Development and Performance evaluation of 3D powerline reconstruction method using airborne LiDAR data
The North American electric power distribution network comprises a vast critical infrastructure of interconnected grids and power lines. Effective management of this system requires timely, accurate power line mapping and monitoring. Scene analysis for powerline change monitoring requires precise detection of all key corridor objects (i.e., powerlines, towers, insulators, splices, switches and other components as well as the terrain, buildings, trees, etc.). The complexity of corridor scene content and large data size currently exceeds the ability of current state-of-the-art geospatial data processing systems to accurately detect corridor objects. In addition, the primary methodology, which is manual-centric, is costly, slow, tedious and expensive. This project primarily focuses on the development of automatically reconstructing 3D powerlines from airborne LiDAR data and evaluates the developed techniques on a large-scale powerline datasets collected by the GeoDigital International (GDI) Inc. During the project period, the GDI will exercise the developed technology's practicality for industrial purpose. The internship student will be participated in the project for four months; for two months, the student will work at the GDI's Hamilton office for onsite training the GDI's commercially driven data processing chain, testing the algorithm, and for the remaining two months, the student will work at the York University to improve the algorithms and conduct a large-scale throughput test in collaboration with GDI.