Detection of Collision and Self-Collision Using QPSO for Deformable Models

In order to improve the speed of deformable objects collision and self-collision detection, proposing a new kind of stochastic collision method which is based on Quantum-behaved Particle Swarm Optimization. In this new algorithm the collision detection problem is treated as a kind of problem which is similar with Dynamic and Multi-objective Optimization Problem(MOP) where as many of the collision pairs satisfying the collision conditions are detected in certain time interval, notice that the detected collision pairs are not necessarily the globally optimal solution. For this problem that is similar with MOP, the iteration searching process for quantum-behaved particle has been optimized, in this algorithm, once a new collision pair satisfying the condition is detected, then the next searching will be converge towards the latest detected collision pair who satisfying the condition. This strategy significantly improved the searching ability for the satisfied collision pairs detection in the limited time interval, and there is no need for clustering, merger, division and any other operations, experiment results show that the efficiency of this algorithm is much better than the similar algorithms.