复杂地理背景下电力线自动提取方法  

Automatic extraction method for power lines in complex geographic background

作　　者：袁晨鑫 陈良良 YUAN Chenxin;CHEN Liangliang(China Nuclear Industry Group Company Limited,Xi'an,Shaanxi 710100,China)

机构地区：[1]中陕核工业集团测绘院有限公司,陕西西安710100

出　　处：《北京测绘》2025年第1期116-120,共5页Beijing Surveying and Mapping

基　　金：陕西省科协青年人才托举计划(20230107)。

摘　　要：为实现电力线走廊更加快捷、有效的巡检,本文提出了一种从机载激光雷达(LiDAR)扫描点云数据中提取电力线的方法。首先,利用不规则三角网(TIN)滤波和角度滤波对原始点云数据进行电力线粗提取;其次,采用基于高斯核的霍夫变换方法来检测和提取电力线;最后,使用最小二乘拟合法对电力线进行分段拟合。实验结果表明,使用该方法提取电力线的准确率达到97.5%,误判率低于1.5%。电力线直线拟合误差在5 cm以下,满足电力线巡线的要求。该方法能够更快速、准确地获取电力线数据,提高电力线检测和监测的效率和可靠性。In order to realize faster and more effective inspection of power line corridors,a method for extracting power lines from airborne light detection and ranging(LiDAR)point cloud data was proposed.Firstly,power lines were roughly extracted from the original point cloud data by triangulated irregular network(TIN)filtering and angle filtering.Secondly,the Hough transform based on the Gaussian kernel was used to detect and extract the power lines.Finally,the least square fitting method was used to fit the power lines by segments.The experimental results show that the accuracy of power line extraction by this method is 97.5%,and the misjudgment rate is less than 1.5%.The linear fitting error of the power lines is less than 5 cm,which meets the requirements of power line inspection.This method can obtain power line data more quickly and accurately and improve the efficiency and reliability of power line detection and monitoring.

关 键 词：机载激光雷达 霍夫变换 电力线提取 最小二乘拟合法 

分 类 号：P258[天文地球—测绘科学与技术]