Skip to Main content Skip to Navigation
Conference papers

A Bayesian framework for the multifractal analysis of images using data augmentation and a Whittle approximation

Abstract : Texture analysis is an image processing task that can be conducted using the mathematical framework of multifractal analysis to study the regularity fluctuations of image intensity and the practical tools for their assessment, such as (wavelet) leaders. A recently introduced statistical model for leaders enables the Bayesian estimation of multifractal parameters. It significantly improves performance over standard (linear regression based) estimation. However, the computational cost induced by the associated nonstandard posterior distributions limits its application. The present work proposes an alternative Bayesian model for multifractal analysis that leads to more efficient algorithms. It relies on three original contributions: A novel generative model for the Fourier coefficients of log-leaders; an appropriate reparametrization for handling its inherent constraints; a data-augmented Bayesian model yielding standard conditional posterior distributions that can be sampled exactly. Numerical simulations using synthetic multifractal images demonstrate the excellent performance of the proposed algorithm, both in terms of estimation quality and computational cost.
Complete list of metadatas

Cited literature [22 references]  Display  Hide  Download
Contributor : Open Archive Toulouse Archive Ouverte (oatao) <>
Submitted on : Friday, April 21, 2017 - 5:17:07 PM
Last modification on : Sunday, November 22, 2020 - 7:48:07 PM
Long-term archiving on: : Saturday, July 22, 2017 - 2:11:49 PM


Files produced by the author(s)


  • HAL Id : hal-01511897, version 1
  • OATAO : 17035


Sébastien Combrexelle, Herwig Wendt, Yoann Altmann, Jean-Yves Tourneret, Stephen Mclaughlin, et al.. A Bayesian framework for the multifractal analysis of images using data augmentation and a Whittle approximation. 41st IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP 2016), Mar 2016, Shanghai, China. pp. 4224-4228. ⟨hal-01511897⟩



Record views


Files downloads