This thesis deals with data processing in Geomatics. It ranges from data acquisition in Photogrammetry to data representation as well as in Cartography. The objective of this research was to use statistical techniques of data processing for the creation of digital surface models starting from photogrammetric images. The main function of photogrammetry is the transformation of data coming from the image space to the object space. An easy solution for three image orientation is proposed. The orientation procedure described has some relevant advantages for environmental and monitoring applications, and makes it a very powerful tool in addition to more traditional methodologies. Among many different applications, an interesting project for the survey of a hydraulic 3D model of a stream confluence in the mountain area has been performed. From a computing point of view, we propose a description of the photogrammetric data based on the XML format for geographic data (Geographic Markup Language). The aim is to optimize the archiving and management of geo-data.
As a conclusion, an original software product which allows to model terrains starting from three-image photogrammetry has been developed and tested.