intelligence and Neuroscience - Advances in Nonnegative Matrix and Tensor Factorization
Unsupervised unmixing of hyperspectral imagery using the constrained positive matrix factorization

Unsupervised unmixing of hyperspectral imagery using the constrained positive matrix factorization
Minimum Determinant Constraint for Non-negative Matrix Factorization

ICA '09 Proceedings of the 8th International Conference on Independent Component Analysis and Signal Separation
Bayesian Non-negative Matrix Factorization

ICA '09 Proceedings of the 8th International Conference on Independent Component Analysis and Signal Separation
Bayesian separation of spectral sources under non-negativity and full additivity constraints

Signal Processing
Joint Bayesian endmember extraction and linear unmixing for hyperspectral imagery

IEEE Transactions on Signal Processing
A convex analysis-based minimum-volume enclosing simplex algorithm for hyperspectral unmixing

IEEE Transactions on Signal Processing
Unmixing of Hyperspectral Images using Bayesian Non-negative Matrix Factorization with Volume Prior

Journal of Signal Processing Systems
Separation of Non-Negative Mixture of Non-Negative Sources Using a Bayesian Approach and MCMC Sampling

IEEE Transactions on Signal Processing

Unmixing of Hyperspectral Images using Bayesian Non-negative Matrix Factorization with Volume Prior

Journal of Signal Processing Systems

Quantified Score

Hi-index	0.00

Visualization

Abstract

Hyperspectral imaging can be used in assessing the quality of foods by decomposing the image into constituents such as protein, starch, and water. Observed data can be considered a mixture of underlying characteristic spectra (endmembers), and estimating the constituents and their abundances requires efficient algorithms for spectral unmixing. We present a Bayesian spectral unmixing algorithm employing a volume constraint and propose an inference procedure based on Gibbs sampling. We evaluate the method on synthetic and real hyperspectral data of wheat kernels. Results show that our method perform as good or better than existing volume constrained methods. Further, our method gives credible intervals for the endmembers and abundances, which allows us to asses the confidence of the results.