利用自适应邻域特征估算与优选提取电力线  

Power Line Extraction Based on Self-adaptive Neighborhood Feature Estimation and Optimization

作　　者：袁超 李元达 陈昌耀 刘建明 YUAN Chao;LI Yuanda;CHEN Changyao;LIU Jianming(Jiangsu Province Institute of Hydrogeological and Engineering Geological Investigation,Huai’an 223400,China;Hangzhou Tongji Surveying&Mapping Co.,Ltd.,Hangzhou 310000,China;Zhejiang Huadong Mapping and Engineering Safety Technology Co.,Ltd.,Hangzhou 310014,China;Henan Institute of Geophysical Space Information Co.,Ltd.,Zhengzhou 450009,China)

机构地区：[1]江苏省水文地质工程地质勘察院,江苏淮安223400 [2]杭州同济测绘有限公司,浙江杭州310000 [3]浙江华东测绘与工程安全技术有限公司,浙江杭州310014 [4]河南省地球物理空间信息研究院有限公司,河南郑州450009

出　　处：《地理空间信息》2025年第3期58-60,65,共4页Geospatial Information

基　　金：基于新一代星载LiDAR数据的中国30 m分辨率森林高度反演研究(42071405)。

摘　　要：针对复杂地形下电力快速巡检对电力线信息提取的需求,基于机载LiDAR数据,提出了一种自适应邻域特征估算与优选的电力线提取算法。首先根据电力线高程特征分离非电力线点;再基于自适应邻域构建多维点云特征;然后联合支持向量机—递归特征消除(SVM-RFE)算法实现特征优选;最后对优选特征进行组合,并采用SVM、随机森林方法进行点云分类,提取电力线。结果表明,该方法提取的电力线信息精度较高,具有较好的适用性。In response to the demand of power line information extraction in rapid power inspection under complex terrain,based on airborne LiDAR data,we proposed a power line extraction algorithm based on self-adaptive neighborhood feature estimation and optimization.Firstly,based on the elevation characteristics of power lines,we separated non power line points.Then,we constructed multi-dimensional point cloud features based on self-adaptive neighborhood,and used the support vector machine-recursive feature elimination(SVM-RFE)algorithm to achieve optimal feature selection.Finally,we combined the optimal features,and used SVM and random forest method to classify point cloud,achieving the extraction of power lines.The result indicates that this method has high accuracy in power line information extraction and high applicability.

关 键 词：机载LIDAR 电力线 自适应邻域 特征优选 

分 类 号：P231[天文地球—摄影测量与遥感]