Sviluppo di un aeromobile prototipale attrezzato per il remote sensing multi-parametrico: sperimentazione del sotto-sistema fotogrammetrico in differenti morfologie e applicazioni.

Airborne remote sensing is so far one of the most important application in the general field of remote sensing. Since from the first aircrafts, different film cameras were used to acquire images lately treated by a photogrammetric approach. The airborne remote sensing potentials are generally opposed to the high cost normally associated to this activity and to the need of high specialized operators for fieldwork and laboratory processing. In the last years, several factors have allowed an increase in the number of users of airborne remote sensing data; among these, it is important to mention the development of processing techniques based on Computer Vision (CV) and the Structure from Motion (SfM) algorithms, the increase of simplicity and power of the remote sensing software, the exponential growth of computing power, and the relative low price of modern workstations. Main goals of this PhD research project are: 1) the development and optimization of a manned prototypal aircraft agile, economical and functional for multi-parametric scientific survey; 2) the acquisition, processing, presentation and accuracy validation of different datasets acquired thanks to the developed aircraft. The device is, technically, an autogyro called “RadGyro” for its initial use in natural radioactivity survey. The autogyro is extremely agile, quite inexpensive, and with a high load capacity (actual maximum payload reaches about 100 kg). In the first phase, during the construction and optimization of the RadGyro, the scientific instrumentation, used in the second phase of the project, was implemented. The second phase was carried out by testing potentials and limits of the system through the experience derived from the most significant case studies. The focus was on the photogrammetric sub-system, but lateral activity has concerned the use of a thermo-camera and of a hyperspectral visible/near infrared pushbroom scanner. The photogrammetric sub-system was assessed trough different case studies in order to check the quality of the processed products; by the creation of Digital Elevation Models (DEM), Digital Terrain Models (DTM) and ortophotos, it was possible to update the existing maps also minimizing the operator interpretation of landforms thanks to a high resolute and spatially continuous acquisition. The RadGyro and the SfM techniques allowed reducing to few days fieldwork activities necessary for traditional topographic survey by GNSS (Global Navigation Satellite System) and Total Station. It is impossible, for the traditional topographic survey to reach a 3D point clouds density as high as the one reachable by the application of image matching techniques, except using terrestrial or airborne LIDAR (Light Detection and Ranging). However, to obtain high positional accuracy, the survey with geodetic instrument of well distributed GCPs (Ground Control Points) still remains indispensable. The class of the Inertial Measurement Unit (IMU) and GPS (Global Positioning System) receivers affects the quality of the acquired data: this accuracy can be considered sufficient in navigation, surveying and pre-processing phases, but when centimetric accuracy is requested (quite typical for an photogrammetric flight), at least a post-processing correction or a field topographic acquisition of GCPs are needed. Aerial photogrammetry by means of RadGyro and SfM gave worthy results also in multi-temporal monitoring, with the big plus that the survey of many sites of interest has been possible only by a single low-altitude flight. The actual configuration of the RadGyro can be used, with good results, in areas where surface ranges between some square kilometres to many dozens of square kilometres. This is especially true when the costs and acquisition times required of a UAV (Unmanned Aerial Vehicle) flight are too high or when the accessibility of the site is complicated for traditional terrestrial surveys. Beside the validation of the photogrammetric sub-system, also sensors operating in the thermal and visible-near infrared range are now operative in the RadGyro thanks to parallel activities that allowed to develop ad hoc software, to make test flights and to complete the implementation of a workflow relative to data processing. At the end of the PhD research project, all these sensors are integrated and operational on board.