Dynamic real-time deformations using space & time adaptive sampling

• DOI: 10.1145/383259.383262
• Publication Date: 2005-10-12T13:24:07Z
• Authors: Gilles Debunne; Mathieu Desbrun; Marie-Paule Cani; Alan H. Barr
• Keywords: Adaptive sampling

This paper presents a robust, adaptive method for animating dynamic visco-elastic deformable objects that provides guaranteed frame rate. Our approach uses novel automatic space and time level of detail technique, in combination with large-displacement (Green) strain tensor formulation. The body is partitioned non-nested multiresolution hierarchy tetrahedral meshes. local resolution determined by quality condition indicates where when the too coarse. As object moves deforms, sampling refined to concentrate computational load into regions deform most. model consists continuous differential equation solved using explicit finite element method. We demonstrate our Green formulation suppresses unwanted artifacts behavior, compared mass-spring other approaches. In particular, damped elastic vibration modes are shown be nearly unchanged several levels refinement. Results presented context virtual reality system. user interacts real-time through control rigid tool, attached haptic device driven forces derived from