Deformable Object Collision Detection Research Based On Parallel Technology

With the continuous development of computer graphics and the emergence of emergingfield including virtual reality, the request for real-time and accuracy is increasingly high.Simulation of the movement of the deformable body may be deformed by the action of forcesexperience. Under the influence of these external forces, deformation will happen somemovement, so it needs real-time collision detection in the movement process.At present, the rigid body collision detection algorithms have been quite mature, but for theresearch for deformable body collision detection is less. The collision detection algorithm ofexisting each has its advantages and disadvantages, the paper conducts the research on flexibleobject collision detection in virtual reality, the main contents of this paper are as follows:(1) In virtual reality, it should build the mode of deformation object before collisiondetection. Paper introduce several common Physical modeling，build the spring mass model,dynamics analysis, gives the numerical procedures, and adjust the constraint exists in modelingof flexible objects.(2) Collision detection technology for deformation objects of study, we chose the AABBbounding box hierarchy to study, detail introduced hierarchical bounding structure and update,hierarchical bounding box tree traversal and intersection tests and hierarchical bounding triangleintersection test. deformation objects collision will occur self-collisions and self-collisiondetection is more complex, so in this article we describes the most commonly used self-collisiondetection method that is based on self-collision algorithm to the cone.(3) Improving classic SOLID algorithm from the bounding box AABB collision detectiontime and storage space. At first, use a temporal correlation search strategy to accelerate thetraversal of AABB bounding box tree. Then, storage optimization for bounding box tree internalnode and a leaf node. The results show that the optimized algorithm not only reduces thecollision detection time, but also reduces the memory space.In this paper, we use the use multi threading technology which decompose the task and thesub-tasks can be performed on multi-core processors. The collision detection algorithm becomeparallel, thereby further improving the efficiency of collision detection.