摘要
激光扫描数据多站配准是扫描数据处理中的关键环节。实际测量中常用定标球实现多站数据配准。该配准方法的精度决定于球心的计算精度,为提高球心的计算精度,本文提出将RANSAC算法用于球心计算。该算法鲁棒性很好,能从含有野值的数据中准确计算出球心。最后通过实验验证了本文的方法具有较高的精度。
Multiple views merging is a key step in point clouds processing,and the Registration-Sphere-Method is widely used in real work,so this paper put great effort on the study of this method.The merging accuracy relies on the accuracy of the sphere center.Aiming at raising the accuracy,the author used RANSAC algorithm to work out the sphere center.RANSAC is a robust algorithm,so the center could be calculated accurately from the points which have errors in them.Experiments and comparisons showed the high accuracy of this algorithm.
出处
《测绘科学》
CSCD
北大核心
2010年第5期58-59,148,共3页
Science of Surveying and Mapping
基金
测绘学院硕士学位论文创新与创优基金
关键词
多站配准
RANSAC
球面拟合
公共点坐标转换
multiple view registration
RANSAC
sphere fitting
common point transformation