Remember how ridiculous Google Street View was the first time you heard about it? One company–an Internet company no less!–was sending cars on every road in the U.S. (and then the world) to build a seamless map of images, a first-person view of what it would look like to be almost anywhere.
Of course, there are places where cars can’t go, like into St. Mark’s Basilica or through the Trevi Fountain. To provide views of these landmarks–what the company calls Photo Tours–Google has turned to crowdsourcing. They’re actually digging through public, geotagged photos on Panoramio and Picasa, snagging shots from various user perspectives and constructing spinning, morphing 3-D tours based upon the images.
“We use state-of-the-art computer vision techniques to organize and position all the photos in 3-D. Then the system groups, or clusters, the photos according to the similarity of the images and selects the best canonical views to include in the photo tour based on the data available,” Google spokesperson Deanna Yick tells Co.Design. “For example, if a lot of people take photos in front of a famous cathedral, then our algorithm will choose to show the best of those photos, taking into considering factors like resolution, photo popularity, etc. The algorithm knows the highlights of a place because users take a lot of photos of those highlights.”
The result is the view of a place from the hive mind, a collective map created from our inner auteurs. We’re a predictable lot when it comes to tourist destinations–just search “Leaning Tower of Pisa” to discover countless photos of people making the same joke.
On one hand, we could judge that human tendency as unoriginal. But on the other, the sheer amount of photos that people have uploaded of a place (and even certain angles of that place) are an undeniable metric for their importance to us. By combining our most popular views, Google is just showing us what we’ve all chosen to look at, time and time again–the world curated by the millions of photographers around the globe.
And while it’s a bit of a side note, it is worth mentioning that Google is attributing the photographer behind each and every shot, and they’re not lifting images from private collections (but your publicly shared photos are totally fair game.)
“If a user has already chosen to share their photos publicly on Panoramio or posted them to a public Picasa album, those photos are automatically eligible for inclusion in a photo tour. On the bottom left corner of the photo tour, attribution will be available listing the site where the photo originated and the selected username of the person who contributed it. Clicking on either of those attribution links will take you to the corresponding photo within Panoramio or the public Picasa album,” Yick writes. “And of course, photos that are removed from public Picasa albums or Panoramio will also be removed from photo tours.”
As of now, the feature is clearly in its infancy. The tours are pre-rendered animations that can’t be explored at-will like Street View, and just a few spots are available, as Google is manually tweaking its new crowdsourcing algorithms as it goes. But let’s fast forward just a few years. As all of our pictures become geotagged by our smartphones, and as we only share more and more to the cloud, Google could combine the perspectives of, not just everyone at a single tourist destination, but everyone everywhere on the planet. They could assemble trillions of images from billions of people to create an interactive map of, not just our streets, but every inch of the entire globe.
Except underwater. They’ll still have to charter some submarines for that part.
Top image by StevanZZ/Shutterstock.