Adaptive clustering of spectral components for band selection in hyperspectral imagery

Research output: Contribution to conferencePaper

Abstract

A novel unsupervised band selection method is proposed, where adaptive clustering of spectral components is employed. For a given hyperspectral image, its spectral bands are grouped into clusters, based on the similarity measured by histogram-determined mutual information and its normalised version. Then, variable numbers of clusters
can be determined automatically in our approach by selecting the most likely clustering boundaries, thus thresholding of image similarity in grouping bands is avoided. Finally, one representative band is extracted from each cluster by minimising the sum of inter-band difference within the band cluster. Using the well-known 92AV3C dataset, the proposed approach is evaluated in terms of efficiency and effectiveness. Experimental results have demonstrated the great potential of our proposed methodology in automatic band selection for many applications of hyperspectral imagery.
LanguageEnglish
Pages90-93
Number of pages4
Publication statusPublished - 18 May 2011
EventHyperspectral Imaging Conference 2011 - University of Strathclyde, Glasgow, United Kingdom
Duration: 17 May 201118 May 2011

Conference

ConferenceHyperspectral Imaging Conference 2011
CountryUnited Kingdom
CityGlasgow
Period17/05/1118/05/11

Fingerprint

imagery
histogram
methodology
spectral band
method

Keywords

  • hyperspectral imaging
  • spectral components
  • band selection

Cite this

Ren, J., Kelman, T., & Marshall, S. (2011). Adaptive clustering of spectral components for band selection in hyperspectral imagery. 90-93. Paper presented at Hyperspectral Imaging Conference 2011, Glasgow, United Kingdom.
Ren, Jinchang ; Kelman, Timothy ; Marshall, Stephen. / Adaptive clustering of spectral components for band selection in hyperspectral imagery. Paper presented at Hyperspectral Imaging Conference 2011, Glasgow, United Kingdom.4 p.
@conference{974e4801394b449bb786671ec092b901,
title = "Adaptive clustering of spectral components for band selection in hyperspectral imagery",
abstract = "A novel unsupervised band selection method is proposed, where adaptive clustering of spectral components is employed. For a given hyperspectral image, its spectral bands are grouped into clusters, based on the similarity measured by histogram-determined mutual information and its normalised version. Then, variable numbers of clusters can be determined automatically in our approach by selecting the most likely clustering boundaries, thus thresholding of image similarity in grouping bands is avoided. Finally, one representative band is extracted from each cluster by minimising the sum of inter-band difference within the band cluster. Using the well-known 92AV3C dataset, the proposed approach is evaluated in terms of efficiency and effectiveness. Experimental results have demonstrated the great potential of our proposed methodology in automatic band selection for many applications of hyperspectral imagery.",
keywords = "hyperspectral imaging, spectral components, band selection",
author = "Jinchang Ren and Timothy Kelman and Stephen Marshall",
year = "2011",
month = "5",
day = "18",
language = "English",
pages = "90--93",
note = "Hyperspectral Imaging Conference 2011 ; Conference date: 17-05-2011 Through 18-05-2011",

}

Ren, J, Kelman, T & Marshall, S 2011, 'Adaptive clustering of spectral components for band selection in hyperspectral imagery' Paper presented at Hyperspectral Imaging Conference 2011, Glasgow, United Kingdom, 17/05/11 - 18/05/11, pp. 90-93.

Adaptive clustering of spectral components for band selection in hyperspectral imagery. / Ren, Jinchang; Kelman, Timothy; Marshall, Stephen.

2011. 90-93 Paper presented at Hyperspectral Imaging Conference 2011, Glasgow, United Kingdom.

Research output: Contribution to conferencePaper

TY - CONF

T1 - Adaptive clustering of spectral components for band selection in hyperspectral imagery

AU - Ren, Jinchang

AU - Kelman, Timothy

AU - Marshall, Stephen

PY - 2011/5/18

Y1 - 2011/5/18

N2 - A novel unsupervised band selection method is proposed, where adaptive clustering of spectral components is employed. For a given hyperspectral image, its spectral bands are grouped into clusters, based on the similarity measured by histogram-determined mutual information and its normalised version. Then, variable numbers of clusters can be determined automatically in our approach by selecting the most likely clustering boundaries, thus thresholding of image similarity in grouping bands is avoided. Finally, one representative band is extracted from each cluster by minimising the sum of inter-band difference within the band cluster. Using the well-known 92AV3C dataset, the proposed approach is evaluated in terms of efficiency and effectiveness. Experimental results have demonstrated the great potential of our proposed methodology in automatic band selection for many applications of hyperspectral imagery.

AB - A novel unsupervised band selection method is proposed, where adaptive clustering of spectral components is employed. For a given hyperspectral image, its spectral bands are grouped into clusters, based on the similarity measured by histogram-determined mutual information and its normalised version. Then, variable numbers of clusters can be determined automatically in our approach by selecting the most likely clustering boundaries, thus thresholding of image similarity in grouping bands is avoided. Finally, one representative band is extracted from each cluster by minimising the sum of inter-band difference within the band cluster. Using the well-known 92AV3C dataset, the proposed approach is evaluated in terms of efficiency and effectiveness. Experimental results have demonstrated the great potential of our proposed methodology in automatic band selection for many applications of hyperspectral imagery.

KW - hyperspectral imaging

KW - spectral components

KW - band selection

UR - http://www.strath.ac.uk/eee/research/events/hsi/

UR - http://www.strath.ac.uk/media/departments/eee/cesip/hsi/HSI_2011_Schedule.pdf

M3 - Paper

SP - 90

EP - 93

ER -

Ren J, Kelman T, Marshall S. Adaptive clustering of spectral components for band selection in hyperspectral imagery. 2011. Paper presented at Hyperspectral Imaging Conference 2011, Glasgow, United Kingdom.