2-DIMENSIONAL LEAST-SQUARES LATTICE ALGORITHM FOR LINEAR PREDICTION

In this paper, we propose a two-dimensional (2-D) least-squares lattic e (LSL) algorithm for the general case of the autoregressive (AR) mode l with an asymmetric half-plane (AHP) coefficient support. The resulti ng LSL algorithm gives both order and space recursions for the 2-D det erministic normal equations. The size and shape of the coefficient sup port region of the proposed lattice filter can be chosen arbitrarily F urthermore, the ordering of the support signal can be assigned arbitra rily. Finally, computer simulation for modeling a texture image is dem onstrated to confirm the proposed model gives rapid convergence.