Authors: Fang Qiu*, University of Texas - Dallas, Haitao Lyu, University of Texas at Dallas
Topics: Remote Sensing, Geographic Information Science and Systems
Keywords: LiDAR, power line extraction, electric pole extraction, bare earth extraction
Session Type: Paper
Presentation File: No File Uploaded
Electricity is the most important energy driving modern industry and life. To distribute electricity from power plants to consumers, energy companies install thousands of miles of electricity delivering facilities (including power lines, poles and other associated equipment). To ensure the normal operation of the facilities, workers have regularly performed inspection by foot, driving between the spans, or from a helicopter/drones flown alongside the powerlines to examine their condition facilities and to identify potential danger caused by the encroachment from surrounding vegetation into the facilities. These methods are either prohibitively labor intensive or expensive. Airborne laser scanning system in the form of light detection and ranging (LiDAR) are being more and more utilized to extract electricity delivering facilities and detect the possible vegetation encroachment. To achieve these goals, a python-based toolset was designed to process the data collected by linear or Geiger-mode LiDAR. The toolset first extracts and eliminate the bare earth based on a progressive TIN method, and the above-ground points are kept for further process. Then an electricity pole detection algorithm was developed, which involves the potential pole tops detection based on a vertical cylindric search, followed by false top removal based on both vertical and horizonal metrics. Once the pole tops are identified, powerlines are extracted based on pairwise linkage of the poles and false powerline are removed based on a horizontal cylindric search. A KD-tree based index is then applied to the powerlines detected, which are used to identify vegetation encroachment based on a range search algorithm.