SPECTRAL UNMIXING WITH SPARSITY AND STRUCTURING CONSTRAINTS

Abstract : This paper addresses the linear spectral unmixing problem, by incorporating different constraints that may be of interest in order to cope with spectral variability: sparsity (few nonzero abundances), group exclusivity (at most one nonzero abundance within subgroups of endmembers) and significance (non-zero abundances must exceed a threshold). We show how such problems can be solved exactly with mixed-integer programming techniques. Numerical simulations show that solutions can be computed for problems of limited, yet realistic , complexity, with improved estimation performance over existing methods, but with higher computing time. Index Terms-sparse spectral unmixing, L0-norm optimization , structured sparsity, mixed-integer programming.
Liste complète des métadonnées

https://hal.archives-ouvertes.fr/hal-02051443
Contributor : Ramzi Ben Mhenni <>
Submitted on : Wednesday, February 27, 2019 - 5:34:25 PM
Last modification on : Friday, April 5, 2019 - 8:05:19 PM

File

BenMhenni_WHISPERS18.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-02051443, version 1

Citation

Ramzi Mhenni, Sébastien Bourguignon, Jordan Ninin, Frédéric Schmidt. SPECTRAL UNMIXING WITH SPARSITY AND STRUCTURING CONSTRAINTS. IEEE Whispers (2018) : 9th Workshop on Hyperspectral Image and Signal Processing, Sep 2018, Amsterdam, Netherlands. ⟨hal-02051443⟩

Share

Metrics

Record views

19

Files downloads

17