Robust linear spectral unmixing using outlier detection

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

This paper presents a Bayesian algorithm for linear spectral unmixing that accounts for outliers present in the data. The proposed model assumes that the pixel reflectances are linear mixtures of unknown endmembers, corrupted by an additional term modelling outliers and additive Gaussian noise. A Markov random field is considered for outlier detection based on the spatial and spectral structures of the anomalies. This allows outliers to be identified in particular regions and wavelengths of the data cube. A Bayesian algorithm is proposed to estimate the parameters involved in the model yielding a joint linear unmixing and outlier detection algorithm. Simulations conducted with synthetic data demonstrate the accuracy of the proposed unmixing and outlier detection strategy for the analysis of hyperspectral images.

Original languageEnglish
Title of host publicationAcoustics, Speech and Signal Processing (ICASSP), 2015 IEEE International Conference on
PublisherIEEE
Pages2464-2468
Number of pages5
Volume2015-August
ISBN (Print)9781467369978
DOIs
Publication statusPublished - 4 Aug 2015
Event40th IEEE International Conference on Acoustics, Speech and Signal Processing 2015 - Brisbane, Australia
Duration: 19 Apr 201524 Apr 2015

Conference

Conference40th IEEE International Conference on Acoustics, Speech and Signal Processing 2015
Abbreviated titleICASSP 2015
CountryAustralia
CityBrisbane
Period19/04/1524/04/15

Fingerprint

Pixels
Wavelength

Keywords

  • Bayesian estimation
  • Hyperspectral imagery
  • MCMC
  • nonlinearity detection
  • unsupervised spectral unmixing

ASJC Scopus subject areas

  • Signal Processing
  • Software
  • Electrical and Electronic Engineering

Cite this

Altmann, Y., McLaughlin, S., & Hero, A. (2015). Robust linear spectral unmixing using outlier detection. In Acoustics, Speech and Signal Processing (ICASSP), 2015 IEEE International Conference on (Vol. 2015-August, pp. 2464-2468). [7178414] IEEE. https://doi.org/10.1109/ICASSP.2015.7178414
Altmann, Yoann ; McLaughlin, Steve ; Hero, Alfred. / Robust linear spectral unmixing using outlier detection. Acoustics, Speech and Signal Processing (ICASSP), 2015 IEEE International Conference on. Vol. 2015-August IEEE, 2015. pp. 2464-2468
@inproceedings{33a20b612a2044f0b5a1aa2454a9f40c,
title = "Robust linear spectral unmixing using outlier detection",
abstract = "This paper presents a Bayesian algorithm for linear spectral unmixing that accounts for outliers present in the data. The proposed model assumes that the pixel reflectances are linear mixtures of unknown endmembers, corrupted by an additional term modelling outliers and additive Gaussian noise. A Markov random field is considered for outlier detection based on the spatial and spectral structures of the anomalies. This allows outliers to be identified in particular regions and wavelengths of the data cube. A Bayesian algorithm is proposed to estimate the parameters involved in the model yielding a joint linear unmixing and outlier detection algorithm. Simulations conducted with synthetic data demonstrate the accuracy of the proposed unmixing and outlier detection strategy for the analysis of hyperspectral images.",
keywords = "Bayesian estimation, Hyperspectral imagery, MCMC, nonlinearity detection, unsupervised spectral unmixing",
author = "Yoann Altmann and Steve McLaughlin and Alfred Hero",
year = "2015",
month = "8",
day = "4",
doi = "10.1109/ICASSP.2015.7178414",
language = "English",
isbn = "9781467369978",
volume = "2015-August",
pages = "2464--2468",
booktitle = "Acoustics, Speech and Signal Processing (ICASSP), 2015 IEEE International Conference on",
publisher = "IEEE",
address = "United States",

}

Altmann, Y, McLaughlin, S & Hero, A 2015, Robust linear spectral unmixing using outlier detection. in Acoustics, Speech and Signal Processing (ICASSP), 2015 IEEE International Conference on. vol. 2015-August, 7178414, IEEE, pp. 2464-2468, 40th IEEE International Conference on Acoustics, Speech and Signal Processing 2015, Brisbane, Australia, 19/04/15. https://doi.org/10.1109/ICASSP.2015.7178414

Robust linear spectral unmixing using outlier detection. / Altmann, Yoann; McLaughlin, Steve; Hero, Alfred.

Acoustics, Speech and Signal Processing (ICASSP), 2015 IEEE International Conference on. Vol. 2015-August IEEE, 2015. p. 2464-2468 7178414.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Robust linear spectral unmixing using outlier detection

AU - Altmann, Yoann

AU - McLaughlin, Steve

AU - Hero, Alfred

PY - 2015/8/4

Y1 - 2015/8/4

N2 - This paper presents a Bayesian algorithm for linear spectral unmixing that accounts for outliers present in the data. The proposed model assumes that the pixel reflectances are linear mixtures of unknown endmembers, corrupted by an additional term modelling outliers and additive Gaussian noise. A Markov random field is considered for outlier detection based on the spatial and spectral structures of the anomalies. This allows outliers to be identified in particular regions and wavelengths of the data cube. A Bayesian algorithm is proposed to estimate the parameters involved in the model yielding a joint linear unmixing and outlier detection algorithm. Simulations conducted with synthetic data demonstrate the accuracy of the proposed unmixing and outlier detection strategy for the analysis of hyperspectral images.

AB - This paper presents a Bayesian algorithm for linear spectral unmixing that accounts for outliers present in the data. The proposed model assumes that the pixel reflectances are linear mixtures of unknown endmembers, corrupted by an additional term modelling outliers and additive Gaussian noise. A Markov random field is considered for outlier detection based on the spatial and spectral structures of the anomalies. This allows outliers to be identified in particular regions and wavelengths of the data cube. A Bayesian algorithm is proposed to estimate the parameters involved in the model yielding a joint linear unmixing and outlier detection algorithm. Simulations conducted with synthetic data demonstrate the accuracy of the proposed unmixing and outlier detection strategy for the analysis of hyperspectral images.

KW - Bayesian estimation

KW - Hyperspectral imagery

KW - MCMC

KW - nonlinearity detection

KW - unsupervised spectral unmixing

UR - http://www.scopus.com/inward/record.url?scp=84946094266&partnerID=8YFLogxK

U2 - 10.1109/ICASSP.2015.7178414

DO - 10.1109/ICASSP.2015.7178414

M3 - Conference contribution

SN - 9781467369978

VL - 2015-August

SP - 2464

EP - 2468

BT - Acoustics, Speech and Signal Processing (ICASSP), 2015 IEEE International Conference on

PB - IEEE

ER -

Altmann Y, McLaughlin S, Hero A. Robust linear spectral unmixing using outlier detection. In Acoustics, Speech and Signal Processing (ICASSP), 2015 IEEE International Conference on. Vol. 2015-August. IEEE. 2015. p. 2464-2468. 7178414 https://doi.org/10.1109/ICASSP.2015.7178414