Modified Error Diffusion Algorithm for Edge Enhancement based on Local Contrast of Input Images

摘要

Error diffusion is a popular algorithm for halftoning of continuous-tone images. Traditional error diffusion often introduces worm artifacts and noise which drowns edges of images. Some modified error diffusions with changing scanning order, and with altering the quantizer input or changing error diffusion filter have been previously used to reduce worm artifacts and to enhance edges, respectively. This paper introduces a modified error diffusion algorithm which modulates threshold based on local contrast of input images to enhance edges and eliminates worm artifacts by adding random weight to the error diffusion filter. Comparing with some existing error diffusion algorithms, the results of the introduced algorithm show good visual properties.