Statistical modelling of multimodal SAR images

Authors:
A. El-Zaart;D. Ziou
Affiliations:
Dépt. d'informatique, Université de Sherbrooke, Sherbrooke, Qc., Canada J1K 2R1;Dépt. d'informatique, Université de Sherbrooke, Sherbrooke, Qc., Canada J1K 2R1
Venue:
International Journal of Remote Sensing
Year:
2007

Citing 0
Cited 4

A combined Markovian and Dirichlet sub-mixture modeling for evidence assignment: Application to image fusion

Pattern Recognition Letters
Dot Detection of Optical Braille Images for Braille Cells Recognition

ICCHP '08 Proceedings of the 11th international conference on Computers Helping People with Special Needs
Finite Gamma mixture modelling using minimum message length inference: application to SAR image analysis

International Journal of Remote Sensing
Expectation-maximization technique for fibro-glandular discs detection in mammography images

Computers in Biology and Medicine

Quantified Score

Hi-index	0.00

Visualization

Abstract

This paper discusses the statistical modelling of multimodal SAR image histograms. We use a generalized mixture of distributions with flexible shapes that are likely to fit the SAR image histogram. We then form a system called GGBL, composed of four parametric distributions (Gaussian, Gamma, Beta and Log-Normal). Selection of a parametric distribution from the GGBL system for each mode of the multimodal SAR histogram is performed according to the location of the skewness and flatness coefficients in this space. We propose a distribution stability method for distribution selection, using the asymmetry and flatness coefficients, and a feature method for the estimation of the parameters of these distributions based on the characteristic points of the histogram. The GGBL system is applied to real histograms of RADARSAT and ERS SAR images with different numbers of looks. The results obtained are promising.