Dottorato in INGEGNERIA AMBIENTALE E IDRAULICA

Titolo della tesi: Advancements in Photogrammetric and LiDAR Techniques: Innovations, Applications, and Implications for Precision Mapping and 3D Modeling

My doctoral project involved the study, deepening, and improvement of the most recent geomatic techniques, with particular reference to the geometric processing of optical images and new sensors based on laser technologies. Photogrammetry has actually been developed since the inception of photographic techniques and, according to some authors, even from the mid-19th century. However, digital techniques have radically changed the approach methodologies, both at the level of software tools and software for processing and handling them. It is noteworthy that until the 20th century, photogrammetric techniques had achieved a certain stabilization, and acquisition and processing tools remained valid for decades. Currently, according to some of the most recognized authors in the field, certain photogrammetric survey instruments become obsolete in about two years. Regarding laser scanning survey techniques, recent years have been characterized by a succession of new sensors with, in some cases, surprisingly high performance, such as new miniaturized laser scanners that can be airborne by drones or installed in SLAM devices. In this climate of technological ferment, this thesis aimed to understand the potential and possible optimization methods of sensors that became available and evolved during the course of the thesis. For a broader overview, the focus was on photogrammetric sensors installed on drones, those within semi-professional cameras, as well as those mounted on simple smartphones and, finally, the more "traditional" panchromatic and hyperspectral sensors installed on satellite platforms. The processing of optical images is an issue that began and was developed during my undergraduate thesis (the results of which are reported in the appendix for completeness); the foundations of this initial research were deepened in the first chapter, where the experimental hyperspectral satellite PRISMA (then recently released) was evaluated and tested, revealing itself to be one of the complex satellite sensors to be geometrically processed. However, "classic" photogrammetric processing must now be integrated with more modern techniques borrowed from computer vision, primarily Structure From Motion (SFM). For this reason, in the second chapter, we will see how SFM techniques, applied to acquisitions with semi-professional cameras, drones, and smartphones, enabled a complete survey (previously impossible) of an area characterized by the presence of cavities of historical and environmental importance, also allowing the development of new theories on the tectonic evolution of the area by an international and multidisciplinary research team. Still by another international research team, the third chapter will illustrate the study of the combined potential of photogrammetric survey and high-precision GNSS positioning enabled by some smartphones, particularly in very remote areas where sensor miniaturization is of primary importance, making otherwise impossible surveys feasible. The fourth chapter will compare various recent survey techniques in a complex site of cultural and environmental significance, highlighting both the potential and the unexpected limitations of some of the latest generation sensors. The thesis concludes with the bibliography consulted for its development and the aforementioned appendix containing a study related to my undergraduate thesis on the development of satellite image orientation procedures, which formed the basis for some of the in-depth studies during my doctorate. Also in the appendix are very preliminary results of an ongoing research on the use of Cosmoskymed images on the island of Ischia and a comparison with permanent GNSS stations for the study of geodynamic movements.