Abstract
Texture analysis is central in many image processing problems. It can be conducted by studying the local regularity fluctuations of image amplitudes, and multifractal analysis provides a theoretical and practical framework for such a characterization. Yet, due to the non Gaussian nature and intricate dependence structure of multifractal models, accurate parameter estimation is challenging: standard estimators yield modest performance, and alternative (semi-)parametric estimators exhibit prohibitive computational cost for large images. This present contribution addresses these difficulties and proposes a Bayesian procedure for the estimation of the multifractality parameter c2 for images. It relies on a recently proposed semi-parametric model for the multivariate statistics of log-wavelet leaders and on a Whittle approximation that enables its numerical evaluation. The key result is a closed-form expression for the Whittle likelihood. Numerical simulations indicate the excellent performance of the method, significantly improving estimation performance over standard estimators and computational efficiency over previously proposed Bayesian estimators.
Original language | English |
---|---|
Title of host publication | 2015 23rd European Signal Processing Conference (EUSIPCO) |
Publisher | IEEE |
Pages | 1003-1007 |
Number of pages | 5 |
ISBN (Print) | 9780992862633 |
DOIs | |
Publication status | Published - 2015 |
Event | 23rd European Signal Processing Conference 2015 - Nice, France Duration: 31 Aug 2015 → 4 Sept 2015 |
Conference
Conference | 23rd European Signal Processing Conference 2015 |
---|---|
Abbreviated title | EUSIPCO 2015 |
Country/Territory | France |
City | Nice |
Period | 31/08/15 → 4/09/15 |
Keywords
- Bayesian estimation
- Hankel transform
- Multifractal analysis
- Texture analysis
- Whittle likelihood
ASJC Scopus subject areas
- Media Technology
- Computer Vision and Pattern Recognition
- Signal Processing