Anon-parametric classification strategy for remotely sensed images using both spectral and textural information

Mukesh Kumar, Douglas Alan Miller

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

Abstract

A classification strategy which does not require a priori assumptions about the statistical distribution of training pixels in each class is proposed. This method uses an indicator kriging approach in feature space to classify remotely sensed images incorporating both spectral and textural information of bands. Texture information is used only in cases where spectral information is not sufficient to resolve the assignment of the pixel to a class. Application of the proposed methodology on a remotely sensed natural scene shows an improvement in the overall classification accuracy with respect to the case when the scenes are classified by the traditional supervised Gaussian maximum likelihood classification (GMLC) method using either spectral band only or using both spectral and textural bands. A marked improvement in classification accuracy is obtained particularly for the classes for which the GMLC's assumption of multivariate normal distribution of training pixels in a class fails miserably.

Original languageEnglish (US)
Title of host publicationProceedings of the Third IASTED International Conference on Signal Processing, Pattern Recognition, and Applications
Pages81-89
Number of pages9
Volume2006
StatePublished - 2006
Event3rd IASTED International Conference on Signal Processing, Pattern Recognition, and Applications - Innsbruck, Austria
Duration: Feb 15 2006Feb 17 2006

Other

Other3rd IASTED International Conference on Signal Processing, Pattern Recognition, and Applications
CountryAustria
CityInnsbruck
Period2/15/062/17/06

Fingerprint

Pixels
Normal distribution
Maximum likelihood
Textures

All Science Journal Classification (ASJC) codes

  • Engineering(all)

Cite this

Kumar, M., & Miller, D. A. (2006). Anon-parametric classification strategy for remotely sensed images using both spectral and textural information. In Proceedings of the Third IASTED International Conference on Signal Processing, Pattern Recognition, and Applications (Vol. 2006, pp. 81-89)
Kumar, Mukesh ; Miller, Douglas Alan. / Anon-parametric classification strategy for remotely sensed images using both spectral and textural information. Proceedings of the Third IASTED International Conference on Signal Processing, Pattern Recognition, and Applications. Vol. 2006 2006. pp. 81-89
@inproceedings{82285626ee2446a08f1094d549476947,
title = "Anon-parametric classification strategy for remotely sensed images using both spectral and textural information",
abstract = "A classification strategy which does not require a priori assumptions about the statistical distribution of training pixels in each class is proposed. This method uses an indicator kriging approach in feature space to classify remotely sensed images incorporating both spectral and textural information of bands. Texture information is used only in cases where spectral information is not sufficient to resolve the assignment of the pixel to a class. Application of the proposed methodology on a remotely sensed natural scene shows an improvement in the overall classification accuracy with respect to the case when the scenes are classified by the traditional supervised Gaussian maximum likelihood classification (GMLC) method using either spectral band only or using both spectral and textural bands. A marked improvement in classification accuracy is obtained particularly for the classes for which the GMLC's assumption of multivariate normal distribution of training pixels in a class fails miserably.",
author = "Mukesh Kumar and Miller, {Douglas Alan}",
year = "2006",
language = "English (US)",
isbn = "0889865477",
volume = "2006",
pages = "81--89",
booktitle = "Proceedings of the Third IASTED International Conference on Signal Processing, Pattern Recognition, and Applications",

}

Kumar, M & Miller, DA 2006, Anon-parametric classification strategy for remotely sensed images using both spectral and textural information. in Proceedings of the Third IASTED International Conference on Signal Processing, Pattern Recognition, and Applications. vol. 2006, pp. 81-89, 3rd IASTED International Conference on Signal Processing, Pattern Recognition, and Applications, Innsbruck, Austria, 2/15/06.

Anon-parametric classification strategy for remotely sensed images using both spectral and textural information. / Kumar, Mukesh; Miller, Douglas Alan.

Proceedings of the Third IASTED International Conference on Signal Processing, Pattern Recognition, and Applications. Vol. 2006 2006. p. 81-89.

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

TY - GEN

T1 - Anon-parametric classification strategy for remotely sensed images using both spectral and textural information

AU - Kumar, Mukesh

AU - Miller, Douglas Alan

PY - 2006

Y1 - 2006

N2 - A classification strategy which does not require a priori assumptions about the statistical distribution of training pixels in each class is proposed. This method uses an indicator kriging approach in feature space to classify remotely sensed images incorporating both spectral and textural information of bands. Texture information is used only in cases where spectral information is not sufficient to resolve the assignment of the pixel to a class. Application of the proposed methodology on a remotely sensed natural scene shows an improvement in the overall classification accuracy with respect to the case when the scenes are classified by the traditional supervised Gaussian maximum likelihood classification (GMLC) method using either spectral band only or using both spectral and textural bands. A marked improvement in classification accuracy is obtained particularly for the classes for which the GMLC's assumption of multivariate normal distribution of training pixels in a class fails miserably.

AB - A classification strategy which does not require a priori assumptions about the statistical distribution of training pixels in each class is proposed. This method uses an indicator kriging approach in feature space to classify remotely sensed images incorporating both spectral and textural information of bands. Texture information is used only in cases where spectral information is not sufficient to resolve the assignment of the pixel to a class. Application of the proposed methodology on a remotely sensed natural scene shows an improvement in the overall classification accuracy with respect to the case when the scenes are classified by the traditional supervised Gaussian maximum likelihood classification (GMLC) method using either spectral band only or using both spectral and textural bands. A marked improvement in classification accuracy is obtained particularly for the classes for which the GMLC's assumption of multivariate normal distribution of training pixels in a class fails miserably.

UR - http://www.scopus.com/inward/record.url?scp=33847236272&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33847236272&partnerID=8YFLogxK

M3 - Conference contribution

SN - 0889865477

SN - 9780889865471

VL - 2006

SP - 81

EP - 89

BT - Proceedings of the Third IASTED International Conference on Signal Processing, Pattern Recognition, and Applications

ER -

Kumar M, Miller DA. Anon-parametric classification strategy for remotely sensed images using both spectral and textural information. In Proceedings of the Third IASTED International Conference on Signal Processing, Pattern Recognition, and Applications. Vol. 2006. 2006. p. 81-89