Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!


Forgot your password?
Graphics Software The Internet Technology

Recreating Cities Using Online Photos 87

Roland Piquepaille writes "The billion of images available from a site like Flickr has stimulated the imagination of many researchers. After designing tools using Flickr to edit your photos, another team at the University of Washington (UW) is using our vacation photos to create 3D models of world landmarks. But recreating original scenes is challenging because all the photos we put on Flickr and similar sites don't exhibit the same quality. With such a large number of pictures available, the researchers have been able to reconstruct with great accuracy virtual 3D model of landmarks, including Notre Dame Cathedral in Paris and the Statue of Liberty in New York City."
This discussion has been archived. No new comments can be posted.

Recreating Cities Using Online Photos

Comments Filter:
  • by BertieBaggio ( 944287 ) * <bob@nOspam.manics.eu> on Sunday November 04, 2007 @05:44PM (#21235065) Homepage

    This story is a bit old (well, it's from Roland after all). There was a demo of this tech by Blaise Aguera y Arcas at TED earlier this year. the two underlying components are Seadragon [live.com] and Photosynth [live.com], both of which are mighty impressive. Also, despite the Mozilla-esque name 'Seadragon', both of these technologies are actually owned by Microsoft. There is a tech preview of Photosynth up for download, but I don't think Seadragon is available yet.

    There is a video of the TED demo [ted.com], which shows off some of the things Seadragon and Photosynth can do, the including Notre Dame example mentioned in T(second)FA. The talk is also on YouTube [youtube.com].

  • Great accuracy? (Score:4, Insightful)

    by DrXym ( 126579 ) on Sunday November 04, 2007 @05:48PM (#21235103)
    It looks like an incredible idea with a lot of promise, but the shots they've shown so far look like some lumpy rejects from a plastic vacuum forming machine. There is some great potential here to involve people to generate better models by asking them to contribute pics of certain monuments with certain characteristics such as resolution, position and so on.
    • Re: (Score:2, Funny)

      by Anonymous Coward
      Imagine a 3D recreation of Goatse. The traditional Goatse photograph is one in a set, of course. Maybe they could even locate the hidden turds that aren't visible in the traditional photo.
    • This looks like a perfect system for making chocolate-bunny-style molds of famous attractions! :)
    • by reason ( 39714 )
      To me, that rather misses the point. There is existing technology ( photosynth [live.com]) that can construct highly accurate models of specific monuments given photos with location information. This is something new and different: it can be used to construct a model of any widely photographed site or monument at short notice from existing photo collections available for free on the web, with no need to gather or commission additional data. Even if the results are not picture-perfect, to me, this is much more exciti
    • The key point is that the software can sort the photos and then use them to create the model with minimal intervention.

      Where it can help a lot, is for system such as google maps, where only a few 3D models have been made. With such a system, a couple of machine could pull photographs out of a free photo database like picasa (which is already linked from Google Earth) and use the pictures to create preliminary gross "3D" models that can be subsequently put back into google earth. (until some artist design ni
  • Terrorists (Score:3, Funny)

    by nurb432 ( 527695 ) on Sunday November 04, 2007 @05:52PM (#21235139) Homepage Journal
    Only a terrorist would want a 3d map of a city. This must be stopped.
    • by pjt33 ( 739471 )
      Well, there's Sony. I think that after the Manchester Cathedral incident they promised not to use Anglican buildings, but Notre Dame is a Catholic cathedral...
  • by It doesn't come easy ( 695416 ) * on Sunday November 04, 2007 @05:52PM (#21235145) Journal
    Online photos of any physical location on any planet in the local galactic group are now forbidden due to the possibility that the photos might be used by terrorists or those who may be helping terrorists to plan terrorist attacks on said locations.

    By order of Ultra Super Secret Chief Intelligence Officer, Department of Homeland Security
  • Home use (Score:5, Interesting)

    by Cillian ( 1003268 ) on Sunday November 04, 2007 @05:53PM (#21235155) Homepage
    I wonder how many photos it needs of an object to produce a decent model and if the software will ever be released to the public - it would be amazing to be able to produce a 3d model of an object just by snapping a bunch of photos from different angles and bunging them into a piece of software. (I know there are things to create a model using a video camera and a turntable, but that's not quite as easy as being able to grab your camera anywhere and make a model.)
    • Re: (Score:1, Insightful)

      by Anonymous Coward
      It's not easy (there's a lot of work involved), but with a calibrated camera, you can, in fact, take lots of pictures of an object and turn them into a 3D model with existing software. So that part of the problem is already solved; it's really just a (very big) linear algebra problem.

      In fact, if you have a stereoscopic setup, it's quite easy, in computer vision terms, to perform the necessary calculations and correlations to do this automatically. It's a little harder with a single camera, positioned at f
      • It's a little harder with a single camera, positioned at free-form positions in space; to make it easy, you'd need to figure out a way for an AI to automatically figure out what points correlated between two pictures. Not a trivial task.

