Adaptive Image Processing Using Nonlinear Inhibition

Recent models of lateral inhibition show a trend toward the use of nonlinear and adaptive processes to model the behaviour observed of biological mechanisms. These developments are of great interest for image processing, suggesting approaches to image processing problems which may avoid the need for arbitrary thresholds and supervision. Pinter's model of Lateral Inhibition using Nonlinear Local Interactions suggests a method of reorganizing image signals that is consistent with the goals of image processing. This paper explores the operational characterstics of a two dimensional version of Pinters model from an image processing perspective.