复杂场景中快速碰撞检测算法及GPU加速  

Fast algorithm for collision detection in complicated scene and GPU acceleration

作　　者：刘秀玲[1] 王冬雨[1] 陈栋[1] 刘京[1] 王洪瑞[1] 

机构地区：[1]河北大学电子信息工程学院,河北保定071002

出　　处：《计算机工程与设计》2012年第5期1847-1851,共5页Computer Engineering and Design

基　　金：国家自然科学基金项目(61074175);河北省自然科学基金项目(F2011201159)

摘　　要：为了保证在大规模复杂场景中,碰撞检测的实时性和精确性,提出了一种基于图形空间与改进的图像空间相结合,并利用GPU加速的快速碰撞检测方法。利用AABB包围盒的检测策略,快速剔除不相交物体,确定潜在碰撞对象。改进传统的基于图像空间的碰撞检测算法,设计了基于向指定平面投影、模板测试和深度测试的碰撞检测算法。在此基础上,利用GPU的并行计算能力加速整个检测过程,有效地减少了碰撞检测时间。通过在虚拟驾驶系统当中的应用,验证了该方法在大规模复杂场景中碰撞检测的实时性和精确性。To ensure the real-time performance and accuracy of collision detection in the large and complicated scene,a fast collision detection methord based on graphics space,improved image space and acceleration using GPU is put forward.First of all,the non-intersect objects are quickly removed and the potential collision objects are identified through the AABB bounding box testing strategy.Then,an improved collision detection algorithm is proposed which is based on projection to the designated plane,template test and depth test.On this basis,the testing process is accelerated by using the parallel computing capabilities-GPU,which effectively reduces the collision detection time.Finally,an application case of the virtual driving system validated the real-time performance and accuracy of collision detection in the large and complicated scene.

关 键 词：碰撞检测 AABB包围盒 模板测试 深度测试 可编程图形单元(GPU) 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]