A model-reference neural network scheme for obtaining laser printer modulation templates

摘要

An adaptive subpixel modulation optimization system for obtaining laser printer modulation templates is introduced in this paper. This paper presents a novel training algorithm for employing a Madaline II network in a model - reference loop. The neural network weights are adapted using the Madaline II rule. Desired response feedback is obtained from a new performance feedback algorithm.