Efficient Pavement Monitoring for South Korea using Unmanned Aerial Vehicles

Abstract

The present study is aimed at using unmanned aerial vehicles (UAVs) for pavement monitoring. It was taken up as an initial pilot study to develop a network level asset management framework for South Korean conditions. A varying pavement stretch containing a bridge, culvert, streams, and merging traffic junctions was selected for the study. Using a quadcopter UAV, 139 overlapping images were acquired and processed using a structure from motion (SfM) program to develop a digital twin of the pavement. The generated orthomosaic and 3D digital twin were used to identify pavement damage and other infrastructural assets. UAV-based image acquisition was found to provide sufficient resolution, exposure, and key point matches, enabling an accurate 3D model generation with detailed feature extraction. Geometric measurements of various features depicted the potential and efficiency of UAV surveys. The research work is expected to aid in effective contactless pavement monitoring and asset management during regular surveys as well as disasters.