Coarse-to-fine Extraction of Power Pylons and Power Lines in LiDAR Point Clouds

With the increasing popularity of LiDAR point clouds in transmission line maintenance, automatic extraction of power pylons and power lines becomes a key function for perceiving and monitoring the status of the transmission line project. However, the point clouds obtained by the 3D laser scanner are noisy. In addition to the elements that need to be extracted, they also contain ground plane data, vegetation data, and other outliers. In this paper, we propose a coarse-to-fine method for power pylon and power line extraction in LiDAR point clouds. In particular, the method first coarsely separates transmission pylons and lines, and then accurately fits the power lines using line or curve models. The effectiveness of the model proposed in this paper is validated through a case study in a point cloud captured in a power transmission line project, in which the results show that the proposed method is accurate in the segmentation of power pylons and power lines.