Fast Hyperspectral Unmixing via Reweighted Sparse Regression

Hongwei HAN; Ke GUO; Maozhi WANG; Tingbin ZHANG; Shuang ZHANG

doi:10.1587/transinf.2018EDP7374

IEICE TRANSACTIONS on Information

Open Access
Fast Hyperspectral Unmixing via Reweighted Sparse Regression

Hongwei HAN, Ke GUO, Maozhi WANG, Tingbin ZHANG, Shuang ZHANG

Full Text Views

54

Cite this

Free PDF (5.1MB)

Summary :

The sparse unmixing of hyperspectral data has attracted much attention in recent years because it does not need to estimate the number of endmembers nor consider the lack of pure pixels in a given hyperspectral scene. However, the high mutual coherence of spectral libraries strongly affects the practicality of sparse unmixing. The collaborative sparse unmixing via variable splitting and augmented Lagrangian (CLSUnSAL) algorithm is a classic sparse unmixing algorithm that performs better than other sparse unmixing methods. In this paper, we propose a CLSUnSAL-based hyperspectral unmixing method based on dictionary pruning and reweighted sparse regression. First, the algorithm identifies a subset of the original library elements using a dictionary pruning strategy. Second, we present a weighted sparse regression algorithm based on CLSUnSAL to further enhance the sparsity of endmember spectra in a given library. Third, we apply the weighted sparse regression algorithm on the pruned spectral library. The effectiveness of the proposed algorithm is demonstrated on both simulated and real hyperspectral datasets. For simulated data cubes (DC1, DC2 and DC3), the number of the pruned spectral library elements is reduced by at least 94% and the runtime of the proposed algorithm is less than 10% of that of CLSUnSAL. For simulated DC4 and DC5, the runtime of the proposed algorithm is less than 15% of that of CLSUnSAL. For the real hyperspectral datasets, the pruned spectral library successfully reduces the original dictionary size by 76% and the runtime of the proposed algorithm is 11.21% of that of CLSUnSAL. These experimental results show that our proposed algorithm not only substantially improves the accuracy of unmixing solutions but is also much faster than some other state-of-the-art sparse unmixing algorithms.

Publication: IEICE TRANSACTIONS on Information Vol.E102-D No.9 pp.1819-1832

Publication Date: 2019/09/01

Publicized: 2019/05/28

Online ISSN: 1745-1361

DOI: 10.1587/transinf.2018EDP7374

Type of Manuscript: PAPER

Category: Image Processing and Video Processing

Authors

Hongwei HAN
  Geomathematics Key Laboratory of Sichuan Province in Chengdu University of Technology,Engineering & Technical College of Chengdu University of Technology
Ke GUO
  Geomathematics Key Laboratory of Sichuan Province in Chengdu University of Technology
Maozhi WANG
  Geomathematics Key Laboratory of Sichuan Province in Chengdu University of Technology
Tingbin ZHANG
  Chengdu University of Technology
Shuang ZHANG
  Neijiang Normal University

Keyword

sparse unmixing, dictionary pruning, hyperspectral imaging, iterative reweighting

Cite this

Copy

Hongwei HAN, Ke GUO, Maozhi WANG, Tingbin ZHANG, Shuang ZHANG, "Fast Hyperspectral Unmixing via Reweighted Sparse Regression" in IEICE TRANSACTIONS on Information, vol. E102-D, no. 9, pp. 1819-1832, September 2019, doi: 10.1587/transinf.2018EDP7374.
Abstract: The sparse unmixing of hyperspectral data has attracted much attention in recent years because it does not need to estimate the number of endmembers nor consider the lack of pure pixels in a given hyperspectral scene. However, the high mutual coherence of spectral libraries strongly affects the practicality of sparse unmixing. The collaborative sparse unmixing via variable splitting and augmented Lagrangian (CLSUnSAL) algorithm is a classic sparse unmixing algorithm that performs better than other sparse unmixing methods. In this paper, we propose a CLSUnSAL-based hyperspectral unmixing method based on dictionary pruning and reweighted sparse regression. First, the algorithm identifies a subset of the original library elements using a dictionary pruning strategy. Second, we present a weighted sparse regression algorithm based on CLSUnSAL to further enhance the sparsity of endmember spectra in a given library. Third, we apply the weighted sparse regression algorithm on the pruned spectral library. The effectiveness of the proposed algorithm is demonstrated on both simulated and real hyperspectral datasets. For simulated data cubes (DC1, DC2 and DC3), the number of the pruned spectral library elements is reduced by at least 94% and the runtime of the proposed algorithm is less than 10% of that of CLSUnSAL. For simulated DC4 and DC5, the runtime of the proposed algorithm is less than 15% of that of CLSUnSAL. For the real hyperspectral datasets, the pruned spectral library successfully reduces the original dictionary size by 76% and the runtime of the proposed algorithm is 11.21% of that of CLSUnSAL. These experimental results show that our proposed algorithm not only substantially improves the accuracy of unmixing solutions but is also much faster than some other state-of-the-art sparse unmixing algorithms.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.2018EDP7374/_p

