PROXIMAL REMOTE SENSING TECHINQUE FOR 3D MAPPING USING UAS AERIAL SYSTEM

The characterization of degraded areas by 3D modeling is mainly conditioned by accurate survey and mapping. In this paper we discuss a method based on proximal sensing photogrammetry technique, to obtain detailed photogrammetric survey by means of an Unmanned Aerial Systems (UAS). This technique allows to collect high definition digital images and generate very dense point clouds for DSM elaboration. In the field of Photogrammetry and Remote Sensing (PaRS; Colomina & Molina, 2014) the acquisition based on UAS platforms allows to perform low-cost survey with reliability and safety, obtaining a better information on morphologically complex landforms and better accuracy of 3D models. The UAS system used for this purpose is a multirotor drone equipped with two high resolution digital camera and 28mm fixed focal lens. The proximal sensing survey is performed at different flight heights (from 5 to 50m), according to the needed ground resolution (GSD), in order to generate DSM by the combination of different photogrammetric shots at consistent resolutions. This reduces errors of digital models due to interpolation, especially in presence of strong steepness, giving a more detailed DSM. After the definition of these parameters the flight is planned by waypoints, according to the photogrammetry rules: longitudinal overlap >70% and cross-overlap >=50%. The flight is monitored with a Ground Control Stations (GCS) using the MkTool software (HiSystems GmbH, 2014), to remotely control the telemetry data in real time. The flight planning is a very important phase, especially when the surfaces that have to be reproduced show different characteristics, strong steepness and different exposition to the incident light. To maximize the acquisition efficiency of the photogrammetric shots we use an experimental gimbal hosting two digital cameras, synchronized with a triggered time-lapse and respectively oriented to a vertical and oblique angle. In this way we can drastically reduce the acquisition time, the costs of the survey and enhance the accuracy, especially in presence of strong surface asperity and vegetation coverage. The resulting digital models could highlight with more details the degradation processes taking place that otherwise could not be properly identified and quantified. This method is particularly designed to characterize abandoned mining areas, allowing to investigate erosion phenomena also into a GIS environment, monitoring the evolution of degradation.