Overlapped Feature-Preserving Multiscale Error Diffusion For Digital Halftoning

A new digital halftoning process established on multiscale error diffusion technique is suggested in this paper. In this work an overlapped feature modified multiscale error diffusion technique is proposed. The noticeable features of the algorithms are 1. Image quadtree, 2. Selection of the dot through maximum intensity guidance theorem, 3. Feature modified non causal diffusion of error. The superiority of the halftoned is assessed through subjective and objective estimation. For the objective evaluation mean square error and universal quality index are considered in this paper.