Image classification based on Markov random field models with Jeffreys divergence

Ryuei Nishii, Shinto Eguchi

Research output: Contribution to journalArticle

11 Citations (Scopus)

Abstract

This paper considers image classification based on a Markov random field (MRF), where the random field proposed here adopts Jeffreys divergence between category-specific probability densities. The classification method based on the proposed MRF is shown to be an extension of Switzer's soothing method, which is applied in remote sensing and geospatial communities. Furthermore, the exact error rates due to the proposed and Switzer's methods are obtained under the simple setup, and several properties are derived. Our method is applied to a benchmark data set of image classification, and exhibits a good performance in comparison with conventional methods.

Original languageEnglish
Pages (from-to)1997-2008
Number of pages12
JournalJournal of Multivariate Analysis
Volume97
Issue number9
DOIs
Publication statusPublished - Oct 1 2006

Fingerprint

Image classification
Image Classification
Random Field
Divergence
Remote sensing
Model
Probability Density
Remote Sensing
Error Rate
Random field
Benchmark

All Science Journal Classification (ASJC) codes

  • Statistics, Probability and Uncertainty
  • Numerical Analysis
  • Statistics and Probability

Cite this

Image classification based on Markov random field models with Jeffreys divergence. / Nishii, Ryuei; Eguchi, Shinto.

In: Journal of Multivariate Analysis, Vol. 97, No. 9, 01.10.2006, p. 1997-2008.

Research output: Contribution to journalArticle

Nishii, Ryuei ; Eguchi, Shinto. / Image classification based on Markov random field models with Jeffreys divergence. In: Journal of Multivariate Analysis. 2006 ; Vol. 97, No. 9. pp. 1997-2008.
@article{78ee679eb577421bbcab235621597ede,
title = "Image classification based on Markov random field models with Jeffreys divergence",
abstract = "This paper considers image classification based on a Markov random field (MRF), where the random field proposed here adopts Jeffreys divergence between category-specific probability densities. The classification method based on the proposed MRF is shown to be an extension of Switzer's soothing method, which is applied in remote sensing and geospatial communities. Furthermore, the exact error rates due to the proposed and Switzer's methods are obtained under the simple setup, and several properties are derived. Our method is applied to a benchmark data set of image classification, and exhibits a good performance in comparison with conventional methods.",
author = "Ryuei Nishii and Shinto Eguchi",
year = "2006",
month = "10",
day = "1",
doi = "10.1016/j.jmva.2006.01.009",
language = "English",
volume = "97",
pages = "1997--2008",
journal = "Journal of Multivariate Analysis",
issn = "0047-259X",
publisher = "Academic Press Inc.",
number = "9",

}

TY - JOUR

T1 - Image classification based on Markov random field models with Jeffreys divergence

AU - Nishii, Ryuei

AU - Eguchi, Shinto

PY - 2006/10/1

Y1 - 2006/10/1

N2 - This paper considers image classification based on a Markov random field (MRF), where the random field proposed here adopts Jeffreys divergence between category-specific probability densities. The classification method based on the proposed MRF is shown to be an extension of Switzer's soothing method, which is applied in remote sensing and geospatial communities. Furthermore, the exact error rates due to the proposed and Switzer's methods are obtained under the simple setup, and several properties are derived. Our method is applied to a benchmark data set of image classification, and exhibits a good performance in comparison with conventional methods.

AB - This paper considers image classification based on a Markov random field (MRF), where the random field proposed here adopts Jeffreys divergence between category-specific probability densities. The classification method based on the proposed MRF is shown to be an extension of Switzer's soothing method, which is applied in remote sensing and geospatial communities. Furthermore, the exact error rates due to the proposed and Switzer's methods are obtained under the simple setup, and several properties are derived. Our method is applied to a benchmark data set of image classification, and exhibits a good performance in comparison with conventional methods.

UR - http://www.scopus.com/inward/record.url?scp=33748414613&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33748414613&partnerID=8YFLogxK

U2 - 10.1016/j.jmva.2006.01.009

DO - 10.1016/j.jmva.2006.01.009

M3 - Article

VL - 97

SP - 1997

EP - 2008

JO - Journal of Multivariate Analysis

JF - Journal of Multivariate Analysis

SN - 0047-259X

IS - 9

ER -