An unofficial blog that watches Google's attempts to move your operating system online since 2005. Not affiliated with Google.

Send your tips to

April 10, 2008

Viewfinder - Integrate Photos in a 3D World Model

Viewfinder is an interesting technology that wants to bring photos to a software like Google Earth and display them as part of the satellite imagery.

"Geotagged photos, geographically indexed on a world map, either manually or via GPS, are an increasingly popular phenomenon. However, current implementations treat maps, and particularly 3D models, in fundamentally different modalities than photographs. The result is that photos tend to hover like playing cards, seemingly suspended over the world, remaining 2D objects in a 3D environment, and negating the transformative experience that we think should occur when combining images and a 3D world. (...) It's possible to place a photo in a 3D model in such a way that it appears seamlessly aligned with the model."

Google has already bought Panoramio, a Spanish photo sharing site that selected around 3 million geotagged photos to be added in a Google Earth layer. In October last year, Flickr had more than 42 million geotagged photos. All these photos could be used to compose a more accurate representation of the world. Combining this with projects such as Microsoft's PhotoSynth should result in new exciting ways to explore the world.

{ via BoingBoing }

Update: More about Viewfinder in this Google techtalk.


  1. To accurately represent every part of the world,we need to have virtually every part of the world picture.Although its awsome what they have done as of now.

  2. PhotoSynth as a viewer is a neat thing to play with, but I expect Microsoft to be beaten to the punch by just about everybody, similar to how their "Surface" system generated a lot of press and then faded away once the iPhone and iPod Touch came out.

    Having seen many cool Microsoft demos over the years but very little cool Microsoft software, I'm confident that they do have some neat technology in there but that they will f(l)ail miserably trying to assemble a decent piece of software that takes advantage of it.

    It's one thing to come up with something cool in a lab, quite another to make a piece of software that can do real work in real situations.

    I look forward to Google pushing this technology forward in a way that people can actually use.