Social tourism is a system for automatically determining geographic location from one photo and then augmenting the photo with additional information. As tourists explore their environment they can add graffiti in the form of text and images to objects in the scene which appear on the same objects in other photos as they are taken. Photos are further augmented with navigation directions in the form of arrows pointing tourists to their desired destination. The system is automatically generated from a collection of geotagged photos using robust local feature correspondence.
To download the structure from motion software used in this system, please see:
The Bundler website.
David Lowe's SIFT website.
For more details on the structure from motion algorithm used in this system, please see:
The Photo Tourism website by the Community Photo Collections team at the University of Washington.