摘要
对待测产品的测量点云和标准设计点云进行配准,得出产品制造偏差量,进行精度测量评定。先采用主元分析(PCA)法作预匹配,再用随机抽样(RANSAC)算法取重合度高的匹配点对,最后利用最近点迭代(ICP)算法得到高精度的点云配准。其中,采用RANSAC算法得到高重合度的匹配点对,便于得出最优空间坐标转换参数,使得配准精度更高;对抽样次数的估计可以推出点云配准迭代次数,进而有效减少运算时间。实验结果显示算法是有效的。
To match measurement point cloud of the products to be tested and the standard point cloud,manufacturing deviation value of products is obtained,so precision measurement can be evaluated. Firstly,principal component analysis is used for pre-matching. Next,by using random sample consensus the matching point pair of high contact ratio can be selected. Finally,closest point iterative is used for getting the point cloud registration with high precision. Using RANSAC to extract the matching point pair of high contact ratio is convenient to get the most optimal space coordinate transformation parameters,making higher registration accuracy. Estimating of the sampling frequency can infer point cloud registration iteration,and then may reduce operation time effectively. Experiment results show that the algorithm is effective.
出处
《电子器件》
CAS
北大核心
2015年第4期929-934,共6页
Chinese Journal of Electron Devices
基金
山西省青年基金项目(201002106-13)
关键词
产品制造
精度测量
主元分析法
随机抽样法
最近点迭代
products manufacturing
precision measurement
principal component analysis
random sample consensus
closest point iterative