Squashing Cubes: Automating Deformable Model Construction for Graphics - Robotics Institute Carnegie Mellon University

Squashing Cubes: Automating Deformable Model Construction for Graphics

Doug James, Jernej Barbic, and Christopher D. Twigg
Conference Paper, Proceedings of ACM SIGGRAPH Conference on Sketches & Applications, August, 2004

Abstract

The vast majority of geometric meshes used in computer graphics are optimized for rendering, and not deformable object simulation. Despite tools for volume (or surface) (re)meshing of geometric models to support physical simulation, in practice, the construction of physically based deformable models from arbitrary graphical models remains a tedious process for animators. Squashing Cubes automates the construction of physically based deformable objects from arbitrary geometric models. During preprocess, the geometric model (typically a surface mesh) is voxelized into tiny elastic cubes, i.e., the (squashing cubes model). Second, a generic deformable object simulator is used to deform the squashing cubes model. Finally, the resulting deformations are interpolated back onto the original model, thus producing the final animation.

Notes
AVI Video available at: http://graphics.cs.cmu.edu/projects/squashingCubes/squashingCubesSig04_divx.avi

BibTeX

@conference{James-2004-9006,
author = {Doug James and Jernej Barbic and Christopher D. Twigg},
title = {Squashing Cubes: Automating Deformable Model Construction for Graphics},
booktitle = {Proceedings of ACM SIGGRAPH Conference on Sketches & Applications},
year = {2004},
month = {August},
publisher = {ACM Press},
}