A fuzzy neural approach to design of a Wiener printer model incorporated into model-based digital halftoning

Abstract

This paper presents the design of a Wiener model for a class of laser electrophotographic printers utilizing a fuzzy neural approach. The feasibility and effectiveness of incorporating the proposed spatial-based Wiener printer model into the framework of model-based digital halftoning is also investigated. The Wiener model comprises a two-dimensional FIR filter followed by a nonlinear static mapping. The nonlinear static mapping is synthesized based on a fuzzy neural network to account for the tone distortion commonly observed in a printing system. A set of systematic steps for parametric identification/optimization and evaluation of the Wiener printer model is proposed. Experimental results as well as a comparative study show that the Wiener model-based halftoning approach is superior (in terms of image quality) to conventional ones and comparable to the one utilizing a comprehensive printer model. Another comparative study reveals that the Wiener printer model, though computationally less efficient than the conventional dot-overlap model, consumes significantly less time to process an image than the comprehensive printer model does.