Multifractal Analysis of Multivariate Images Using Gamma Markov Random Field Priors

Herwig Wendt, Sebastien Combrexelle, Yoann Altmann, Jean-Yves Tourneret, Stephen McLaughlin, Patrice Abry

Research output: Contribution to journalArticle

Abstract

Texture characterization of natural images using the mathematical framework of multifractal (MF) analysis, enables the study of the fluctuations in the regularity of image intensity. Although successfully applied in various contexts, the use of MF analysis has so far been limited to the independent analysis of a single image, while the data available in applications are increasingly multivariate. This paper addresses this limitation and proposes a joint Bayesian model and associated estimation procedure for MF parameters of multivariate images. It builds on a recently introduced generic statistical model that enabled the Bayesian estimation of MF parameters for a single image and relies on the following original key contributions: First, we develop a novel Fourier domain statistical model for a single image that permits the use of a likelihood that is separable in the MF parameters via data augmentation. Second, a joint Bayesian model for multivariate images is formulated in which prior models based on gamma Markov random elds encode the assumption of the smooth evolution of MF parameters between the image components. The design of the likelihood and of conjugate prior models is such that exploitation of the conjugacy between the likelihood and prior models enables an ecient estimation procedure that can handle a large number of data components. Numerical simulations conducted using sequences of multifractal images demonstrate that the proposed procedure significantly outperforms previous univariate benchmark formulations at a competitive computational cost.
LanguageEnglish
Pages1294–1316
Number of pages23
JournalSIAM Journal on Imaging Sciences
Volume11
Issue number2
DOIs
StatePublished - 22 May 2018

Fingerprint

Textures
Computer simulation
Costs
Statistical Models

Cite this

Wendt, Herwig ; Combrexelle, Sebastien ; Altmann, Yoann ; Tourneret, Jean-Yves ; McLaughlin, Stephen ; Abry, Patrice. / Multifractal Analysis of Multivariate Images Using Gamma Markov Random Field Priors. In: SIAM Journal on Imaging Sciences . 2018 ; Vol. 11, No. 2. pp. 1294–1316
@article{c5f85fb6136c4cacb2beabc0b92c997d,
title = "Multifractal Analysis of Multivariate Images Using Gamma Markov Random Field Priors",
abstract = "Texture characterization of natural images using the mathematical framework of multifractal (MF) analysis, enables the study of the fluctuations in the regularity of image intensity. Although successfully applied in various contexts, the use of MF analysis has so far been limited to the independent analysis of a single image, while the data available in applications are increasingly multivariate. This paper addresses this limitation and proposes a joint Bayesian model and associated estimation procedure for MF parameters of multivariate images. It builds on a recently introduced generic statistical model that enabled the Bayesian estimation of MF parameters for a single image and relies on the following original key contributions: First, we develop a novel Fourier domain statistical model for a single image that permits the use of a likelihood that is separable in the MF parameters via data augmentation. Second, a joint Bayesian model for multivariate images is formulated in which prior models based on gamma Markov random elds encode the assumption of the smooth evolution of MF parameters between the image components. The design of the likelihood and of conjugate prior models is such that exploitation of the conjugacy between the likelihood and prior models enables an ecient estimation procedure that can handle a large number of data components. Numerical simulations conducted using sequences of multifractal images demonstrate that the proposed procedure significantly outperforms previous univariate benchmark formulations at a competitive computational cost.",
author = "Herwig Wendt and Sebastien Combrexelle and Yoann Altmann and Jean-Yves Tourneret and Stephen McLaughlin and Patrice Abry",
year = "2018",
month = "5",
day = "22",
doi = "10.1137/17M1151304",
language = "English",
volume = "11",
pages = "1294–1316",
journal = "SIAM Journal on Imaging Sciences",
issn = "1936-4954",
publisher = "Society for Industrial and Applied Mathematics Publications",
number = "2",

}

Multifractal Analysis of Multivariate Images Using Gamma Markov Random Field Priors. / Wendt, Herwig; Combrexelle, Sebastien; Altmann, Yoann; Tourneret, Jean-Yves; McLaughlin, Stephen; Abry, Patrice.

In: SIAM Journal on Imaging Sciences , Vol. 11, No. 2, 22.05.2018, p. 1294–1316.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Multifractal Analysis of Multivariate Images Using Gamma Markov Random Field Priors

AU - Wendt,Herwig

AU - Combrexelle,Sebastien

AU - Altmann,Yoann

AU - Tourneret,Jean-Yves

AU - McLaughlin,Stephen

AU - Abry,Patrice

PY - 2018/5/22

Y1 - 2018/5/22

N2 - Texture characterization of natural images using the mathematical framework of multifractal (MF) analysis, enables the study of the fluctuations in the regularity of image intensity. Although successfully applied in various contexts, the use of MF analysis has so far been limited to the independent analysis of a single image, while the data available in applications are increasingly multivariate. This paper addresses this limitation and proposes a joint Bayesian model and associated estimation procedure for MF parameters of multivariate images. It builds on a recently introduced generic statistical model that enabled the Bayesian estimation of MF parameters for a single image and relies on the following original key contributions: First, we develop a novel Fourier domain statistical model for a single image that permits the use of a likelihood that is separable in the MF parameters via data augmentation. Second, a joint Bayesian model for multivariate images is formulated in which prior models based on gamma Markov random elds encode the assumption of the smooth evolution of MF parameters between the image components. The design of the likelihood and of conjugate prior models is such that exploitation of the conjugacy between the likelihood and prior models enables an ecient estimation procedure that can handle a large number of data components. Numerical simulations conducted using sequences of multifractal images demonstrate that the proposed procedure significantly outperforms previous univariate benchmark formulations at a competitive computational cost.

AB - Texture characterization of natural images using the mathematical framework of multifractal (MF) analysis, enables the study of the fluctuations in the regularity of image intensity. Although successfully applied in various contexts, the use of MF analysis has so far been limited to the independent analysis of a single image, while the data available in applications are increasingly multivariate. This paper addresses this limitation and proposes a joint Bayesian model and associated estimation procedure for MF parameters of multivariate images. It builds on a recently introduced generic statistical model that enabled the Bayesian estimation of MF parameters for a single image and relies on the following original key contributions: First, we develop a novel Fourier domain statistical model for a single image that permits the use of a likelihood that is separable in the MF parameters via data augmentation. Second, a joint Bayesian model for multivariate images is formulated in which prior models based on gamma Markov random elds encode the assumption of the smooth evolution of MF parameters between the image components. The design of the likelihood and of conjugate prior models is such that exploitation of the conjugacy between the likelihood and prior models enables an ecient estimation procedure that can handle a large number of data components. Numerical simulations conducted using sequences of multifractal images demonstrate that the proposed procedure significantly outperforms previous univariate benchmark formulations at a competitive computational cost.

UR - https://www.irit.fr/~Herwig.Wendt/software.html#bayesc2

U2 - 10.1137/17M1151304

DO - 10.1137/17M1151304

M3 - Article

VL - 11

SP - 1294

EP - 1316

JO - SIAM Journal on Imaging Sciences

T2 - SIAM Journal on Imaging Sciences

JF - SIAM Journal on Imaging Sciences

SN - 1936-4954

IS - 2

ER -