Tree-Based Approach for Vegetation Monitoring and Risk Assessment Along Powerline Using High Resolution Satellite Image

Vegetation management with remote sensing is receiving growing attention as a low-cost and highly scalable solution to replace the traditional labor-intensive approach. Our research focused on the use of high-resolution satellite imagery for risk assessment along the right of way due to the wide area covered by the infrastructure and ability to provide occasional updates at a lower cost. We proposed a top-down framework for a tree-based risk analysis approach and validated the predicted powerline risk levels against results generated using LiDAR input. We summarized and discussed the results for each component of the framework, from tree height estimation, treetop detection, crown delineation to risk analysis. The method achieved a high F1 score of 0.87. The result demonstrates the potential for a cost-effective risk assessment solution to facilitate maintenance and planning along powerlines.