Statistics of target spectra in HSI scenes

J Scott Tyo, J. Robertson, J. Wollenbecker, Richard C. Olsen

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

10 Citations (Scopus)

Abstract

The majority of spectral imagery classifiers make a decision based on information from a particular spectrum, often the mean, that best represents the spectral signature of a particular target. It is known, however, that the spectral signature of a target can vary significantly due to differences in illumination conditions, shape, and material composition. Furthermore, many targets of interest are inherently mixed, as is the case with camouflaged military vehicles, leading to even greater variability. In this paper, a detailed statistical analysis is performed on HYDICE imagery of Davis Monthan AFB. Several hundred pixels are identified as belonging to the same target class and the distribution of spectral radiance within this group is studied. It is found that simple normal statistics do not adequately model either the total radiance or the single band spectral radiance distributions, both of which can have highly skewed histograms even when the spectral radiance is high. Goodness of fit tests are performed for maximum likelihood normal, lognormal, Γ, and Weibull distributions. It is found that lognormal statistics can model the total radiance and many single-band distributions reasonably well, possibly indicative of multiplicative noise features in remotely sensed spectral imagery.

Original languageEnglish (US)
Title of host publicationProceedings of SPIE - The International Society for Optical Engineering
PublisherSociety of Photo-Optical Instrumentation Engineers
Pages306-314
Number of pages9
Volume4132
DOIs
StatePublished - 2000
Externally publishedYes
EventImaging Spectrometry VI - San Diego, USA
Duration: Jul 31 2000Aug 2 2000

Other

OtherImaging Spectrometry VI
CitySan Diego, USA
Period7/31/008/2/00

Fingerprint

radiance
Statistics
statistics
Military vehicles
imagery
Weibull distribution
Normal distribution
spectral signatures
Maximum likelihood
Statistical methods
Classifiers
Lighting
Pixels
military vehicles
goodness of fit
Chemical analysis
spectral bands
classifiers
histograms
statistical analysis

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Condensed Matter Physics

Cite this

Tyo, J. S., Robertson, J., Wollenbecker, J., & Olsen, R. C. (2000). Statistics of target spectra in HSI scenes. In Proceedings of SPIE - The International Society for Optical Engineering (Vol. 4132, pp. 306-314). Society of Photo-Optical Instrumentation Engineers. https://doi.org/10.1117/12.406599

Statistics of target spectra in HSI scenes. / Tyo, J Scott; Robertson, J.; Wollenbecker, J.; Olsen, Richard C.

Proceedings of SPIE - The International Society for Optical Engineering. Vol. 4132 Society of Photo-Optical Instrumentation Engineers, 2000. p. 306-314.

