Graphics, Imaging, and Games Lab

Image abstraction traditionally eliminates texture, but doing so ignores the more elegant alternative of texture indication, e.g., suggesting the presence of texture through irregular silhouettes and locally chosen details. We propose a variant of geodesic image filtering which preserves the locally strongest edges, leading to preservation of both strong edges and weak edges depending on the surrounding context.

Our contribution is to introduce cumulative range geodesic filtering, where the distance in the image plane is lengthened proportional to the color distance. We apply the new filtering scheme to abstraction applications and demonstrate that it has powerful structure-preserving capabilities, especially regarding preservation and indication of textures.