Spectral image processing using sparse linear transforms

Research output: Chapter in Book/Report/Conference proceedingConference contributionResearchpeer-review

Abstract

We propose the employment of nonnegative sparse linear feature extraction as a tool for unsupervised spectral unmixing. Sparse feature extraction can be seen as a general linear unmixing approach that maps the data into a new dimensional space in which each of the components has only a limited number of non-zero values. Unlike other transforms that target decorrelation or statistical independence, our focus is on the enforcement of sparseness by imposing restrictions (such as cardinality or norm relationships), as well as nonnegativity. When compared to the linear mixing model, the sparse components can be naturally associated to the abundance of endmembers, and the inverse transform to the endmembers. Our approach is a variant of a well known technique based on Nonnegative Matrix Factorization (NMF). In most of the cases, the NMF components are produced using a gradient descent optimization algorithm that was previously shown to converge. To validate our approach we use quantitative (classification) and qualitative (visualization) analysis of hyperspectral data sets.

Original languageEnglish
Title of host publication2009 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2009 - Proceedings
Volume4
DOIs
StatePublished - 1 Dec 2009
Event2009 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2009 - Cape Town, South Africa
Duration: 12 Jul 200917 Jul 2009

Other

Other2009 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2009
CountrySouth Africa
CityCape Town
Period12/07/0917/07/09

Fingerprint

Factorization
image processing
Feature extraction
Image processing
transform
Inverse transforms
matrix
Visualization
visualization

Keywords

  • Hyperspectral imagery
  • Linear mixing model
  • Sparse nonnegative matrix factorization

Cite this

Robila, S. (2009). Spectral image processing using sparse linear transforms. In 2009 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2009 - Proceedings (Vol. 4). [5417431] https://doi.org/10.1109/IGARSS.2009.5417431
Robila, Stefan. / Spectral image processing using sparse linear transforms. 2009 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2009 - Proceedings. Vol. 4 2009.
@inproceedings{3e895c4fbdfd487290729dd2483979bf,
title = "Spectral image processing using sparse linear transforms",
abstract = "We propose the employment of nonnegative sparse linear feature extraction as a tool for unsupervised spectral unmixing. Sparse feature extraction can be seen as a general linear unmixing approach that maps the data into a new dimensional space in which each of the components has only a limited number of non-zero values. Unlike other transforms that target decorrelation or statistical independence, our focus is on the enforcement of sparseness by imposing restrictions (such as cardinality or norm relationships), as well as nonnegativity. When compared to the linear mixing model, the sparse components can be naturally associated to the abundance of endmembers, and the inverse transform to the endmembers. Our approach is a variant of a well known technique based on Nonnegative Matrix Factorization (NMF). In most of the cases, the NMF components are produced using a gradient descent optimization algorithm that was previously shown to converge. To validate our approach we use quantitative (classification) and qualitative (visualization) analysis of hyperspectral data sets.",
keywords = "Hyperspectral imagery, Linear mixing model, Sparse nonnegative matrix factorization",
author = "Stefan Robila",
year = "2009",
month = "12",
day = "1",
doi = "10.1109/IGARSS.2009.5417431",
language = "English",
isbn = "9781424433957",
volume = "4",
booktitle = "2009 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2009 - Proceedings",

}

Robila, S 2009, Spectral image processing using sparse linear transforms. in 2009 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2009 - Proceedings. vol. 4, 5417431, 2009 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2009, Cape Town, South Africa, 12/07/09. https://doi.org/10.1109/IGARSS.2009.5417431

Spectral image processing using sparse linear transforms. / Robila, Stefan.

2009 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2009 - Proceedings. Vol. 4 2009. 5417431.

Research output: Chapter in Book/Report/Conference proceedingConference contributionResearchpeer-review

TY - GEN

T1 - Spectral image processing using sparse linear transforms

AU - Robila, Stefan

PY - 2009/12/1

Y1 - 2009/12/1

N2 - We propose the employment of nonnegative sparse linear feature extraction as a tool for unsupervised spectral unmixing. Sparse feature extraction can be seen as a general linear unmixing approach that maps the data into a new dimensional space in which each of the components has only a limited number of non-zero values. Unlike other transforms that target decorrelation or statistical independence, our focus is on the enforcement of sparseness by imposing restrictions (such as cardinality or norm relationships), as well as nonnegativity. When compared to the linear mixing model, the sparse components can be naturally associated to the abundance of endmembers, and the inverse transform to the endmembers. Our approach is a variant of a well known technique based on Nonnegative Matrix Factorization (NMF). In most of the cases, the NMF components are produced using a gradient descent optimization algorithm that was previously shown to converge. To validate our approach we use quantitative (classification) and qualitative (visualization) analysis of hyperspectral data sets.

AB - We propose the employment of nonnegative sparse linear feature extraction as a tool for unsupervised spectral unmixing. Sparse feature extraction can be seen as a general linear unmixing approach that maps the data into a new dimensional space in which each of the components has only a limited number of non-zero values. Unlike other transforms that target decorrelation or statistical independence, our focus is on the enforcement of sparseness by imposing restrictions (such as cardinality or norm relationships), as well as nonnegativity. When compared to the linear mixing model, the sparse components can be naturally associated to the abundance of endmembers, and the inverse transform to the endmembers. Our approach is a variant of a well known technique based on Nonnegative Matrix Factorization (NMF). In most of the cases, the NMF components are produced using a gradient descent optimization algorithm that was previously shown to converge. To validate our approach we use quantitative (classification) and qualitative (visualization) analysis of hyperspectral data sets.

KW - Hyperspectral imagery

KW - Linear mixing model

KW - Sparse nonnegative matrix factorization

UR - http://www.scopus.com/inward/record.url?scp=77951270121&partnerID=8YFLogxK

U2 - 10.1109/IGARSS.2009.5417431

DO - 10.1109/IGARSS.2009.5417431

M3 - Conference contribution

SN - 9781424433957

VL - 4

BT - 2009 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2009 - Proceedings

ER -

Robila S. Spectral image processing using sparse linear transforms. In 2009 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2009 - Proceedings. Vol. 4. 2009. 5417431 https://doi.org/10.1109/IGARSS.2009.5417431