GNGTS 2017 - 36° Convegno Nazionale

636 GNGTS 2017 S essione 3.2 and mapping. Moreover the UAV was also equipped with a thermal imaging camera Optris PI 640, with an optical resolution of 640x480 pixels, the PI 640 with temperature ranges from –20 °C to 1500 °C and accuracy: ±2 °C. Data acquisition from the entire survey was accomplished by a combination of two missions in a uniform crossed grid pattern. The coverage of the entire area has been achieved by acquisition of 5400 frames for the visible spectrum and 120 for thermal. The survey has been performed at an altitude of 30 m and a speed of 3 m/s. Photo overlap is user-set in the 75-85% range, allowing for high-quality photogrammetric image matching. A set of ground control points have been used for geo-referencing the Digital Model. The aerial photographs were processed into georeferenced orthoimages, a high-resolution Digital Surface Model (DSM) and Digital Elevation Model (DEM), using the photogrammetric 3D reconstruction technology software by AgiSoft PhotoScan (Agisoft LLC, 2015). The DEM generated has a ground resolution up to 4.0 cm. Advances in computer vision and image analysis are, however, generating innovative developments in photogrammetry through the technique of Structure-from-Motion (SfM), which offers an automated method for the production of high- resolution DSMs with standard cameras (Fonstad et al. , 2013; Micheletti et al. , 2015). Terrestrial survey. The magnetic survey was carried out by means of the Overhauser magnetometer-gradiometer GEM-19 from GEM Systems which has an accuracy of 0.2 nT. The device was set as gradiometer with two parallel sensors spaced by 56 cm and carried by the operator on a specific backpack. Measurements were performed walking along parallel lines spaced by 0.5 m and materialized on the field with measuring tapes; t�� ������ ���� ������� ��� he survey area extends for about 2500 m 2 . ��� ������������ �� �������� ���� � ��� ��������� ��� �������� ��� ��� ���� The magnetometer is equipped with a GPS providing the position and the time for each magnetic measures (5 Hz sampling frequency), for more than 37,600 readings. Data were interpolated on a regular grid to draw a map of the total magnetic field and a map of the magnetic gradient. The values of the total magnetic field range from about 45000 to 45080 nT (Fig. 1a). There is evidence of magnetic features with magnitude in the order of ~5 - 20 nT. Such features are preferable ESE-WNW oriented (and, in a minor way, perpendicularly) and have dimension ranging for 2 to 6-8 m. For their characteristics they are ascribable to buried structures (e.g. roads or buildings). Differently, the anomalous low magnetic values in the southernmost part of the study area, are clearly influenced by the occurrence of a metal fencing that delimits the archaeological site. Fig. 1 - a) Magnetometric map. b) GPR depth-slice ranging from 0.5 m to 1 m depth.