Deep Nonnegative Dictionary Factorization for Hyperspectral Unmixing
As a powerful blind source separation tool, Nonnegative Matrix Factorization (NMF) with effective regularizations has shown significant superiority in spectral unmixing of hyperspectral remote sensing images (HSIs) owing to its good physical interpretability and data adaptability. However, the major...
Main Authors: | , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2020-09-01
|
Series: | Remote Sensing |
Subjects: | |
Online Access: | https://www.mdpi.com/2072-4292/12/18/2882 |
id |
doaj-69bb6ad1b97b4f7d989d45fce7350ae7 |
---|---|
record_format |
Article |
spelling |
doaj-69bb6ad1b97b4f7d989d45fce7350ae72020-11-25T02:55:14ZengMDPI AGRemote Sensing2072-42922020-09-01122882288210.3390/rs12182882Deep Nonnegative Dictionary Factorization for Hyperspectral UnmixingWenhong Wang0Hongfu Liu1College of Computer Science, Liaocheng University, Liaocheng 252059, ChinaVolen National Center for Complex Systems, Departments of Computer Science, Brandeis University, Waltham, MA 02453, USAAs a powerful blind source separation tool, Nonnegative Matrix Factorization (NMF) with effective regularizations has shown significant superiority in spectral unmixing of hyperspectral remote sensing images (HSIs) owing to its good physical interpretability and data adaptability. However, the majority of existing NMF-based spectral unmixing methods only adopt the single layer factorization, which is not favorable for exploiting the complex and structured representation relationship of endmembers implied in HSIs. In order to overcome such an issue, we propose a novel two-stage Deep Nonnegative Dictionary Factorization (DNDF) approach with a sparseness constraint and self-supervised regularization for HSI unmixing. Beyond simply extending one-layer factorization to multi-layer, DNDF conducts fuzzy clustering to tackle the mixed endmembers of HSIs. Moreover, self-supervised regularization is integrated into our DNDF model to impose an effective constraint on the endmember matrix. Experimental results on three real HSIs demonstrate the superiority of DNDF over several state-of-the-art methods.https://www.mdpi.com/2072-4292/12/18/2882deep nonnegative dictionary factorizationhyperspectral unmixingself-supervised learningsparse coding |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Wenhong Wang Hongfu Liu |
spellingShingle |
Wenhong Wang Hongfu Liu Deep Nonnegative Dictionary Factorization for Hyperspectral Unmixing Remote Sensing deep nonnegative dictionary factorization hyperspectral unmixing self-supervised learning sparse coding |
author_facet |
Wenhong Wang Hongfu Liu |
author_sort |
Wenhong Wang |
title |
Deep Nonnegative Dictionary Factorization for Hyperspectral Unmixing |
title_short |
Deep Nonnegative Dictionary Factorization for Hyperspectral Unmixing |
title_full |
Deep Nonnegative Dictionary Factorization for Hyperspectral Unmixing |
title_fullStr |
Deep Nonnegative Dictionary Factorization for Hyperspectral Unmixing |
title_full_unstemmed |
Deep Nonnegative Dictionary Factorization for Hyperspectral Unmixing |
title_sort |
deep nonnegative dictionary factorization for hyperspectral unmixing |
publisher |
MDPI AG |
series |
Remote Sensing |
issn |
2072-4292 |
publishDate |
2020-09-01 |
description |
As a powerful blind source separation tool, Nonnegative Matrix Factorization (NMF) with effective regularizations has shown significant superiority in spectral unmixing of hyperspectral remote sensing images (HSIs) owing to its good physical interpretability and data adaptability. However, the majority of existing NMF-based spectral unmixing methods only adopt the single layer factorization, which is not favorable for exploiting the complex and structured representation relationship of endmembers implied in HSIs. In order to overcome such an issue, we propose a novel two-stage Deep Nonnegative Dictionary Factorization (DNDF) approach with a sparseness constraint and self-supervised regularization for HSI unmixing. Beyond simply extending one-layer factorization to multi-layer, DNDF conducts fuzzy clustering to tackle the mixed endmembers of HSIs. Moreover, self-supervised regularization is integrated into our DNDF model to impose an effective constraint on the endmember matrix. Experimental results on three real HSIs demonstrate the superiority of DNDF over several state-of-the-art methods. |
topic |
deep nonnegative dictionary factorization hyperspectral unmixing self-supervised learning sparse coding |
url |
https://www.mdpi.com/2072-4292/12/18/2882 |
work_keys_str_mv |
AT wenhongwang deepnonnegativedictionaryfactorizationforhyperspectralunmixing AT hongfuliu deepnonnegativedictionaryfactorizationforhyperspectralunmixing |
_version_ |
1724717457755602944 |