Comic-Strip Rendering

CS-TR-95-2

Author: Peter Hall
Source: GZipped PostScript (439kb); Adobe PDF (384kb)


Photorealistic rendering methods have been at the centre of much computer graphic research and remain in that position to date. The aim of photorealism is to produce computer graphic images that cannot be distinguished from photographs. Such methods have produced impressive results, and in some cases the aim has been met. However, photorealism is not always a desirable goal, sometimes a hand-drawn quality to the image is wanted and this requires a non photorealistic renderer. Non photorealistic rendering methods have received very little attention, our work addresses this gap. We have developed a method of rendering images in a style often used in popular comic strips. Our renderer accepts three dimensional models and lighting information as input. Initially, the scene is lit using standard techniques. Next shadow effects are drawn directly on to the models and these are rendered without a lighting model. The contribution our work makes is to introduce a general mechanism by which a new, non-photorealistic, style of rendering can be achieved. We demonstrate this with a particular mechanism that produces images which resemble the drawings that appear in many popular comic-strips. This paper describes our method.

[Up to Computer Science Technical Report Archive: Home Page]