摘要： 针对传统误差扩散法在中频区域存在的结构纹理，提出了一种基于视觉差的误差扩散算法(PEB Error Diffusion)。PEB-ED借鉴了DBS中考察人类视觉误差的思想，对误差扩散结果中最容易发生误处理的中间调区域像素进行二次处理，因而明显减少了由固定的误差扩散滤波器所产生的结构性纹理，提高了图像的整体视觉效果。此外，它还可以减弱实际打印中“点增益现象”的影响。

关键词: 数字半色调, 误差扩散法, 直接二值搜索, 人类视觉系统, 点增益

Abstract: This paper presents a perceived error-based error diffusion (PEB Error Diffusion)which aims at reducing the structured artifacts introduced by error diffusion. First the continous-tone image is halftoned by error diffusion, then the halftoned is rectified by the error between the perceived continous-tone image and the output halftone image. The algorithm efficiently reduces the structured artifacts in midtone area and results in higher visual quality. Besides, dot-overgain is weakened in some way.

Key words: Digital halftoning, Error diffusion, DBS, HVS, Dot-overgain

中图分类号: 

• TP391