摘要
古建筑保护的重要性已越来越为人们所认识,针对古建筑的支撑性部件提取效率低、精度差等问题,文章首先采用全局八叉树和局部K-D树的二级空间索引的方法将点云模型数据进行空间划分,然后利用随机采样一致性(RANdom Sample Consensus,RANSAC)改进算法结合线性最小二乘拟合算法提取古建筑的几何特征面,最后以庐山天主教堂三维激光扫描数据为例,验证了该方法的有效性、高效性、精确性。
The importance of ancient architecture protection has gradually been recognized.For the problems of low extraction efficiency and poor precision of supporting components of ancient architecture,this paper firstly uses the second-order spatial index of global Octree and local K-D tree to divide the point cloud model data into space,and then uses the RANSAC to improve algorithm and combine the linear least squares fitting algorithm to extract the geometric features of the ancient architecture,final y takes the 3 D laser scanning data of Lushan Catholic Church for example to verify the effectiveness,efficiency and accuracy of the proposed method in this paper.
作者
黄慧敏
朱雪辉
HUANG Hui-min;ZHU Xue-hui(Jiangxi Institute of Basic Surveying and Mapping,Nanchang Jiangxi 330209,China)
出处
《地矿测绘》
2020年第2期13-15,41,共4页
Surveying and Mapping of Geology and Mineral Resources
关键词
古建筑
点云模型
三维激光扫描
特征提取
ancient architecture
point cloud model
3D laser scanning
feature extraction
