A Piecewise Catenary Curve Model Growing for 3D Power Line Reconstruction

Abstract

The three-dimensional (3D) modeling of power lines is one of the most critical processing steps for identifying vegetation encroachment on high-voltage power lines, which often results in electrical shock hazards, fires, and outages, as well as verifying the current capacity of existing power lines and thus supplying the increasing demand for electricity. Today’s state-of-the art techniques are used to reconstruct power line models, however, still require human intervention to a certain extent, which hinders the rapid response needed for effectively managing power line risks. We propose an alternative solution to automate the procedures to reconstruct a 3D power line model from Airborne Laser Scanning (ALS) data. The proposed method starts by extracting power line candidate points, which are converted into a catenary curve model representing a power line. This initial model is allowed to progressively grow to produce a complete power line model by producing hypothetical growing models and selecting optimal one. A stochastic constrained non-linear adjustment method is developed for estimating catenary curve model parameters. An evaluation of the proposed approach over a complex power line scene demonstrates that the proposed method is promising for automatically reconstructing 3D power line model from ALS data, and thus provides benefits for efficient and reliable power line risk management.