Authors: XIAO LI*, UTDallas, Fang Qiu, UTDallas
Topics: Remote Sensing, Energy, Spatial Analysis & Modeling
Keywords: Lidar, Power-line detection
Session Type: Paper
Start / End Time: 8:00 AM / 9:40 AM
Room: Muses, Sheraton, 8th Floor
Presentation File: No File Uploaded
The safety of electricity infrastructure becomes more and more significant for modern society. Power outage and interruptions could cause tremendous economic loss. However, due to the huge amount and complexity of modern electricity transmission network, traditional manual inspection, which is time-consuming and expensive, become insufficient and unacceptable. Therefore, an automatic and efficient monitoring method for transmission network is urgently required. Airborne Lidar technology, which collects dense 3D points measurements of objects on the ground, appears as a promising alternative for monitoring the power line. Currently, some researchers have proposed many algorithms for power line extraction and reconstruction using lidar data. But most of them use classification of entire Lidar point dataset as the first step, which is very time consuming due to the large amount of lidar points. We are going to introduce a robust efficient algorithm for powerline detection and localization, which use raw lidar points as input and does not require points classification. We use lidar dataset with different point density as test data and successfully achieve high accuracy.