关键词: 图像分割, 误差分散, 数字半调

Abstract:

To reduce the visual artifacts produced by the conventional error diffusion algorithm in the digital halftone process, an error diffusion algorithm based on image segmentation is presented. In this algorithm the image segmentation technology is used to segment an image into two types of region. In one region, pixels have the same gray level while in the other they have a different gray level. In the halftone process, threshold modulation is only applied to the region which has the same gray level and at the same time the input dependent error diffusion coefficients which have four directions are used to process the halftone error. Experimental results show that our algorithm can reduce visual artifacts effectively and outperforms Floyd-Steinberg's error diffusion, Ostromoukhov's error diffusion and Bingfeng Zhou's error diffusion in the PSNR(Peak Signal Noise Ratio),WSNR(Weighted Signal Noise Ratio) and UIQI(Universal Image Quality Index) criteria.

Key words: image segmentation, error diffusion, digital halftone