Much of the research in computer graphics in the last twenty years has been directed toward achieving photorealism for increasingly complex models. However, another equally valid approach to handling complexity is abstraction, a classic example of which is illustration. Indeed, in many applications, such as architectural, industrial, automotive, and graphic design, and whenever effective communication is the goal, illustration has certain advantages over photorealism. Illustrations convey information better, consume less storage, are more easily reproduced, are more capable of conveying information at various levels of detail, and are in many respects more attractive than photorealistic images.
The focus of this project is to develop techniques for creating non-photorealistic illustrations as a means for handling complexity. In particular, we are investigating algorithms for multiresolution painting, multiresolution curve editing, interactive digital halftoning, and architectural rendering.
Principal investigators: DeRose, Salesin