Sviluppo e applicazioni di tecniche di automazione in fotogrammetria dei vicini

Automation plays a central role in photogrammetry: surveying operations rely on the use of increasingly advanced and performing computers and the gap between human resources costs and machine costs tends to increase exponentially. On the other hand, different techniques where three-dimensional reconstruction is more direct (e.g. laser scanner) still lacks efficient restitution methods. Automation, in this sense, is necessary to keep the photogrammetry discipline competitive and for leverage its integration with other technologies. The thesis addresses, illustrating the main results and the most important applications developed in the last three years, different techniques that allow the almost complete automation of the photogrammetric procedures in close-range, characterized by arbitrary block structures. Image block orientation (Structure and Motion), dense point cloud restitution based on image correlation algorithms, analysis and segmentation of structured and unstructured three-dimensional data are all activities that can be automated. A final section shows an application in which all the different techniques are used in sequence, creating a fully automated measurement and analysis system.