A new method for 2-D AR model parameter estimation

Abstract

This paper presents a new method for two dimensional (2-D) autoregressive (AR) model parameters estimation. Based on Levinson-Durbin method, a new approach is extended for 2-D series, and for this purpose, an algorithm is presented. The presented method preserves in the 2-D case advantages of Levinson-Durbin such as recursive and online estimation similar 1-D case. This approach is illustrated by a numerical example. 1