carrier image

ENTKERNER: A System for Removal of Globally Invisible Triangles from Large Meshes

Ernst, Manfred, Frank Firsching and Roberto Grosso

Proceedings, 13th International Meshing Roundtable, Williamsburg, VA, Sandia National Laboratories, SAND #2004-3765C, pp.449-458, September 19-22 2004

IMR
PROCEEDINGS

13th International Meshing Roundtable
Willimasburg, Virginia, USA
September 19-22, 2004

Manfred Ernst and Frank Firsching
Bytes+Lights GmbH, Erlangen, Germany.
{manfred.ernst, frank.}rschingg@bytes-lights.de

Roberto Grosso
Computer Graphics Group, University of Erlangen
grosso@informatik.uni-erlangen.de

Abstract
We present a method that computes a global potentially visible set for the complete region outside the convex hull of an object. The technique is used to remove invisible parts (triangles) from complex tessellated CAD models. Such optimizations are required to achieve interactive frame rates for the visual exploration of huge data sets on graphics workstations. Our algorithm is subdivided into three stages. At the rst step the the tesselated object is rendered with OpenGL from various camera positions to detect visible triangles very fast. A hardware-accelerated hemicube test is applied as the second stage, marking all triangles that can directly exchange energy with an infinitely distant environment. Finally, a Monte Carlo ray tracing pass is applied to each remaining triangle, sampling its visibility with arbitrary accuracy. All invisible triangles are completely removed from the mesh. It is therefore not necessary to store visibility information, allowing the reduced mesh to be processed and displayed by any software.

Download Full Paper (PDF Format)


Contact author(s) or publisher for availability and copyright information on above referenced article