The Research On Fast Collision Detection Based On Oriented Bounding Box

Collision detection has been researched in many fields such as robot motion planning and computer graphics for a long time. In the later years, with the rising of virtual reality and distributed interactive simulation, many researches focus on collision detection and Exact collision detection is very important to improve reality and enhance immersion for virtual environment, and the complexity and real-time of virtual environment bring new requirement to collision detection.Bounding volume hierarchy provides an effective method to resolve the intrinsic time complexity in collision detection. The idea behind it is to approximate the object with a simpler bounding volume that is a little bigger than the object. In building hierarchies on object, one can obtain increasingly more accurate approximations of the objects. So during traversing bounding volume hierarchy, it speeds up collision detection by prune away primitive pairs, which will not intersect clearly though rapid intersection test between bounding volumes and just deal with those whose bounding volume is intersected. The choice of bounding volume is the basic and key problem of this approach.This thesis focuses on the rigid bodies of large-scale complex environment as the research target and emphasizes on the relevant problems of OBB collision detection algorithm. Based on the full research of inherent characters of OBB hierarchical box, this thesis implements a series of key techniques such as building OBB and OBB tree, and overlap test between OBB boxes, verifies the attribute that OBB has the advantage that it is fit for the collision detection between rigid bodies and optimizesthe shortcomings and disadvantages of OBB. Based on the limitations of OBB, this thesis proposes an optimizing overlap test algorithm of bounding box and primitive. This method eliminates the bounding boxes of leaf nodes and streamlines the OBB hierarchical layer. This method also introduces a separating axis theory in the overlapping test between the primitives in order to accelerate the velocity. By scientifically arranging groups of overlapping methods, this thesis proposes an optimizing hierarchical hybrid algorithm based on OBB to enhance the reliability and actual running efficiency.