Halftoning Algorithm For Color Images Based On Human Visual System's Model

Error Diffusion halftoning is an important method in digital halftoning. Most of color halftoning that does not take the correlation among color planes into account leads to artifacts such as worms. Vector error diffusion captures correlation among color planes by using an error filter with matrix-valued coefficients. A vector error diffusion halftoning algorithm for color images based on human visual system is proposed in this paper, against the damage of color relativity and get the output images of high quality. The idea of Pattern-Color separable is adaptive to the Human Visual System by analyzing its characters. The proposing method is based on a Vector Error Diffusion which the optimal error diffusion kernel can get through transforming the original image from RGB color space to linear CIELAB color space and filtering the image in different planes by two kinds of HVS filters.