Copy

@ARTICLE{e102-d_9_1819,
author={Hongwei HAN, Ke GUO, Maozhi WANG, Tingbin ZHANG, Shuang ZHANG, },
journal={IEICE TRANSACTIONS on Information},
title={Fast Hyperspectral Unmixing via Reweighted Sparse Regression},
year={2019},
volume={E102-D},
number={9},
pages={1819-1832},
abstract={The sparse unmixing of hyperspectral data has attracted much attention in recent years because it does not need to estimate the number of endmembers nor consider the lack of pure pixels in a given hyperspectral scene. However, the high mutual coherence of spectral libraries strongly affects the practicality of sparse unmixing. The collaborative sparse unmixing via variable splitting and augmented Lagrangian (CLSUnSAL) algorithm is a classic sparse unmixing algorithm that performs better than other sparse unmixing methods. In this paper, we propose a CLSUnSAL-based hyperspectral unmixing method based on dictionary pruning and reweighted sparse regression. First, the algorithm identifies a subset of the original library elements using a dictionary pruning strategy. Second, we present a weighted sparse regression algorithm based on CLSUnSAL to further enhance the sparsity of endmember spectra in a given library. Third, we apply the weighted sparse regression algorithm on the pruned spectral library. The effectiveness of the proposed algorithm is demonstrated on both simulated and real hyperspectral datasets. For simulated data cubes (DC1, DC2 and DC3), the number of the pruned spectral library elements is reduced by at least 94% and the runtime of the proposed algorithm is less than 10% of that of CLSUnSAL. For simulated DC4 and DC5, the runtime of the proposed algorithm is less than 15% of that of CLSUnSAL. For the real hyperspectral datasets, the pruned spectral library successfully reduces the original dictionary size by 76% and the runtime of the proposed algorithm is 11.21% of that of CLSUnSAL. These experimental results show that our proposed algorithm not only substantially improves the accuracy of unmixing solutions but is also much faster than some other state-of-the-art sparse unmixing algorithms.},
keywords={},
doi={10.1587/transinf.2018EDP7374},
ISSN={1745-1361},
month={September},}

Copy

TY - JOUR
TI - Fast Hyperspectral Unmixing via Reweighted Sparse Regression
T2 - IEICE TRANSACTIONS on Information
SP - 1819
EP - 1832
AU - Hongwei HAN
AU - Ke GUO
AU - Maozhi WANG
AU - Tingbin ZHANG
AU - Shuang ZHANG
PY - 2019
DO - 10.1587/transinf.2018EDP7374
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E102-D
IS - 9
JA - IEICE TRANSACTIONS on Information
Y1 - September 2019
AB - The sparse unmixing of hyperspectral data has attracted much attention in recent years because it does not need to estimate the number of endmembers nor consider the lack of pure pixels in a given hyperspectral scene. However, the high mutual coherence of spectral libraries strongly affects the practicality of sparse unmixing. The collaborative sparse unmixing via variable splitting and augmented Lagrangian (CLSUnSAL) algorithm is a classic sparse unmixing algorithm that performs better than other sparse unmixing methods. In this paper, we propose a CLSUnSAL-based hyperspectral unmixing method based on dictionary pruning and reweighted sparse regression. First, the algorithm identifies a subset of the original library elements using a dictionary pruning strategy. Second, we present a weighted sparse regression algorithm based on CLSUnSAL to further enhance the sparsity of endmember spectra in a given library. Third, we apply the weighted sparse regression algorithm on the pruned spectral library. The effectiveness of the proposed algorithm is demonstrated on both simulated and real hyperspectral datasets. For simulated data cubes (DC1, DC2 and DC3), the number of the pruned spectral library elements is reduced by at least 94% and the runtime of the proposed algorithm is less than 10% of that of CLSUnSAL. For simulated DC4 and DC5, the runtime of the proposed algorithm is less than 15% of that of CLSUnSAL. For the real hyperspectral datasets, the pruned spectral library successfully reduces the original dictionary size by 76% and the runtime of the proposed algorithm is 11.21% of that of CLSUnSAL. These experimental results show that our proposed algorithm not only substantially improves the accuracy of unmixing solutions but is also much faster than some other state-of-the-art sparse unmixing algorithms.
ER -

IEICE TRANSACTIONS on Information