Reconnaissance of Two Liquefaction Sites Using Small Unmanned Aerial Vehicles and Structure from Motion Computer Vision Following the April 1, 2014 Chile Earthquake

Abstract

Small unmanned aerial vehicles (sUAVs) are used to reconnoiter, image, and model the effects of soil liquefaction at two separate sites following the M8.2 and M7.6 earthquakes near Iquique, Chile on April 1 and 2, 2014, respectively. Geotechnical investigators performed conventional observations/measurements at the two sites using traditional reconnaissance methods, as well as virtual observations/measurements using sUAV-based remote sensing and structure from motion (SfM) computer vision. Two types of sUAV platform/sensor combinations areevaluated for this study. Comparisons between the traditional and sUAV-based reconnaissance methods are performed and observations are discussed. The sUAVbasedpoint cloud models of the sites are between 770 to 2600 points per square meter depending on the camera that is used, with average model dimensional erro ranging from ±1.0% to ±3.5%. Lateral spread displacement measurements from the UAV-based three-dimensional models are generally within 9 cm of the lateral spread displacements measured in the field, and measured vertical deformations from the models generally agree well with the deformations reported by previous investigators. This study demonstrates that sUAVs can be useful instruments for post-disaster geotechnical surficial site reconnaissance.