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

Tyo, JS, Robertson, J, Wollenbecker, J & Olsen, RC 2000, Statistics of target spectra in HSI scenes. in Proceedings of SPIE - The International Society for Optical Engineering. vol. 4132, Society of Photo-Optical Instrumentation Engineers, pp. 306-314, Imaging Spectrometry VI, San Diego, USA, 7/31/00. https://doi.org/10.1117/12.406599
Tyo JS, Robertson J, Wollenbecker J, Olsen RC. Statistics of target spectra in HSI scenes. In Proceedings of SPIE - The International Society for Optical Engineering. Vol. 4132. Society of Photo-Optical Instrumentation Engineers. 2000. p. 306-314 https://doi.org/10.1117/12.406599
Tyo, J Scott ; Robertson, J. ; Wollenbecker, J. ; Olsen, Richard C. / Statistics of target spectra in HSI scenes. Proceedings of SPIE - The International Society for Optical Engineering. Vol. 4132 Society of Photo-Optical Instrumentation Engineers, 2000. pp. 306-314
@inproceedings{6a0f61d6b77b4b4cb4203459153f015e,
title = "Statistics of target spectra in HSI scenes",
abstract = "The majority of spectral imagery classifiers make a decision based on information from a particular spectrum, often the mean, that best represents the spectral signature of a particular target. It is known, however, that the spectral signature of a target can vary significantly due to differences in illumination conditions, shape, and material composition. Furthermore, many targets of interest are inherently mixed, as is the case with camouflaged military vehicles, leading to even greater variability. In this paper, a detailed statistical analysis is performed on HYDICE imagery of Davis Monthan AFB. Several hundred pixels are identified as belonging to the same target class and the distribution of spectral radiance within this group is studied. It is found that simple normal statistics do not adequately model either the total radiance or the single band spectral radiance distributions, both of which can have highly skewed histograms even when the spectral radiance is high. Goodness of fit tests are performed for maximum likelihood normal, lognormal, Γ, and Weibull distributions. It is found that lognormal statistics can model the total radiance and many single-band distributions reasonably well, possibly indicative of multiplicative noise features in remotely sensed spectral imagery.",
author = "Tyo, {J Scott} and J. Robertson and J. Wollenbecker and Olsen, {Richard C.}",
year = "2000",
doi = "10.1117/12.406599",
language = "English (US)",
volume = "4132",
pages = "306--314",
booktitle = "Proceedings of SPIE - The International Society for Optical Engineering",
publisher = "Society of Photo-Optical Instrumentation Engineers",

}

TY - GEN

T1 - Statistics of target spectra in HSI scenes

AU - Tyo, J Scott

AU - Robertson, J.

AU - Wollenbecker, J.

AU - Olsen, Richard C.

PY - 2000

Y1 - 2000

N2 - The majority of spectral imagery classifiers make a decision based on information from a particular spectrum, often the mean, that best represents the spectral signature of a particular target. It is known, however, that the spectral signature of a target can vary significantly due to differences in illumination conditions, shape, and material composition. Furthermore, many targets of interest are inherently mixed, as is the case with camouflaged military vehicles, leading to even greater variability. In this paper, a detailed statistical analysis is performed on HYDICE imagery of Davis Monthan AFB. Several hundred pixels are identified as belonging to the same target class and the distribution of spectral radiance within this group is studied. It is found that simple normal statistics do not adequately model either the total radiance or the single band spectral radiance distributions, both of which can have highly skewed histograms even when the spectral radiance is high. Goodness of fit tests are performed for maximum likelihood normal, lognormal, Γ, and Weibull distributions. It is found that lognormal statistics can model the total radiance and many single-band distributions reasonably well, possibly indicative of multiplicative noise features in remotely sensed spectral imagery.

AB - The majority of spectral imagery classifiers make a decision based on information from a particular spectrum, often the mean, that best represents the spectral signature of a particular target. It is known, however, that the spectral signature of a target can vary significantly due to differences in illumination conditions, shape, and material composition. Furthermore, many targets of interest are inherently mixed, as is the case with camouflaged military vehicles, leading to even greater variability. In this paper, a detailed statistical analysis is performed on HYDICE imagery of Davis Monthan AFB. Several hundred pixels are identified as belonging to the same target class and the distribution of spectral radiance within this group is studied. It is found that simple normal statistics do not adequately model either the total radiance or the single band spectral radiance distributions, both of which can have highly skewed histograms even when the spectral radiance is high. Goodness of fit tests are performed for maximum likelihood normal, lognormal, Γ, and Weibull distributions. It is found that lognormal statistics can model the total radiance and many single-band distributions reasonably well, possibly indicative of multiplicative noise features in remotely sensed spectral imagery.

UR - http://www.scopus.com/inward/record.url?scp=0034504281&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0034504281&partnerID=8YFLogxK

U2 - 10.1117/12.406599

DO - 10.1117/12.406599

M3 - Conference contribution

VL - 4132

SP - 306

EP - 314

BT - Proceedings of SPIE - The International Society for Optical Engineering

PB - Society of Photo-Optical Instrumentation Engineers

ER -