Image classification based on a multiresolution two dimensional hidden Markov model

Jia Li, Robert M. Gray

Research output: Contribution to conferencePaper

7 Citations (Scopus)

Abstract

This paper presents an image classification algorithm based upon a two dimensional multiresolution hidden Markov model (MHMM). This model represents an image by feature vectors in several resolutions and considers the feature vectors statistically dependent through an underlying state process assumed to be a multiscale Markov mesh. To estimate the model by the maximum likelihood criterion, approximations are made successively based on the EM algorithm to reach feasible computation. To classify an image, the algorithm attempts to find the optimal set of states with the maximum a posteriori probability. The states are then mapped into classes. The multiresolution model enables multiscale context information to be incorporated into classification. Suboptimal algorithms based on the model provide progressive classification which greatly speeds up classification based on single resolution HMMs.

Original languageEnglish (US)
Pages348-352
Number of pages5
StatePublished - Dec 1 1999
EventInternational Conference on Image Processing (ICIP'99) - Kobe, Jpn
Duration: Oct 24 1999Oct 28 1999

Other

OtherInternational Conference on Image Processing (ICIP'99)
CityKobe, Jpn
Period10/24/9910/28/99

Fingerprint

Image classification
Hidden Markov models
Maximum likelihood

All Science Journal Classification (ASJC) codes

  • Hardware and Architecture
  • Computer Vision and Pattern Recognition
  • Electrical and Electronic Engineering

Cite this

Li, J., & Gray, R. M. (1999). Image classification based on a multiresolution two dimensional hidden Markov model. 348-352. Paper presented at International Conference on Image Processing (ICIP'99), Kobe, Jpn, .
Li, Jia ; Gray, Robert M. / Image classification based on a multiresolution two dimensional hidden Markov model. Paper presented at International Conference on Image Processing (ICIP'99), Kobe, Jpn, .5 p.
@conference{d79491b8e27e4599bbc74a9ad3cfb9d9,
title = "Image classification based on a multiresolution two dimensional hidden Markov model",
abstract = "This paper presents an image classification algorithm based upon a two dimensional multiresolution hidden Markov model (MHMM). This model represents an image by feature vectors in several resolutions and considers the feature vectors statistically dependent through an underlying state process assumed to be a multiscale Markov mesh. To estimate the model by the maximum likelihood criterion, approximations are made successively based on the EM algorithm to reach feasible computation. To classify an image, the algorithm attempts to find the optimal set of states with the maximum a posteriori probability. The states are then mapped into classes. The multiresolution model enables multiscale context information to be incorporated into classification. Suboptimal algorithms based on the model provide progressive classification which greatly speeds up classification based on single resolution HMMs.",
author = "Jia Li and Gray, {Robert M.}",
year = "1999",
month = "12",
day = "1",
language = "English (US)",
pages = "348--352",
note = "International Conference on Image Processing (ICIP'99) ; Conference date: 24-10-1999 Through 28-10-1999",

}

Li, J & Gray, RM 1999, 'Image classification based on a multiresolution two dimensional hidden Markov model' Paper presented at International Conference on Image Processing (ICIP'99), Kobe, Jpn, 10/24/99 - 10/28/99, pp. 348-352.

Image classification based on a multiresolution two dimensional hidden Markov model. / Li, Jia; Gray, Robert M.

1999. 348-352 Paper presented at International Conference on Image Processing (ICIP'99), Kobe, Jpn, .

Research output: Contribution to conferencePaper

TY - CONF

T1 - Image classification based on a multiresolution two dimensional hidden Markov model

AU - Li, Jia

AU - Gray, Robert M.

PY - 1999/12/1

Y1 - 1999/12/1

N2 - This paper presents an image classification algorithm based upon a two dimensional multiresolution hidden Markov model (MHMM). This model represents an image by feature vectors in several resolutions and considers the feature vectors statistically dependent through an underlying state process assumed to be a multiscale Markov mesh. To estimate the model by the maximum likelihood criterion, approximations are made successively based on the EM algorithm to reach feasible computation. To classify an image, the algorithm attempts to find the optimal set of states with the maximum a posteriori probability. The states are then mapped into classes. The multiresolution model enables multiscale context information to be incorporated into classification. Suboptimal algorithms based on the model provide progressive classification which greatly speeds up classification based on single resolution HMMs.

AB - This paper presents an image classification algorithm based upon a two dimensional multiresolution hidden Markov model (MHMM). This model represents an image by feature vectors in several resolutions and considers the feature vectors statistically dependent through an underlying state process assumed to be a multiscale Markov mesh. To estimate the model by the maximum likelihood criterion, approximations are made successively based on the EM algorithm to reach feasible computation. To classify an image, the algorithm attempts to find the optimal set of states with the maximum a posteriori probability. The states are then mapped into classes. The multiresolution model enables multiscale context information to be incorporated into classification. Suboptimal algorithms based on the model provide progressive classification which greatly speeds up classification based on single resolution HMMs.

UR - http://www.scopus.com/inward/record.url?scp=0033335553&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0033335553&partnerID=8YFLogxK

M3 - Paper

SP - 348

EP - 352

ER -

Li J, Gray RM. Image classification based on a multiresolution two dimensional hidden Markov model. 1999. Paper presented at International Conference on Image Processing (ICIP'99), Kobe, Jpn, .