Date of this Version
Mohammadi ME, Wood RL. (2018). Damage Assessment of Built-up Areas via UAS-SfM Derived Point Cloud Data. 11th US National Conference on Earthquake Engineering, Los Angeles, CA, June 25-29, 2018, 5pp.
In the aftermath of extreme events (e.g., earthquakes, tsunami, tornados, etc.), rapid and reliable identification of the damage in a built-up area are crucial in to rescue, recovery, and reconstruction operations. While it is critical to conduct efficient emergency response management, lack of classified or tagged damaged regions due to communications and accessibility limitations can further delay recovery operations, rescue efforts, and resource management. Furthermore, critical and perishable damage scenes can also be lost during recovery and cleanup operations immediately following the event. In recent decades, advances in remote sensing technologies demonstrate a great potential to perform rapid reconnaissance and damage assessments. One such technology is an unmanned aerial systems (UAS). UAS platforms require minimal ground support and can be easily and efficiently deployed to collect detailed images from the damaged areas within short duration of time. These images not only depict the current conditions, but can also be used for three-dimensional point cloud reconstructions using a computer vision algorithm, Structure-from-Motion (SfM). Within this study, an algorithm is developed to objectively evaluate SfM point cloud data using local geometric features to detect and identify the locations of damaged structures. This method outlines a rapid, cost-effective, and safe solution for damage assessment of built up areas, which could be used in efficient decision making for emergency response management and estimating the intensity of an earthquake or severity of a tornadic event.