Research on Parallel Collision Detection Algorithm for polyhedron Models

摘要

Collision detection among complex polyhedron models is an open question because of its algorithmic complexity. To satisfy the real-time collision detection requirement of polyhedrons, we presented a parallel algorithm to perform collision detection for polyhedrons models. The algorithm decomposes nonconvex polyhedron into tetrahedrons, then detect thetetrahedrons in parallel, and box tree of the polyhedron is built to improve the efficiency of the parallel algorithm. We detailed described the parallel algorithm based on box-tree, and give the experiment data. reseach result shows that the parallel method based on box tree is more efficient and scalable than the classic algorithm.