Abstract:

In this paper, a hybrid collision detection algorithm based on the combination of AxisAligned Bounding Boxes and Oriented Bounding Box is proposed to achieve the purpose of improving the realtimeness of collision detection. Firstly, we determine whether the objects are intersection with the object projection, and detect the objects which can not be intersection. Secondly, we project on the objects which may intersect and reversally construct AxisAligned Bounding Boxes with a quick test. Finally, we do the accurate detection with Oriented Bounding Box to achieve efficient collision detection of objects in virtual environments. This algorithm is also improved about the lowlevel efficiency of the traditional overlapping test between the primitiVes. The experimental results show that compared with other two algorithms, the method can effectively improve the efficiency of collision detection in the same environment.

Key words: virtual reality;axisaligned bounding boxe;oriented bounding box;hybrid collision detection;overlapping test between the primitives