摘要
针对涡轮叶片密集点云数据与计算机辅助设计模型配准速率慢、耗费时间长的问题,提出基于简化点云带动涡轮叶片快速配准的方法。基于叶片的三维光学扫描数据和计算机辅助设计模型,采用均匀采样法和曲率采样法简化点云模型数据,通过最近点迭代和奇异值分解相结合的配准方法,实现叶片密集点云模型与计算机辅助设计模型的快速配准,并对配准精度和速度进行了统计分析。结果表明,基于均匀采样法简化的点云数据模型带动的配准,在保证配准精度的前提下能有效提高配准效率。
To solve the slow and long time-consuming problem of registration between dense cloud data and Computer Aided Design（CAD） model,a simplified point cloud data based rapid registration method of turbine blade was proposed.Based on three-dimensional optical scanning data and CAD model of turbine blade,point cloud data was simplified by using uniform sampling method and curvature sampling method.Through Singular Value Decomposition Interactive Closest Point（SVD-ICP） method,the rapid registration between turbine blade’s dense cloud model and CAD model was realized,and accuracy as well as speed of registration were analyzed.The results indicated that the proposed method could increase the efficiency of registration without influencing registration accuracy.
出处
《计算机集成制造系统》
EI
CSCD
北大核心
2012年第5期988-992,共5页
Computer Integrated Manufacturing Systems
关键词
点云数据
光学扫描
均匀采样
精确配准
涡轮叶片
cloud data
optical scanning
uniform sampling
accurate registration
turbine blades
