To realize refined inspection of transmission towers using UAVs

it is necessary to extract the coordinates of these towers from complex laser point cloud data.This paper first preprocesses the point cloud data by downsampling and elevation filtering.Subsequently

each tower is separated and extracted according to the spatial characteristics of the towers

wires

and ground point clouds.Finally

the independent tower is precisely extracted

leveraging their inherent vertical and symmetrical characteristics.Using the laser point cloud data from two 220 kV transmission lines in Zhangjiagang as a case study

the proposed algorithm can extract all transmission towers efficiently and completely.The modeling of a single tower is effective

with an average extraction time of about 2 minutes per tower

facilitating the subsequent trajectory planning for UAV-based detailed tower inspections.