Five Neighbor Stochastic Error Diffusion for Digital Halftoning

摘要

In this paper we present a five neighbor stochastic error diffusion technique. The technique assumes a Gaussian process for weights adjustment. Five neighboring pixels are considered for uniform error distribution. The direct and indirect paths from a particular pixel to all of its five neighboring pixels are considered to calculate the weights of the error filter. The distribution of quantization error equally among the neighboring pixels, results in reduction of noise patterns as compared to other error diffusion techniques.