目的 为了解决现有半色调算法产生的图像边界模糊以及视觉蠕虫等问题，研究一种视觉效果较好的半色调图像。方法 提出一种在半色调过程中动态分配误差扩散系数的数字半色调方法，减少由固定误差扩散系数和固定扩散方向带来的边界模糊、视觉蠕虫以及结构性纹理等现象。结果 文中提出的算法相较于传统的误差扩散算法其PSNR值提高了约0.5，SSIM值提高了约0.06，NSME值下降了约0.06。不仅解决了传统误差扩散算法中的边界模糊现象，同时也更好地表达了半色调图像的细节纹理信息。结论 依据提出的算法产生的半色调图像视觉效果较好，并且算法简单易行，运行效率高。

The work aims to study a halftone image with better visual effect to solve such problems as fuzzy image boundary and visual worm generated in the existing halftone algorithm. A digital halftone method for dynamic distribution of error diffusion coefficients in the halftone process was proposed to reduce such phenomena as fuzzy boundary, visual worm and structural texture brought by fixed error diffusion coefficients and fixed diffusion direction. Compared to the traditional error diffusion algorithm, PSNR value of the proposed method was improved by about 0.5, SSIM value improved by about 0.06 and NSME value reduced by about 0.06. In such case, not only the fuzzy boundary in the traditional error diffusion algorithm was solved, but also the information on detailed texture of halftone image was better expressed. The visual effect of the halftone image generated based on the proposed algorithm is good, and the algorithm is simple and easy and of high running efficiency.