Unmixing Sparse Hyperspectral Mixtures

Finding an accurate sparse approximation of a spectral vector described by a linear model, when there is available a library of possible constituent signals (called endmembers or atoms), is a hard combinatorial problem which, as in other areas, has been increasingly addressed. This paper studies the efficiency of the sparse regression techniques in the spectral unmixing problem by conducting a comparison between four different approaches: Moore-Penrose Pseudoinverse, Orthogonal Matching Pursuit (OMP) [1], Iterative Spectral Mixture Analysis (ISMA) [2] and 12 ll - sparse regression techniques, which are of widespread use in compressed sensing. We conclude that the 12 ll - sparse regression techniques, implemented here by Iterative Shrinkage/Thresholding (TwIST) algorithm [3], yield the state-ofthe-art in the hyperspectral unmixing area.

• Type: Journal
• Year: 2009
• Where: IGARSS
• Authors: Marian-Daniel Iordache, José M. Bioucas-Dias, Antonio Plaza
• Keywords: Accurate Sparse Approximation, IGARSS 2009, Remote Sensing, Sparse Regression Techniques, Spectral Unmixing Problem