Real-time tonal depiction method by reaction-diffusion mask

摘要

Various methods such as stippling, hatching, or hedcut are being used to express the brightness of images as a tone. In this paper, as a new method of tonal depiction, we use the mathematical model used in biology for the purpose of expanding the area of non-photorealistic rendering. The model of reaction-diffusion, which is used to depict the skin patterns of diverse animals, has relatively recently been used as the technique of NPR. The biggest obstacle in processing this technique in real time is that it is very time-consuming process of the repeated calculation. In this contribution, we proposed how to build a mask creating similar results in order to gain instant results from the brightness of images. Mask is a two-dimensional table, which is beforehand calculated and used in digital halftoning, can process each pixel of pictures and produce the quickest result. Although the mask obtained as a result uses a uniform size repeatedly throughout the whole images, it has the merit of the repeated parts not being exposed because of the continuity of the connected parts. As the result, we compare the images through the masks made by both the existing method by repetition and the suggested method.