基于离群点探测准则和主成分分析的点云平面拟合效果研究  

Point Cloud Surface Fitting Based on Outliers Detection Criterion and Principal Component Analysis

作　　者：丁鸽[1,2] 赵若鹏[1,2] 卞磊[1,2] 陈艳华 

机构地区：[1]济南市勘察测绘研究院,山东济南250013 [2]山东省城市空间信息工程技术研究中心,山东济南250013 [3]济南市房产测绘研究院,山东济南250001

出　　处：《测绘地理信息》2017年第3期25-28,共4页Journal of Geomatics

基　　金：国家自然科学基金资助项目(41174010;41374011)

摘　　要：针对三维激光扫描测量中离群值对点云质量的影响,提出了基于离群点探测准则和主成分分析的点云平面拟合方法,给出了方法的基本原理和实现步骤。实验结果表明,该方法较稳定,能够在不同条件下较为准确地剔除噪声,提高了点云拟合的平面精度。Aimed at the effect of outliers on the quality of point cloud in 3D laser scanning, point cloud surface fitting method based on outliers detection criteria and principal corn ponent analysis is proposed, and the basic principle and im- plementation steps of this method is given in this paper. Re- sults show that this method is more stable, can accurate to eliminate noise in different conditions, improve the fitting precision of point cloud.

关 键 词：离群值探测准则 主成分分析 点云平面拟合 效果分析 

分 类 号：P258[天文地球—测绘科学与技术] P228.41