Abstract
The aim of this paper is automatic generation of painting-style images from photographic images. Several work has been done for modeling painting tools and materials so that strokes of brushes can be automatically generated from input images. However, when strokes arc overlapped, colors of overlapped regions are decided without considering gradations between different colors. In this paper, therefore, we propose a graze model for deciding colors of overlapped strokes by considering optical transmission characteristic of paints and human vision properties.
