Level set segmentation of hyperspectral images using joint spectral edge and signature information

Radford R. Juang, Philippe Burlina, Amit Banerjee

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

Abstract

This paper describes a new method for segmenting hyperspectral imagery (HSI) using dynamic curves. We are concerned about challenging HSI target segmentation/detection use cases where the scene includes confusers exhibiting a spectral return similar to the desired signature and in close proximity of the object of interest. Our method is based on a level sets approach. It fuses all available spectral bands and incorporates spectral as well as spatial information to obtain a finer target segmentation. The proposed method applies level set segmentation to HSI by defining an expansion force field that combines both hyperspectral gradient information as well as the desired spectral signature. We carry out experiments on HSI datacubes obtained from a sensor spanning visible and near IR and show improved results when compared to direct spectral matching in challenging close range scenes including significant level of nearby confusers.

Original languageEnglish (US)
Title of host publicationProceedings of the 11th International Conference on Information Fusion, FUSION 2008
DOIs
StatePublished - Dec 1 2008
Event11th International Conference on Information Fusion, FUSION 2008 - Cologne, Germany
Duration: Jun 30 2008Jul 3 2008

Other

Other11th International Conference on Information Fusion, FUSION 2008
CountryGermany
CityCologne
Period6/30/087/3/08

Fingerprint

Electric fuses
Sensors
Experiments

Keywords

  • Hyperspectral imagery
  • Level set segmentation
  • Spectral gradient
  • Spectral signature

ASJC Scopus subject areas

  • Computational Theory and Mathematics
  • Computer Science Applications
  • Information Systems

Cite this

Juang, R. R., Burlina, P., & Banerjee, A. (2008). Level set segmentation of hyperspectral images using joint spectral edge and signature information. In Proceedings of the 11th International Conference on Information Fusion, FUSION 2008 [4632446] https://doi.org/10.1109/ICIF.2008.4632446

Level set segmentation of hyperspectral images using joint spectral edge and signature information. / Juang, Radford R.; Burlina, Philippe; Banerjee, Amit.

Proceedings of the 11th International Conference on Information Fusion, FUSION 2008. 2008. 4632446.

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

Juang, RR, Burlina, P & Banerjee, A 2008, Level set segmentation of hyperspectral images using joint spectral edge and signature information. in Proceedings of the 11th International Conference on Information Fusion, FUSION 2008., 4632446, 11th International Conference on Information Fusion, FUSION 2008, Cologne, Germany, 6/30/08. https://doi.org/10.1109/ICIF.2008.4632446
Juang RR, Burlina P, Banerjee A. Level set segmentation of hyperspectral images using joint spectral edge and signature information. In Proceedings of the 11th International Conference on Information Fusion, FUSION 2008. 2008. 4632446 https://doi.org/10.1109/ICIF.2008.4632446
Juang, Radford R. ; Burlina, Philippe ; Banerjee, Amit. / Level set segmentation of hyperspectral images using joint spectral edge and signature information. Proceedings of the 11th International Conference on Information Fusion, FUSION 2008. 2008.
@inproceedings{c7865b55c7d0445f8d5b33ba8675b12c,
title = "Level set segmentation of hyperspectral images using joint spectral edge and signature information",
abstract = "This paper describes a new method for segmenting hyperspectral imagery (HSI) using dynamic curves. We are concerned about challenging HSI target segmentation/detection use cases where the scene includes confusers exhibiting a spectral return similar to the desired signature and in close proximity of the object of interest. Our method is based on a level sets approach. It fuses all available spectral bands and incorporates spectral as well as spatial information to obtain a finer target segmentation. The proposed method applies level set segmentation to HSI by defining an expansion force field that combines both hyperspectral gradient information as well as the desired spectral signature. We carry out experiments on HSI datacubes obtained from a sensor spanning visible and near IR and show improved results when compared to direct spectral matching in challenging close range scenes including significant level of nearby confusers.",
keywords = "Hyperspectral imagery, Level set segmentation, Spectral gradient, Spectral signature",
author = "Juang, {Radford R.} and Philippe Burlina and Amit Banerjee",
year = "2008",
month = "12",
day = "1",
doi = "10.1109/ICIF.2008.4632446",
language = "English (US)",
isbn = "9783000248832",
booktitle = "Proceedings of the 11th International Conference on Information Fusion, FUSION 2008",

}

TY - GEN

T1 - Level set segmentation of hyperspectral images using joint spectral edge and signature information

AU - Juang, Radford R.

AU - Burlina, Philippe

AU - Banerjee, Amit

PY - 2008/12/1

Y1 - 2008/12/1

N2 - This paper describes a new method for segmenting hyperspectral imagery (HSI) using dynamic curves. We are concerned about challenging HSI target segmentation/detection use cases where the scene includes confusers exhibiting a spectral return similar to the desired signature and in close proximity of the object of interest. Our method is based on a level sets approach. It fuses all available spectral bands and incorporates spectral as well as spatial information to obtain a finer target segmentation. The proposed method applies level set segmentation to HSI by defining an expansion force field that combines both hyperspectral gradient information as well as the desired spectral signature. We carry out experiments on HSI datacubes obtained from a sensor spanning visible and near IR and show improved results when compared to direct spectral matching in challenging close range scenes including significant level of nearby confusers.

AB - This paper describes a new method for segmenting hyperspectral imagery (HSI) using dynamic curves. We are concerned about challenging HSI target segmentation/detection use cases where the scene includes confusers exhibiting a spectral return similar to the desired signature and in close proximity of the object of interest. Our method is based on a level sets approach. It fuses all available spectral bands and incorporates spectral as well as spatial information to obtain a finer target segmentation. The proposed method applies level set segmentation to HSI by defining an expansion force field that combines both hyperspectral gradient information as well as the desired spectral signature. We carry out experiments on HSI datacubes obtained from a sensor spanning visible and near IR and show improved results when compared to direct spectral matching in challenging close range scenes including significant level of nearby confusers.

KW - Hyperspectral imagery

KW - Level set segmentation

KW - Spectral gradient

KW - Spectral signature

UR - http://www.scopus.com/inward/record.url?scp=56749138135&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=56749138135&partnerID=8YFLogxK

U2 - 10.1109/ICIF.2008.4632446

DO - 10.1109/ICIF.2008.4632446

M3 - Conference contribution

AN - SCOPUS:56749138135

SN - 9783000248832

BT - Proceedings of the 11th International Conference on Information Fusion, FUSION 2008

ER -