Photo Tourism2

The Colosseum as seen in the digital reconstruction. Each triangle is where a person was standing when he or she took a photo. The building’s shape is determined by analyzing photos taken from all these different perspectives.

U. WASHINGTON (US)—Using a new computer algorithm, researchers were able to take 150,000 tourist photos tagged “Roma” or “Rome” downloaded from the photo sharing Web site Flickr and combine them into a single 3-D digital model in about 21 hours.

“How to match these massive collections of images to each other was a challenge,” says Sameer Agarwal, acting assistant professor of computer science and engineering at the University of Washington.

Until now, he explains, “even if we had all the hardware we could get our hands on and then some, a reconstruction using this many photos would take forever.”

Photo Tourism, an earlier version of the photo-stitching technology developed at the University of Washington, was licensed in 2006 to Microsoft, and is now available as a free tool called Photosynth.

“With Photosynth and Photo Tourism, we basically reconstruct individual landmarks. Here we’re trying to reconstruct entire cities,” explains coauthor Noah Snavely, who developed Photo Tourism as part of his doctoral work at Washington and is now an assistant professor at Cornell University.

Transitioning from landmarks to cities and going from hundreds of photos to hundreds of thousands of photos is not trivial, Agarwal explains.

As the number of photos increases the number of matches explodes, increasing with the square of the number of photos. Without the new technology, a set of 250,000 images would take at least a year for 500 computers to process; a million photos would take more than a decade.

Working more than 100 times faster that the earlier version, the newly developed code  first establishes likely matches and then concentrates on those parts. The code also uses parallel processing techniques, allowing it to run simultaneously on many computers, or even on remote servers connected through the Internet.

The new faster code makes it possible to tackle more ambitious projects, says computer science professor and coauthor Steve Seitz. “On a timeline of one day, you can methodically start going through all the cities and start building models of them.”

The technology has reach far beyond tourism, the researchers believe. Not only could it be used to build cities for video games automatically, rather than by hand; it could also be used in architecture for digital preservation of cities, or integrated with online maps, Seitz says.

In addition to Rome, the team also recreated the Croatian coastal city of Dubrovnik, processing 60,000 images in less than 23 hours using a cluster of 350 computers, and Venice, Italy, processing 250,000 images in 65 hours using a cluster of 500 computers.

The research was supported by the National Science Foundation, the Office of Naval Research and its Spawar lab, Microsoft Research, and Google.

University of Washington news: