I stumbled upon this video of a new (still in development?) product from Microsoft, called PhotoSynth. It takes a set of photos, and creates a 3d enviromnent from them, by analzying similarities from the photos. Think of it as a super-powered VR image-stitcher.
The end result is an application that can take your photos of the Taj Mahal that you took on your vacation, and assemble them into a 3D photo album, that you can explore. It becomes a bit more interesting when you consider the possibilities of linking to network photos. Your virtual Taj Mahal could be constructed from not only your photos, but everyone else's who have tagged the image with 'Taj Mahal'.
The possibility of integrating this with Flickr is almost too much to think about. A giant server running PhotoSynth could probably 3D-map a fair portion of the planet, just from Flickr photos. This could be good and bad. While 3D renderings of tourist attractions may seem kind of harmless, what about 3D renderings of banks? Police stations? Airports? Shopping malls? One person taking 18,704 photographs of LAX might catch someone's attention, but gathering thousands of photos taken in the airport by tourists would take about 30 seconds using Flickr.
I also wonder how long it will take people to start to 'punk' these rendered environments, by photo-shopping in advertisements, naked pictures, or political slogans. You read it here first.
In any case, PhotoSynth is impressive technology. Check out the video here.
