基于法向量的多视点定标球球心算法  

Normal-based Algorithm for Registration Sphere Center in Multiple Views

作　　者：魏江[1] 熊邦书[1] 冯燕[1] 李龙[1] 

机构地区：[1]西北工业大学电子信息学院,西安710072

出　　处：《计算机工程与应用》2005年第19期15-17,共3页Computer Engineering and Applications

基　　金：国家自然科学基金(编号:60172040);国家863高技术研究发展计划项目(编号:2001AA131022);西北工业大学青年教师创新基金(编号:M016141)

摘　　要：逆向工程中常用基于定标球的方法实现多视点合并,定标球球心的计算精度决定了合并的精度。该文提出了一种基于法向量的定标球球心算法,利用扫描数据点的法向量等相关性信息和视点扫描特性,从扫描点云数据中提取出定标球的球面数据点,从而准确计算出定标球球心。大量真实数据对比实验表明,算法具有较高的精度和较好的稳定性。Multiple-view merging method based on registration sphere is often used in reverse engineering.The merging accuracy relies on the calculating accuracy of registration sphere center.A normal-based algorithm of registration sphere center is presented in this paper.The points of sphere surface are filtered from all scanned point cloud data according to the normals of data points and characteristic of viewpoint,then the center of registration sphere can be calculated accurately.Experiments and comparisons on real data shows the high accuracy and good stability of this algorithm.

关 键 词：法向量 定标球 多视点 点云数据合并 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]