Classification of high-voltage power line structures in low density ALS data acquired over broad non-urban areas

Jean Romain Roussel, Alexis Achim, David Auty

Research output: Contribution to journalArticlepeer-review

3 Scopus citations

Abstract

Airborne laser scanning (ALS) has gained importance over recent decades for multiple uses related to the cartography of landscapes. Processing ALS data over large areas for forest resource estimation and ecological assessments requires efficient algorithms to filter out some points from the raw data and remove human-made structures that would otherwise be mistaken for natural objects. In this paper, we describe an algorithm developed for the segmentation and cleaning of electrical network facilities in low density (2.5 to 13 points/m2) ALS point clouds. The algorithm was designed to identify transmission towers, conductor wires and earth wires from high-voltage power lines in natural landscapes. The method is based on two priors i.e. (1) the availability of a map of the high-voltage power lines across the area of interest and (2) knowledge of the type of transmission towers that hold the conductors along a given power line. It was tested on a network totalling 200 km of wires supported by 415 transmission towers with diverse topographies and topologies with an accuracy of 98.6%. This work will help further the automated detection capacity of power line structures, which had previously been limited to high density point clouds in small, urbanised areas. The method is open-source and available online.

Original languageEnglish (US)
Pages (from-to)1-22
Number of pages22
JournalPeerJ Computer Science
Volume7
DOIs
StatePublished - 2021

Keywords

  • Classification
  • Forest
  • LiDAR
  • Powerlines
  • Segmentation
  • Transmission towers

ASJC Scopus subject areas

  • General Computer Science

Fingerprint

Dive into the research topics of 'Classification of high-voltage power line structures in low density ALS data acquired over broad non-urban areas'. Together they form a unique fingerprint.

Cite this