Mass movements detection from UAV images analysus

Abstract

Biophysical conditions, the lack of governance and the weak applications of planning and territorial planning policies are aggravating factors of affections caused by natural disasters in Latin America. In Ecuador, landslides have caused not only material but also human losses, which shows the lack of immediate actions to avoid human settlements in risk areas or more controlled evacuations protocols, in case of early detection of risk areas due to this type of events. This article is part of an applied research project for the City of Cuenca allowing to take advantage of the information provided by citizens with a PPGIS. Furthermore, remote sensor images are analyzed, in order to identify in a semi-automated way, risk areas due to mass movements problems. Two approaches for semi-automated detecting areas where mass movements happen have been explored, i) Temporal images obtained by a LiDAR; and ii) Temporary images obtained by flights of an unmanned aerial vehicle, UAV. This paper present the procedure followed by comparison of image outcomes. The resulting analysis clearly presents the applicability of the two techniques thus allowing the determination of maps and early detection of landslides.