Predictive Lazy Amplification: Synthesis and Rendering of Massive Procedural Scenes in Real Time
Departamento de Ciencia da Computacao, Universidade Federal de Minas Gerais, Belo Horizonte, Brasil
23rd SIBGRAPI Conference on Graphics, Patterns and Images (SIBGRAPI), 2010
@inproceedings{cordeiro2010predictive,
title={Predictive Lazy Amplification: Synthesis and Rendering of Massive Procedural Scenes in Real Time},
author={Cordeiro, C.S. and Chaimowicz, L.},
booktitle={2010 23rd SIBGRAPI-Conference on Graphics, Patterns and Images},
pages={263–270},
year={2010},
organization={IEEE}
}
In this paper we propose a new paradigm for procedural modeling that enables the real time visualization of massive procedural scenes. For this, we use a combination of memory and task management with two well known procedural modeling paradigms: data amplification and lazy evaluation. Experimental results show that, in addition to obtaining performance gains through parallelism, the implemented system can generate and visualize a procedural scene far greater than the available memory in real time using only a single PC equipped with a GPU and a multicore processor.
August 11, 2011 by hgpu