A model reduction method for a class of 2-D systems

A decomposition-aggregation scheme for reduction of dimensionality for a class of 2-D systems is introduced. This method, which is based upon the extension of the singular perturbation method in two dimensions, is used to decompose the original 2-D system into two reduced-order 2-D subsystems. These reduced order subsystems are shown to effectively capture the dynamical behavior of the original full-order system. Two numerical examples are provided that indicate the effectiveness of this method when used in image modeling applications.<
>