机载激光点云中电力线的自动提取与重建  

Automatic power line extraction and reconstruction from airborne LiDAR point cloud

作　　者：程宇航 白征东[1] 辛浩浩 张强 郭锦萍 CHENG Yuhang;BAI Zhengdong;XIN Haohao;ZHANG Qiang;GUO Jinping(Department of Civil Engineering, Tsinghua University, Beijing 100084, China)

机构地区：[1]清华大学土木工程系,北京100084

出　　处：《测绘工程》2021年第5期58-63,共6页Engineering of Surveying and Mapping

基　　金：国家重点研发计划资助项目(2017YFB0504202)。

摘　　要：针对传统电力线点云自动提取方法在抗差性、准确性上的不足,文中从输电线路走廊中不同地物点云的空间特征差异出发,提出一种新的基于激光点云的电力线自动提取与重建方法。此算法利用栅格化处理和拟合残差判别提取电力线候选点云,通过双重K-means算法实现电力线重建。利用4组实际输电线路点云进行可行性试验。试验结果表明,方法对电力线候选点集的提取正确率均在97%以上,双重K-means算法能够准确地对电力线进行提取与重建,具有较高的提取准确率与较好的算法效率,可为实际工程的应用提供参考。Aiming at the lack of robustness and accuracy of traditional automatic power line extraction methods,a new automatic power line extraction and reconstruction method is put forward based on the spatial characteristics difference of the point cloud data of different features in transmission corridor.The candidate points are extracted by rasterization and fitting residual discrimination,and the power line are reconstructed by dual K-means algorithm.In this paper,the feasibility and accuracy is evaluated through experiments.The result shows that the algorithm proposed in this paper can achieve accurate automatic extraction and reconstruction of power line with a high extraction accuracy above 97%,which can provide references for practical engineering applications.

关 键 词：机载激光雷达 点云 输电线路 电力线提取 双重K-means算法 

分 类 号：P225[天文地球—大地测量学与测量工程]