多级误差扩散加网技术研究

Research on Multi-level Error Diffusion Screening Technology

为了提高输出图像的质量,使人眼视觉产生连续调效果,使用了多级误差扩散算法生成网目调图像。多级加网技术在传统二值(0和1)之间增加了中间阶调,突破了长久以来人们一贯应用的"非黑即白"输出理念。通过分析传统误差扩散和多级误差扩散理论基础,提出了多级多阶误差扩散算法。多级多阶误差扩散采用迭代算法,引用图像四叉树来表示原始图像与输出网目调图像的差异。算法结果表明,多级多阶误差扩散加网技术使加网后的图像具有更为精确的阶调复制和更为丰富的层次再现,图像之间的过渡更为平整自然。

In order to improve the quality of the output image, and make the human eye vision generate continuous tone effect, it uses the multi-level error diffusion algorithm to generate halftone image. Multi-level screening technology has increased the intermediate tone in the traditional two values （0 and 1 ）. It has broken out of the ＂ non-black or white＂ output concept which has long been consistent application. Analyzing the theoretical basis of the traditional error diffusion and multi- scale error diffusion algorithm came up with the muhi-stage error diffusion algorithm. Multi-stage error diffusion used iterative algorithm, and quoted image quadtree to show the difference of original image and output halftone image. The results of this algorithm showed that multi-stage error diffusion screening technology made the image after screening more accurate tone reproduction and richer gradation reappearance, transition between the images were more smooth and natural.