Choice of a 2-D Causal Autoregressive texture model using information criteria
Pattern Recognition Letters, Volume 24, Number 9-10, pages 1191-1201 - June 2003
In the context of parametric modeling for image processing, we derive
an estimation method for both the order and the parameters of
2-D causal autoregressive model with different geometries of
support. Model parameters are estimated from a
lattice representation, i.e. based on reflection
coefficients. Lattice parameter estimation algorithms offer
advantages compared to the
Yule-Walker method: they do not require matrix inversion
and their computation are robust and fast. For order selection, Information
Criterion (IC) methods are the most commonly used.
Therefore our order selection method is based on the
combination of an IC and the prediction errors
of models computed from the lattice parameter estimation
algorithm. In this paper, we favour two consistent criteria
compared to the non-consistent Akaike criterion: the first criterion is a 2-D
extension of Bayesian Information Criterion (BIC); the second
criterion, noted $phi_beta$, extended here to the
2-D case,
is a generalization drawn on Rissanen's works. Simulations are
provided on synthetic and natural textures with Quarter Plane support and
Non-Symmetrical Half Plane support. We validate our
results on natural textures using the Kullback divergence. The results
show the interest of the combination of 2-DFLRLS algorithm and
$phi_beta$ criterion to characterize natural textures.
Références BibTex
@Article{AO2003_1305,
}
author | = {Alata, O. and Olivier, C.}, | |
title | = {Choice of a 2-D Causal Autoregressive texture model using information criteria.}, | |
journal | = {Pattern Recognition Letters}, | |
number | = {9-10}, | |
volume | = {24}, | |
pages | = {1191-1201}, | |
month | = {June}, | |
year | = {2003}, | |
keywords | = {Characterization of textures; 2-D autoregressive model; Order selection by information criteria; Lattice representation; Kullback divergence}, | |
url | = {http://www.sciencedirect.com/science?_ob=IssueURL\&_tockey=%23TOC%235665%232003%23999759990%23386685%23FLA%23\&_auth=y\&view=c\&_acct=C000028278\&_version=1\&_urlVersion=0\&_userid=554142\&md5=c59c250cfdca012491bd95900f7161d9}, |