AUTHOR=Stone Harriette , Putrino Valentina , D'Ayala Dina TITLE=Earthquake Damage Data Collection Using Omnidirectional Imagery JOURNAL=Frontiers in Built Environment VOLUME=4 YEAR=2018 URL=https://www.frontiersin.org/journals/built-environment/articles/10.3389/fbuil.2018.00051 DOI=10.3389/fbuil.2018.00051 ISSN=2297-3362 ABSTRACT=

The unique perspectives and viewpoints offered by omnidirectional camera technology has the potential to help improve the outcomes of technical post-earthquake reconnaissance missions. Omnidirectional imagery can be used to virtually “walk through” damaged streets post hoc with a 360°, immersive view. A common reconnaissance mission aim is to accurately collect damage data; however, there are time challenges for surveyors in the field. The manuscript explores the potential for using omnidirectional imagery to improve damage surveying, firstly by comparing results from damage surveys completed in the field with results obtained using omnidirectional images collected during a mission and surveyed by an experienced engineer virtually and secondly by comparing damage assessment obtained through omnidirectional imagery collected on the ground with the EU Copernicus damage assessment maps. The omnidirectional imagery data was collected during two separate Earthquake Engineering Field Investigation Team post-earthquake reconnaissance missions, namely the area affected by the 2016, 7.8 Muisne Earthquake in Ecuador and the area struck by the 2016, 6.2 Amatrice earthquake in Italy. Notwithstanding the diverse geographic scale, terrain and urban context of the two reconnaissance missions, the results consistently show significant capabilities for this technology in the identification of construction typologies, number of stories, aggregated “low” and “high” damage grades, and failure modes. The work highlights potential issues with correct identification of disaggregated lower damage grades (e.g., European Macroseismic Scale (EMS-98) damage grades 0–3). Challenges identified in the virtual survey process included poor image quality, insufficient photo sphere captures, and obstructions such as trees, walls or vehicles. The omnidirectional imagery represents a substantial improvement in damage assessment accuracy in respect to satellite imagery, especially for lower damage grades, while it is an essential tool for comprehensive surveys in reduced access zones with high levels of damage.