DocumentCode
1905061
Title
Evaluation of splitting methods in top-down OBB-tree construction
Author
Diktas, Engin Deniz ; Sahiner, Ali Vahit
Author_Institution
Dept. of Comput. Eng., Bogazici Univ., Istanbul
fYear
2008
fDate
27-29 Oct. 2008
Firstpage
1
Lastpage
6
Abstract
The Oriented Bounding Box trees (OBB-trees) offer efficient means for real-time collision detection between geometric models. The efficiency of the OBB-tree in answering the queries depends mainly on the way the triangles are splitted and distributed to the child nodes. In this paper we examine four different partitioning methods for top-down construction of OBB-trees and compare their performance in three different types of queries: collision detection between two objects, finding the triangle closest to a given point and extremal point queries. In addition to the examined methods we also offer a method for determining and splitting disconnected sets of triangles at an early stage during the hierarchy construction.
Keywords
trees (mathematics); collision detection; geometric models; hierarchy construction; oriented bounding box trees; partitioning methods; splitting methods; top-down OBB -tree construction; Geometry; Haptic interfaces; Object detection; Performance evaluation; Solid modeling;
fLanguage
English
Publisher
ieee
Conference_Titel
Computer and Information Sciences, 2008. ISCIS '08. 23rd International Symposium on
Conference_Location
Istanbul
Print_ISBN
978-1-4244-2880-9
Electronic_ISBN
978-1-4244-2881-6
Type
conf
DOI
10.1109/ISCIS.2008.4717855
Filename
4717855
Link To Document