单站地面激光点云中孤立噪声的快速检测方法  

A Fast Isolated Noise Detection Method for Single Station Point Cloud in Terrestrial Laser Scanning

作　　者：殷义程 石宏斌 王利华 YIN Yi-cheng;SHI Hong-bin;WANG Li-hua(Surveying and Mapping Engineering Institute of Yunnan Province, Kunmin Yunnan 650033, China;School of Urban-Rural Planning andLandscape Architecture, Xuchang University ,Xuchang Henan 461000, China)

机构地区：[1]云南省测绘工程院,云南昆明650033 [2]许昌学院城乡规划与园林学院,河南许昌461000

出　　处：《地矿测绘》2018年第1期6-8,共3页Surveying and Mapping of Geology and Mineral Resources

基　　金：河南省高等学校重点科研项目(1713420001);河南省科技攻关项目(182102310924)

摘　　要：提出了一种基于遮挡边界检测的单站激光点云孤立噪声检测方法。该方法首先根据单站点云数据的栅格结构,计算任意非空点间的邻域距离;然后根据突变阈值确定遮挡边界点,同时对遮挡边界范围内的点在三维空间中进行空间聚类;最后根据预设点集数目,确定孤立噪声点集。实验结果表明,该方法能够快速确定单站地面激光扫描数据集中的孤立噪声。This paper proposed an isolated noise detection method for single station point cloud in terrestrial laser scanning based on occluding edge detection. The method firstly calculated the distances between the non-null point and its neighbors in 2 D raster structure determined by the single station; Then the occluding points were determined by the threshold and clustered in 3 D space; Finally the isolated noise sets were determined by the predefined number of the set.The experiment shows that the method proposed in this paper can determine the noise sets in single station point cloud fast and efficiently.

关 键 词：地面激光三维扫描 噪声 遮挡边界 空间聚类 单站点云 

分 类 号：P204[天文地球—测绘科学与技术] TP391.41[自动化与计算机技术—计算机应用技术]