Today at Web 2.0, Gary Flake from Live Labs released a Technology Preview of PhotoSynth, a next-generation 3D photoviewer. They had released a video of it this past summer (Radar post). Photosynth takes a group of photos and stitches them together. These photos are then used to construct a 3D model of the location. Once the model is constructed on the server the client allows you to fly around it, zoom in (using technology acquired via SeaDragon), find similar photos, and see where the photographer stood when they took a particular photo. You can learn more about the process of building the models on their site.
Photosynth is being released with 4 collections that can be viewed. To use Photosynth you will need to download an ActiveX and have an IE browser (Macs need not apply). I’ll give them more of a pass than Spaceland on being Windows/IE-only as this is coming from a research org, but it is still disappointing. I hope the final release is more inclusive.
It’s interesting to see this coming out the same week as SpaceLand (Radar post). Both use algorithms to construct their 3D models, but while Spaceland uses data to generate the images you see, Photosynth stitches together photos and they comprise the visual images. Could the Photosynth technology be used as a layer on top of Spaceland? Taking it a step further, could the Photosynth technology be used to construct new 3-D models in areas where it is to expensive to deploy the Vexcel data-gathering technologies? It would be an amazing way to fill out the maps of the world.