Bayesian sensor image fusion using local linear generative models

R. K. Sharma, T. K. Leen, M. Pavel

Research output: Contribution to journalArticle

30 Citations (Scopus)

Abstract

We present a probabilistic method for fusion of images produced by multiple sensors. The approach is based on an image formation model in which the sensor images are noisy, locally linear functions of an underlying true scene (latent variable). A Bayesian framework then provides for maximum-likelihood or maximum a posteriori estimates of the true scene from the sensor images. Least-squares estimates of the parameters of the image formation model involve (local) second-order image statistics, and are related to local principal-component analysis. We demonstrate the efficacy of the method on images from visible-band and infrared sensors.

Original languageEnglish (US)
Pages (from-to)1364-1376
Number of pages13
JournalOptical Engineering
Volume40
Issue number7
DOIs
StatePublished - Jul 2001
Externally publishedYes

Fingerprint

Image fusion
Image sensors
Image processing
fusion
sensors
Sensors
Principal component analysis
Maximum likelihood
Statistics
Infrared radiation
estimates
principal components analysis
statistics

Keywords

  • Image formation model
  • Image fusion
  • Infrared imaging
  • Local linear models
  • Maximum a posteriori fusion
  • Multisensor fusion
  • Probabilistic fusion
  • Sensor fusion
  • Sensors

ASJC Scopus subject areas

  • Atomic and Molecular Physics, and Optics

Cite this

Bayesian sensor image fusion using local linear generative models. / Sharma, R. K.; Leen, T. K.; Pavel, M.

In: Optical Engineering, Vol. 40, No. 7, 07.2001, p. 1364-1376.

Research output: Contribution to journalArticle

Sharma, R. K. ; Leen, T. K. ; Pavel, M. / Bayesian sensor image fusion using local linear generative models. In: Optical Engineering. 2001 ; Vol. 40, No. 7. pp. 1364-1376.
@article{64f157fb48e44f15b0c7e1e54e64b22f,
title = "Bayesian sensor image fusion using local linear generative models",
abstract = "We present a probabilistic method for fusion of images produced by multiple sensors. The approach is based on an image formation model in which the sensor images are noisy, locally linear functions of an underlying true scene (latent variable). A Bayesian framework then provides for maximum-likelihood or maximum a posteriori estimates of the true scene from the sensor images. Least-squares estimates of the parameters of the image formation model involve (local) second-order image statistics, and are related to local principal-component analysis. We demonstrate the efficacy of the method on images from visible-band and infrared sensors.",
keywords = "Image formation model, Image fusion, Infrared imaging, Local linear models, Maximum a posteriori fusion, Multisensor fusion, Probabilistic fusion, Sensor fusion, Sensors",
author = "Sharma, {R. K.} and Leen, {T. K.} and M. Pavel",
year = "2001",
month = "7",
doi = "10.1117/1.1384886",
language = "English (US)",
volume = "40",
pages = "1364--1376",
journal = "Optical Engineering",
issn = "0091-3286",
publisher = "SPIE",
number = "7",

}

TY - JOUR

T1 - Bayesian sensor image fusion using local linear generative models

AU - Sharma, R. K.

AU - Leen, T. K.

AU - Pavel, M.

PY - 2001/7

Y1 - 2001/7

N2 - We present a probabilistic method for fusion of images produced by multiple sensors. The approach is based on an image formation model in which the sensor images are noisy, locally linear functions of an underlying true scene (latent variable). A Bayesian framework then provides for maximum-likelihood or maximum a posteriori estimates of the true scene from the sensor images. Least-squares estimates of the parameters of the image formation model involve (local) second-order image statistics, and are related to local principal-component analysis. We demonstrate the efficacy of the method on images from visible-band and infrared sensors.

AB - We present a probabilistic method for fusion of images produced by multiple sensors. The approach is based on an image formation model in which the sensor images are noisy, locally linear functions of an underlying true scene (latent variable). A Bayesian framework then provides for maximum-likelihood or maximum a posteriori estimates of the true scene from the sensor images. Least-squares estimates of the parameters of the image formation model involve (local) second-order image statistics, and are related to local principal-component analysis. We demonstrate the efficacy of the method on images from visible-band and infrared sensors.

KW - Image formation model

KW - Image fusion

KW - Infrared imaging

KW - Local linear models

KW - Maximum a posteriori fusion

KW - Multisensor fusion

KW - Probabilistic fusion

KW - Sensor fusion

KW - Sensors

UR - http://www.scopus.com/inward/record.url?scp=0035388480&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0035388480&partnerID=8YFLogxK

U2 - 10.1117/1.1384886

DO - 10.1117/1.1384886

M3 - Article

AN - SCOPUS:0035388480

VL - 40

SP - 1364

EP - 1376

JO - Optical Engineering

JF - Optical Engineering

SN - 0091-3286

IS - 7

ER -