This document describes four projects that, taken together, serve as a proof by example of the efficacy of a perception-motivated strategy for making graphics tools. More specifically, this strategy involves first selecting local features motivated by human perception and domain understanding, and then building algorithms that allow users to interactively edit these features. While the tools presented herein range over the domains of painting, compositing, and animation, all share a common philosophy &ndash- they spring from the isolation of a basic task-relevant feature and the creation of support algorithms that allow local edits of this feature at interactive rates.

Gradient Paint is a paint program inspired by theories of color perception which allows artists to edit edges between colors rather than colors themselves. The system includes a new real-time integrator for image display and a set of brushes suited for working with edges.

Local Layering introduces a notion of local overlap which allows artists to stack images as if they were paper cut-outs. In this work, I present a new way of representing stacked images and provide a provably correct and sufficient set of operators for navigating this representation.

Soft Stacking is a continuous-domain extension of Local Layering which allows artists to stack images as if they were volumes of fog. I present both brush-based and optimization-based creation techniques for such continuous-domain stackings.

3D-like Texturing for 2D Animation is an attempt to provide 2D artists with a way to reap one of the primary benefits @ndash; coherent texturing – of creating 3D models. The system I describe, though imperfect, is able to construct a correspondence between various 2D drawings using only local shape information, and in such a way that the computer never creates a 3D model.

These four projects showcase a new, and somewhat effective, strategy for creating graphics tools that transcend traditional art materials and computer constructs.

102 pages

Return to: SCS Technical Report Collection
School of Computer Science

This page maintained by reports@cs.cmu.edu