Photographs are among the most-widely exchanged media in our lives and we know they carry a high persuasive power ("seeing is believing"). As a matter of fact, two questions arises:

1. How can we trust such a powerful media?
2. How is it possible to infer what really happened from a set of different, heterogeneous, sometimes contradictory pictures?

The focus of this thesis work is the reconstruction of real life event from crowdsourced data (images, videos and twits harvested online). Starting from SfM solutions and software, the research will focus on localizing and synchronizing groups of heterogeneous pictures taken during the same event.

.

(Edited by Friso Simone - original submission Friday, 2 October 2020, 3:05 PM)