Fast Inspection in War-affected Areas Using Airborne Remote Sensing Point Cloud

¹ Faculty of Civil Engineering, Idlib University, Syria.
² Faculty of Health, Engineering and Sciences, University of Southern Queensland, Springfield Campus, Springfield, QLD 4300, Australia

^* Corresponding author: jwan_abdo_zoaa@idlib.edu.sy

Abstract

Rapidly assessing damage in war-affected areas is critical for emergency response and reconstruction planning. Traditional inspection methods often compromise personnel safety and lack comprehensive spatial coverage. This investigation employed drone-based methods with advanced imaging systems and LiDAR technology to reconstruct precise three-dimensional models for damage assessment in disaster zones. Aerial imagery was captured using a DJI Mavic 2 Pro drone over a 12-hectare area in Kafr Nabl, Syria—a region devastated by armed conflict. The dataset, comprising 122 nadir images acquired at an altitude of 150 meters, was used to reconstruct a coloured 3D point cloud. Advanced segmentation algorithms classified the point cloud into ground class and building class, enabling targeted structural damage analysis. Three different types of damage were identified: 1- Complete damage (the building was completely demolished), 2- Damage to the roof only, with the load-bearing parts remaining intact, and 3- Partial damage to the roof and the load-bearing structure. The results demonstrate the system's capability to map damage distribution with accuracy. The framework provides decision-makers with criteria for prioritizing reconstruction efforts, including safety risks, resource allocation, and cost-benefit analysis of repair versus demolition. This approach significantly reduces human exposure to hazardous environments while delivering rapid results..