Optimized Pattern-Based Adaptive Mesh Refinement Using GPU

Ricardo Lenz, Joaquim Bento Cavalcante-Neto, Creto Augusto Vidal
Federal University of Ceara, Fortaleza, CE, Brazil
XXII Brazilian Symposium on Computer Graphics and Image Processing (SIBGRAPI), 2009


   title={Optimized Pattern-Based Adaptive Mesh Refinement Using GPU},

   author={Lenz, R. and Cavalcante-Neto, J.B. and Vidal, C.A.},

   booktitle={Computer Graphics and Image Processing (SIBGRAPI), 2009 XXII Brazilian Symposium on},





Download Download (PDF)   View View   Source Source   



The high performance of GPUs and the increasing use of their programming mechanisms have fostered the development of graphics applications that better exploit the raw power of these devices to achieve higher levels of realism. Silhouette refinement, as one of the techniques that help to improve realism, has profited from GPUs’ advances in recent years. In this paper, we present a method for triangular mesh refinement which alleviates the problem of rugged silhouettes. We demonstrate that, through a clever indexing scheme, our method is able to use adaptive patterns in an optimized way, taking full advantage of the GPU’s parallelism. The ideas we used were adapted from distinct previous works, but our method presents astounding performance gains. Also, our method works very well with existing meshes, therefore, it can improve the visual appearance of existing models without mesh redesign.
No votes yet.
Please wait...

* * *

* * *

HGPU group © 2010-2021 hgpu.org

All rights belong to the respective authors

Contact us: