古建筑三维点云数据的精简与配准方法  

3D point cloud data simplification and registration method of ancient buildings

作　　者：刘冉 刘军廷 LIU Ran;LIU Junting(Ningbo Metallurgical Survey and Design Research Company Limited,Ningbo Zhejiang 315000,China)

机构地区：[1]宁波冶金勘察设计研究股份有限公司,浙江宁波315000

出　　处：《北京测绘》2022年第7期881-885,共5页Beijing Surveying and Mapping

摘　　要：针对古建筑点云数据量大,配准效率低,质量差的问题,提出古建筑三维点云数据精简与配准的方法研究。首先对获取的古建筑源点云和目标点云数据通过曲率与包围盒随机抽样相结合的点云精简方法缩小点云的数据量,点云数据精简既保留了古建筑点云中的特征点,又使平坦区域的点云得到了均匀分布;然后从精简点云数据中提取四点集,利用超四点快速鲁棒匹配算法(super four point robust matching algorithm,Super 4PCS)实现源点云和目标点云的粗配准;最后利用最近点迭代算法(iterative closest point,ICP)实现点云间的精配准。实验结果表明,本文方法有效地提高了点云配准效率和配准精度。Aiming at solving with the problems of large amount of point cloud data,low registration efficiency and poor quality of ancient buildings,a method of simplification and registration of three-dimensional(3D)point cloud data of ancient buildings was proposed.Firstly,the data of the source point cloud and the target point cloud of the ancient buildings were reduced by the point cloud simplification method combining the curvature and the random sampling of the bounding box.The point cloud data simplification not only retained the feature points in the ancient building point cloud,but also made the point cloud in the flat area uniformly distributed.Then the four-point consistent set was extracted from the simplified point cloud data,and the super four-point fast robust matching algorithm(Super 4PCS)was applied to realize the coarse registration of source point cloud and target point cloud.Finally,the iterative closest point(ICP)algorithm was proposed to achieve fine-tuning between point clouds.The experimental results showed that the proposed method effectively improved the efficiency and accuracy of point cloud registration.

关 键 词：曲率-包围盒随机抽样 点云数据精简 古建筑点云配准 

分 类 号：P237[天文地球—摄影测量与遥感] TP391.41[天文地球—测绘科学与技术]