Microsoft Photosynth

WilloxH asked:

Photosynth takes a large collection of photos of a place or object, analyzes them for similarities, and displays them in a reconstructed 3-Dimensional space

Related Posts

bookmark bookmark bookmark bookmark bookmark bookmark bookmark bookmark bookmark bookmark bookmark bookmark


  1. jayseepingkian Says:

    Totally amazing. Microsoft brings innovation once again.

  2. bv90andy Says:

    man, this vid is from July 2006, they didn’t have i7, maybe a Core 2. but the Graphics card is more important here, I think.

  3. filipkir Says:

    If you look carefully it lags even on that demonstration PC. And they use the most powerful core i7 processor.

  4. bennyb220380 Says:

    Your face seems stupid.

  5. Dratio Says:

    Wouldn’t it be cool if it could generate a simple sky and ground backdrop so it feels more like you’re in a place rather than a massive row of photos floating in black space?

  6. bv90andy Says:

    This is how progress is made, by inventing, and then finding a useful task for that invention, and making it better.

  7. bv90andy Says:

    this is not new technology, only this is on a much larger scale . And with many many features, I think it needs a good CPU for it to work fast enough.

  8. sinnison23 Says:

    Another example would be, say, NASA collects several hundred still images from a space probe, or the Mars rover, etc. They can then take every one of those images and collate them into a panoramic image which you can scroll through in a near-3D image to gain different perspectives and details of the surface of Mars, even incorporating overhead aerial images. Et cetera.

  9. sinnison23 Says:

    The best examples of what this program is capable of are Obama’s inauguration. John King of CNN had viewers send in any pictures they could take from any part of the assembly, from any kind of capture device. He then used Photosynth to create a panoramic image, which he could literally pick a point and zoom in or out for greater or less detail, pan to another point and change focus, even pan to different angles of the same point. All done with single images.

  10. GanjaOnDeck Says:

    It reminds me of the movie deja vu

  11. ApeWithBone Says:

    seems stupid

  12. esoeso122 Says:

    That is pretty cool.
    Microsoft is still the main dog.

  13. Philscbx Says:

    Simply review the first 30 seconds of this clip.
    He explains it clearly for anyone.

  14. grandgret Says:

    This is what happens with a group of people with too much time get together. I’m kind of irritated by the whole virtual organic model, the social objects of living oganisms in the digital realms of social combinations. Seriously, everyone is trying to come up with some crap to get bought up by wither google microsoft. Why don’t they spend more time trying to figure out how make their existing software more reliable.

  15. DustyMooseHamilton Says:

    Just like the internet.

    lololololololololololololololo lzzzzzzz

  16. DustyMooseHamilton Says:

    Just like the internet.


  17. Philscbx Says:

    Because you no longer have the Ford Probe after you totaled it out. If there are enough views, it is appearing to be 3D like. I think any 3D physical model was taken out of context to a point. There are 3D model creations for machining and manufacturing prototypes out of foam beads. So it is possible in a small scale.
    If your told or read an actual point, then a link would clear up the fact.

  18. antoniosoul Says:

    Thanks for your explanation, but I’m still missing the point, as are a lot of other people, apparently. I checked out this video because I was told this software automatically creates 3D models, and a lot of the commenters here seem to believe that it does. I mean, why not just shut off the computer and go outside and sit in an actual Ford Probe? And maybe drive somewhere while your at it, you know, experience actual reality vs. fake reality. Cheers.

  19. Philscbx Says:

    Apparently You missed the whole concept.
    The photos you took of your car. It could be a Ford Probe. You over time took 27 images of specific points.
    Other people with the same car each took 27 images.
    Multiply 17,000 cars times 27 images each.
    No 2 images are actually alike. Close, but different.
    Therefore, selecting one image you took will automatically seek out images by others that let you see even more detail or angle.
    How’s that, pretty dam cool if you ask me.

  20. antoniosoul Says:

    I don’t get it, most 3D animations are of imaginary places. How are you going to get a huge collection of photographs of an imaginary place? This doesn’t appear to create an actual 3D model, it just simulates the experience of an actual 3D space in 2 dimensions. There were no smooth transitions or pans, he just zoomed in and out of different views. Or am I missing something here?

  21. Maketsup Says:

    Nice! When this technology gets better, gamemakers can use it to make 3D worlds, without 3D programming.

  22. 25mmpolycarbonate Says:

    Wow imagine being able to go on the web and find places or thinkgs of interest and then view them in real time through streaming images…oh wait a minute…we created this over a hundred years ago and now call it video…wanna see St Peters then search it right here on you tube and voila…also many places have now produced 3d virtual tours…this product is everything thats wrong with technological developments today…no direction just random ideas allowed to become finished products…

  23. antago Says:

    so cool! Can’t wait til google images becomes a 3D tour

  24. RT7 Says:

    This is pretty kewl, but think about it.. you could only do this if you had like thousands of pictures, with the same lighting and everything? like lol not kewl.


RSS feed for comments on this post. TrackBack URL

Leave a comment