[automatically translated] In the memory are reported the results of the verifications of applicability of an image-based technique for the monitoring of channeled erosion processes. The survey, in particular, has had the aim of testing the use of three-dimensional terrain models (DTM) obtained with the use of a large number of photographs of the same scene acquired from different points of view (Technique "Structure-From Motion "SFM and" Multi-View Stereo-MVS). The DTM-dimensional (3D) was generated using the open source software and Autodesk 123D Catch to create the Digital Elevation Model DEM (2.5D) has been employed CloudCompare software. Of the two channels in the ground used in the experimentation, a rectilinear triangular section and the other curvilinear trapezoid section, It was also carried out the survey of some sections by means of a profilometer. The actual volume of each channel has been measured by filling it, after it is sealed with a plastic film, with a known volume of water. The comparison between the three methods of relief (3D model, 2.5D and profilometer) model showed in all cases a slight underestimation of the total volume that for the case of the 3D model assumes the minimum value, equal to -0.3% for the straight channel and at the - 0.5% for the curvilinear. Hired as a "reference" the 3D model the morphological and hydraulic characteristics relating to relevant sections were compared. The analysis shows that the profiler provides values of the maximum depth, H, systematically higher than those in the 3D model, as the model 2. 5D no systematic differences in H than in 3D. The wetted perimeter is underestimated by both the profilometer is from 2.5D method. The volumes of the individual sections, in the case of the curvilinear channel, are systematically underestimated (-0.73% for the 3D model and -2.71% for the profilometer) as a result of underestimation of the area of the cross sections.
|Number of pages||9|
|Publication status||Published - 2016|