Collision Detection Algorithm Based On Bounding Box Of The Research And Application

As one important part of Virtual Reality(VR), Collide Detection(CD) has beenresearched in many fields such as computer graphics, computer animation, robotics andso on with the development of VR. The recently research indicate that the requirementsfor real-time of collision detection have become higher than before, because theperceived quality of most interactive3D applications does not depend on exactsimulation, but rather on real-time response to collisions.In order to improve the requirement of real-time of CD, this paper researched animportant branch of CD algorithm-hierarchical bounding volume(HBV)algorithm onthe premise of keeping the accuracy of CD. The main contents and innovational pointsare shown as follows:Firstly, this paper made a summary introduction of CD algorithms used frequently.Among these algorithms, the HBV method was mentioned especially. And it is thereason for we chose the bounding volumes(BV)in the algorithm of this paper when themerits and shortcomings of the BVs were enumerated.Secondly, we researched deep into the RAPID algorithm, which was a HBV treealgorithm based on oriented bounding box(OBB)bounding volume. And we optimizedthe structure of HBV tree and the construct of OBB bounding volume in RAPIDalgorithm. The experiments showed that the optimized algorithm was better in the scenewhich had two or more objects compared with RAPID algorithm, because the optimizedalgorithm could reduce the amount of intersection tests between OBBs in the scene ofhaving two objects and exclude the objects which were apart from each other quickly inthe scene of having more than two objects.Thirdly, we researched deep into the SOLID algorithm, which was a HBV treealgorithm based on axis-aligned bounding box(AABB)bounding volume. And weoptimized the storage of the HBV Tree’s nodes in SOLID algorithm. The experimentsshowed that compared with SOLID algorithm, the optimized algorithm not onlyreduced the time of CD but also saved the memory space. And if it was applied in thescene which had complex objects we could get better results. Finally, the production of CD algorithm was applied to item--Network attacksub-munition(NAS), and the result proved that the optimized algorithm based on AABBbounding volume could resolve the real-time of CD in system simulation on the premiseof accuracy of CD was not destroyed, and this optimized algorithm also gave aneffective method which could apply the CD algorithm into the item exploitationeffectively.