Title
A Fast Method for Real-time Computation of Approximated Global Illumination
Abstract
We present a fast method for real-time computation of approximated global illumination for fully dynamic scenes under area light sources. To accelerate the computation, we use simplified models to calculate the indirect illumination, while render the direct illumination with original complex models. After direct illumination is computed with convolution soft shadow maps algorithm, color, position and normal textures are generated, from which a Halton quasi-random sampling method is used to produce the pixel lights for the second bounce. Our testing models show that thousands of passes can be rendered with the simplified scenes in one second for the indirect illumination, and it can dramatically improve the frame rate for relatively complex scenes. Final image is generated by blending direct image and all indirect images.
Year
DOI
Venue
2009
10.1109/CGIV.2009.24
CGIV
Keywords
Field
DocType
pixel,real time systems,random sampling,image texture,lighting,global illumination,mesh generation,real time computing
Computer vision,Convolution,Computer science,Image texture,Shadow mapping,Pixel,Global illumination,Frame rate,Artificial intelligence,Mesh generation,Computation
Conference
Volume
Issue
Citations 
null
null
0
PageRank 
References 
Authors
0.34
21
4
Name
Order
Citations
PageRank
Weiwei Lv131.49
Jian Lu200.68
Xue-Hui Liu325626.39
Enhua Wu4916115.33