DocumentCode :
3325901
Title :
Accelerating Line-of-Sight computation with structure guided hybrid approach
Author :
Zhu, Jiejie ; Hai, Wei ; Zhang, Lei ; Liu, Xiaoqing ; Yadegar, Jacob ; de La Cruz, Jorge ; Gonzalez, Hector J.
Author_Institution :
UtopiaCompression Corp., Los Angeles, CA, USA
fYear :
2012
fDate :
12-14 Jan. 2012
Firstpage :
103
Lastpage :
106
Abstract :
This paper proposes a hybrid Line-of-Sight (LoS) computation method for terrains and terrain surfaces, such as buildings and trees, represented using different formats. Unlike the unified triangular terrain representation used in 3D games, the representation of terrains and features is different in most of government training systems. The proposed hybrid approach improves LoS performance around 50% by successfully reducing several queries into one. We extend the state-of-the-art surface area heuristic in constructing a bounding volume hierarchy guided by terrain structures. Experimental results are tested from 10 datasets.
Keywords :
computer graphics; image texture; stereo image processing; terrain mapping; 3D games; bounding volume hierarchy; buildings; government training systems; hybrid line-of-sight computation method; structure guided hybrid approach; surface area heuristic; terrain structures; terrain surfaces; trees; unified triangular terrain representation; Acceleration; Algorithm design and analysis; Clocks; Computational modeling; Ray tracing; Training; Vegetation; Bounding Volume Hierarchy; Hybrid; Line-of-Sight; Structure Guided; Surface Area Heuristic;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Emerging Signal Processing Applications (ESPA), 2012 IEEE International Conference on
Conference_Location :
Las Vegas, NV
Print_ISBN :
978-1-4673-0899-1
Type :
conf
DOI :
10.1109/ESPA.2012.6152456
Filename :
6152456
Link To Document :
بازگشت