3D Geometry Compression

Polyhedral models, which are used in most graphics applications, require considerable amounts of storage, even when they only approximate precise shapes with limited accuracy. This course analyzed storage costs for 3D shape representations and covered several recent schemes for lossy and loss-less compression of triangle, polygonal, and tetrahedral meshes. It reviewed both single-resolution compression schemes for triangle and polygonal meshes, which result in compressed formats of less than a byte per triangle, and multi-resolution progressive-refinement approaches.

Prerequisites
Some familiarity with basic concepts of meshes and standard representations of polygonal models.

Topics
Representation of polygonal models, single-resolution geometry compression schemes, progressive transmission of multi-resolution models, hierarchical geometry compression schemes, prediction schemes based on spectral methods, motion compression, compression of animated models, compression of tetrahedral meshes, hardware-oriented geometry compression, geometry compression in Java3D, geometry compression in MPEG-4.

Organizer
Gabriel Taubin
IBM T.J. Watson Research Center

Lecturers
Michael Deering
Sun Microsystems, Inc.

Craig Gotsman
Technion - Israel Institute of Technology

Stefan Gumhold
Universität Tübingen

Jarek Rossignac
Georgia Institute of Technology

Gabriel Taubin
IBM T.J. Watson Research Center