Fast collision detection algorithm for deformable objects
Year: 2014
Type: article
Abstract: In order to achieve a real-time collision detection between deformable objects and rigid body in the virtual environment,a fast parallel collision detection algorithm based on hybrid bounding box hierarchy is proposed.Taking into account the different characteristics of the bounding box in the detection speed and accuracy,this algorithm builds a sphere-AABB hybrid bounding box hierarchy tree for deformable object,and establishes a sphere-OBB hybrid bounding box hierarchy tree for rigid body; in addition,every hybrid bounding box hierarchy tree is divided into upper layer,middle layer and lower layer,and each layer uses a different bounding box; when traversing the trees,this algorithm uses spheresphere test in the upper layer,sphere-OBB test in the middle layer and AABB-OBB test in the lower layer to detect collision.This paper uses multi-threading technology to achieve parallel collision detection algorithm on multi-core devices.Experimental results show that this algorithm has obvious advantages compared with the classic AABB algorithm in the aspect of efficiency,and can meet the requirement of collision detection between deformable objects and rigid body.
Author: Tan Rui-p
Institution: Jilin Agricultural University
Language: English
Field: Computer Science
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W2378576994
