By Marilee DeGoede
College of Engineering and Applied Sciences
The “Outstanding Student Paper” Excellence Award was received by the Downtown Denver Campus’ Sunhwa Jung, Min Hong, and Min-Hyung Choi, Department of Computer Science and Engineering, for the paper “An Adaptive Collision Detection and Resolution for Deformable Objects Using Spherical Implicit Surface” at the fifth International Conference on Computational Science (ICCS). The event was May 22-25 at Emory University in Atlanta, Georgia.
The ICCS Technical Program featured four keynote speakers; talks from representatives of government, funding agencies, and industry; and seven invited tutorials. The conference was attended by more than 1,000 scientists, researchers, students and industry representatives from computing disciplines and provided a forum for discussing computational science research. The main conference included 134 contributed papers selected from 464 submissions; this article was one of the six award-winning papers selected by the conference committee.
Simply stated, the paper describes research based on the computer modeling of deformable objects, of which cloth is an example, and the paper describes a realistic simulation of falling cloth, folding cloth, and crumpled cloth.
(a)
(b)
Simulation snapshots showing cloth falling down [(a) and cloth folded and crumpled (b).
Less simply stated, the paper describes the collision management theory used to model interactive simulations of deformable objects. For example, the collision of two cylindrical volumes (that are meshed in tetrahedral) is shown below in (c).
(c)
The deformation of two cylindrical volumetric models in collision (c).
Because the shape and topology of a deformable object changes continuously, unlike that of a rigid body, the challenge is to find computational collision detection and contact resolution schemes. Human perception is used to determine the necessary accuracy of the collision detection. By using a new, enhanced, spherically implicit surface hierarchy, Jung et al can adjust the criteria for penetration depth and separating distance, depending on application-specific error tolerance. Their comparative experiments show that the proposed method performs substantially faster than existing algorithms for deformable object simulation with collisions at each iteration step. This hierarchical approach enables the achievement of a real-time simulation rate that is well suited for interactive applications.
In “real world” application, this type of simulation of deformable objects is indispensable for modern character animation and medical simulation.
The work described in the paper is part of Jung’s recently completed master’s thesis on computer graphics. Jung will begin working on his PhD this fall to continue his studies in the area of computer modeling and simulation. Co-author Hong is also a PhD student at UCD; his area of specialty is bioinformatics and he is currently studying in the Bioinformatics Program at Health Sciences Center. Co-author Choi is faculty advisor to Jung and Hong; he focuses his research on modeling and simulation of deformable objects.