Automatización de la vectorización de conductores eléctricos a partir de datos LiDAR en QGIS

Abstract

This research develops an automated methodology for the vectorization of electrical conductors from classified LiDAR data through the development of the LineDetect plugin in QGIS. The study addresses the high dependence on manual processes for generating electrical line geometries, which increases processing times and operational costs in electrical infrastructure projects. The implemented methodology integrates LiDAR point cloud processing, geometric analysis, spatial tracking of conductors, and automated generation of three-dimensional LineStringZ entities using PyQGIS and specialized Python libraries. Additionally, a PostgreSQL/PostGIS spatial database was incorporated for the automated storage and management of the information generated during processing. During the development of the project, different commercial and open-source tools for electrical conductor vectorization were evaluated, including Bentley MicroStation, ArcGIS Pro, and QGIS. The obtained results demonstrated that the developed plugin reconstructs electrical conductors with adequate spatial continuity and geometric adjustment relative to the original LiDAR point cloud, achieving RMSE values close to 0.018 m and R² coefficients near 0.99998. Finally, the research demonstrates the feasibility of using open-source GIS tools to automate vectorization processes for electrical infrastructure, contributing to reducing manual intervention and providing a flexible and replicable alternative for LiDAR data processing in geospatial environments.