以机载LiDAR点云为数据源的电力线自动提取算法研究

Research on power line automatic extraction algorithm using airborne LiDAR point cloud as data source

目前,从激光雷达点云中提取电力线已经成为电力巡检的重要方法。为了得到更准确的电力线提取算法,本文提出一种基于激光雷达(LiDAR)点云的电力线自动提取算法。首先,利用布料模拟滤波算法分离地面点和非地面点;其次,通过格网划分、高差分割点云数据,对电力线进行粗提取;最后,利用Hough直线检测和改进的DBSCAN聚类算法实现电力线点云的精细提取。试验结果表明,该方法能从机载激光点云数据中提取出完整的电力线,并具有较好的鲁棒性,对电力巡检具有一定的实际意义。

At present,extracting power line from LiDAR point cloud has become an important method of power inspection.In order to get a more accurate power line extraction algorithm,this paper presents an automatic extraction algorithm of power line based on LiDAR point cloud.Firstly,the ground points and non ground points are separated by CSF algorithm.Secondly,the power line are extracted by grid segmentation and height difference segmentation of point cloud data.Finally,Hough line detection and improved DBSCAN clustering algorithm are used to achieve fine extraction of power line.The experimental results show that this method can extract complete power line from airborne laser point cloud data,and has good robustness,which has certain practical significance for power line inspection.