Docente
Mirko Reguzzoni (mail)
Area di ricerca
Intelligenza artificiale, robotica e computer vision
Keyword (max 3 separate da virgola)
Photogrammetry, Least-squares, Digital-images
Descrizione (max 500 caratteri)
The goal is to write a software that is able to read two or more digital images and, by using photogrammetric techniques, compute their orientation and, if possible, the corresponding 3D dense cloud of points. Double points between couples of images can be acquired as additional input from an external software or manually selected on the images through a visual interface. The core of the algorithm is a non-linear least-squares adjustment based on collinearity equations. The algorithm can be tested on drone images.