Research on Parallel Collision Detection Algorithm for polyhedron Models

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.