Research On Collision Detection Algorithm Based On Bounding Volume And Spatial Subdivision

Fast and accurate collision detection is very important to improve immersion and authenticity for virtual environment. Because there exists a large number of objects and the complexity of the geometry, collision detection process often spend a lot of storage space and processing time, real-time and authenticity of virtual environment are determined by the efficiency of collision detection. Collision detection is always the hard question in virtual reality.There are mainly two algorithms of collision detection in virtual reality: hierarchical bounding volume algorithm and space subdivision algorithm. The purpose of these two algorithms is to reduce the number of pairs of objects or primitives that need to be checked for contact. To improve the efficiency of collision detection, on the basis of our in-depth research, we choose the following as our research direct:Due to the convenient construction and simple intersection test for AABB, this paper deals with the problem of collision detection from it. Based on the nature of AABB and construction process of AABB tree, we introduce a method that can reduce the amount of byte of AABB tree for internal node, so the algorithm is speeded up.We propose a uniform spatial subdivision approach to collision and self-collision detection of deformable objects. The proposed approach employs hash function as data structure, and tetrahedron meshes as the primitives. We experiment the parameter of the collision detection algorithm, such as hash function, hash table size, spatial cell size. Our approach has been implemented based on tetrahedron, however, it is not restricted to tetrahedrons and could handle other object primitives.