Photosynth
I wouldn’t normally post a link to something so corporately hyped, but I do think that new imaging technology needs to be grabbed creatively and critically as soon as possible in order to keep it democratic. Especially when said tech is depending on accessing and using everybody’s personal photos, apparently no matter where they are posted on the web. So, since Photosynth is new enough to still only (afaik) exist in Corporateland, I apologize for sending you to a bunch of Microsoft spin-o-matic vids, but Go Forth and Make It Your Own.
Specifically, this stuff is geared to ruling the universe of photography, but as we draughts-people types showed Flickr, we can appropriate tech according to our illustratorly whims. According to a spokesbot, “… we take the the world’s photos and organize them in 3d…” and “the ultimate dream is that we merge the real world and the virtual world until it is ultimately seamless…”. OK, so now imagine doing that with our drawings. Also, this goes some way to providing a solution to the lack-of-detail-available-in-web-images issue I complained about recently. Here’s the MS site, but the link above is of a live presentation and much more watchable.
Awesome, how about a view of St. Peter’s composed entirely of sketches from different views? I’ll get started on that right now :)
This must’ve been so corporately hyped, I missed it altogether. I think this is totally and absolutely mindblowingly awesome.
Beyond just getting really great 3d reference of just about anything you could want, this technology could eventually allow people to virtually transport themselves into a compellingly real virtual location anywhere in the world. And if you can interpolate architecture from a series of still images (and I’m sure, eventually moving ones) you can interpolate a crowd of people, or damn near anything. Everyone will have their own hollywood studio on their PC, and be able to generate realistic virtual environments that look very much like real environments in practically real time. These images will have more fidelity, in some ways, than what we can see with our eyes.
Anyway, this is completely awesome.
This is based on Photo Tourism, shown at siggraph last year (watch the longer video if you’ve got the bandwidth). Lower resolution youtube variant is here.
excuse the above semi-coherant ramble. Let me just leave it at I was impressed.
…first of all ,i have to say this very cool
but there are some things that can’t be done:
1) it will never work with illustrations, art, …nEver
2) it will never work with people, animals, etc.
…the above two will only work if you do it in a studio or something, which already happens (in studios, with a lot of work, nOt computationally)
to do this automatically, you need a staggering number of pictures, of solid, stable things. Otherwise, image processing tools won’t recognize the same thing, from sLiGhtly (…this is important) different angles. iT only works with a lot of useless picture data, not art.