Title: 3D post-disaster damage assessment with drone data
Rapid and comprehensive information is essential following disaster events, and remote sensing has long been a principal tool in the acquisition of this information. Despite a range of advantages, satellites can only provide partial insight into the structural damage on a disaster site, which has been leading to a growing deployment of drone platforms. While limited in operational range (distance and flight duration) and payload, drones are an interesting platform to aid damage assessment. In this seminar I will explain how we use the optical image data taken by cameras carried by drones, and how a richer analysis is possible when using also 3D geometrical and texture information extracted from those images. Among others we use machine learning approaches to develop semi-automatic analysis methods. I will also talk about the significant challenges that remain, ranging from restrictions related to regulation to the damage mapping process fundamentally being a subjective one.