基于优化Snake模型的变形物体碰撞检测算法研究  

Collision Detection Algorithm of Deformable Object Based on Snake Model Optimization

作　　者：李照[1] 靳雁霞[1] 秦志鹏[1] 任超 

机构地区：[1]中北大学计算机与控制工程学院,山西太原030051

出　　处：《系统仿真学报》2018年第1期62-68,共7页Journal of System Simulation

基　　金：国家自然科学基金(61379080);山西省自然科学基金(2013011017-7)

摘　　要：针对现有碰撞检测算法难以解决变形物体碰撞的实时性和逼真性问题,提出一种基于可变形物体的改进碰撞检测算法。该算法在K-dops(K-discrete orientation polytopes)包围盒基础上使用Snake模型应用于包围盒更新过程中,使用改进的Pso(Particle swarm optimization)算法对Snake模型进行边界优化。在优化过程中依据碰撞部位的内凹特性,在Snake模型控制点簇与中心点构成的多直线上,采用改进的粒子选取方式和多种群粒子优化算法思想构成多直线群体。实验结果表明,改进后的碰撞检测算法可以快速准确的模拟物体轮廓,简化了包围盒更新过程,提高了碰撞检测效率。In view of the existing collision detection algorithm which is difficult to solve the collision of real-time and reality, an improved algorithm of collision detection was propsed based on deformable objects. The algorithm based on K-DOPs bounding box applied to update process used the improved PSO algorithm to optimize the Snake model boundary. The concave in collision region, the control point clusters and the center formed multi-linear swarm with the improved selection method and the ulti PSO algorithm. Experimental results show that the improved collision detection algorithm can accurate simulate object contour, simplify the bounding box to improve the efficiency of collision detection.

关 键 词：碰撞检测 K-dops包围盒 多直线群体 PSO算法 SNAKE模型 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]