        I was thinking that Steve Mann had solved that, but on looking up his paper it seems that he's only addressed a couple of special cases.
      • Re:Home use (Score:4, Insightful)

        by Chris Pimlott ( 16212 ) on Monday November 05, 2007 @12:57AM (#21238051)
        The SIFT algorithm [wikipedia.org] created by David Lowe, might work for that; it's commonly used in panorama stitching software to automatically determine the common points between two overlapping photos. Provided you had only a small amount of rotation between photos, it could probably give you some good results.
    • I don't think there is a version of this available for home, use, but this has definitely been done. As sib. points out, the problem that needed to be solved was determining the position of the camera in space relative to the subject. Generally speaking, this is the field of Photogrammetry. One researcher I know of doing fascinating stuff with this is here:

      Marc Pollefeys at Chapel Hill NC [unc.edu]. Scroll down to see his 3d from video demo, pretty amazing stuff. Somewhere on this site is also a 3d model of a Chapel
    • This method of modelling has actually been around for some time in the consumer arena, going back to the mid-to-late 90's. The earliest one I can think of was from a company called MetaCreations (formerly Specular, the makers of Carrara Studio's older cousin, Infini-D). However, I can't recall the name of the software that did the photo-to-model stuff. I do know it functioned very similar to Google's SketchUp, which does the exact same thing.

      Google's setup is a bit more grandiose in scale though. They're ho
  • by tommyhj ( 944468 ) on Sunday November 04, 2007 @06:10PM (#21235307)
    This has nothing to do with the Photosynth/seadragon project, other than the fact that both teams use a whole bunch of different photos of the same object/setting.

    This approach tries to make the actual 3D objects from a bunch of 2D photos of varying quality. Photosynth just tried to MAP the photos in a rough 3D space. Making the actual 3D model to any degree of accuracy is really a challenge when you can't control the input images.

    The goal is different in the two cases, but they should definitely get together and exchange technologies and algorithms, because I SO much want this tech built into Google Earth!!
    • by ozphx ( 1061292 )
      You silly little man. Both Seadragon and whatever the fuck they are calling this perform a technique called "feature extraction", where they identify small features common to several images. Then they map these to a relative 3D location.

      Seadragon then displays images at the appropriate angle. This "new" stuff does a bit of micro-feature mapping between the points to make a more detailed model. Its cool, but not new. You can see it in action by getting the Seadragon preview and looking at the feature point c
    • Actually, it's based in part on Photosynth. Photosynth is the commercial version of Photo Tourism [washington.edu], which was a joint project between UW CSE and Microsoft Research. From what I understand, this new technology uses the camera pose (location, orientation, etc.) estimator from Photo Tourism and combines it with some multiview stereo algorithms [washington.edu] (PDF) to generate a 3D model.
    • Why wait for it to be built into Google Earth when it is already built into Live Maps [live.com]?


  • by TheVelvetFlamebait ( 986083 ) on Sunday November 04, 2007 @06:17PM (#21235353) Journal
    If so, I think it's time to rent a few Jessica Alba movies.
    • by BlueParrot ( 965239 ) on Sunday November 04, 2007 @06:28PM (#21235459)

      If so, I think it's time to rent a few Jessica Alba movies.

      You may also want some of the CSI image analysis software which support zooming beyond the image resolution. ; )
      • Screw that. Get the Blade Runner image analysis software that supports rotating within the image to views not even visible in the original photo!
    • by tommyhj ( 944468 )

      If so, I think it's time to rent a few Jessica Alba movies.
      actually, yes - that would be entirely possible! Altough you'd have to choose scenes with her in the same stance, wearing the same clothes, not moving a muscle. Wether ot not that is possible, is pretty much up to your opinion of her acting abilities :-D
    • by Squalish ( 542159 ) <Squalish AT hotmail DOT com> on Sunday November 04, 2007 @07:20PM (#21235835) Journal
      We certainly have software that can take a video of static objects and turn it into a 3d scene.

      We have, in TFA, software that can take frames of static objects, remove the dynamic objects among them, and leave us with a 3d scene.

      We probably have software that can interpolate a static object which is bounding a nonstatic/elastic layer (the shape of a statue under a swaying tarp, the dimensions of a box inside a grocery bag someone is swinging).

      We probably do not, however, have software that can efficiently calculate the at-rest dimensions of an elastic, mobile object(Jessica Alba) beneath a nonstatic/elastic layer (clothes). We've just barely reached the point where we can depict the behavior of the squishy, bony, muscular, hairy human body accurately, much less interpolate a hidden body.

      One wonders what it would cost to develop such software to the satisfaction of a pervert, compared to what it would cost to simply fund a movie where the pervert gets [robbscelebs.co.uk] to [fortunecity.com] do [robbscelebs.co.uk] this [robbscelebs.co.uk].
      • Re: (Score:3, Funny)

        We probably do not, however, have software that can efficiently calculate the at-rest dimensions of an elastic, mobile object(Jessica Alba) beneath a nonstatic/elastic layer (clothes). We've just barely reached the point where we can depict the behavior of the squishy, bony, muscular, hairy human body accurately, much less interpolate a hidden body.

        So what you're saying is we need naked pictures of Jessica Alba... for science.
  • Warsaw (Score:5, Interesting)

    by flyingfsck ( 986395 ) on Sunday November 04, 2007 @06:46PM (#21235609)
    After the 2nd war, Warsaw was rebuilt from photos. If you visit Warsaw today, you'd think that it is an old city. In fact, it is all new, the Poles just like it that way.
    • by Kjella ( 173770 )
      They're like most historic cities - there's an "old city" and a "new city", since inevitably the cramped roads, building restrictions and whatnot means that corporations, shopping malls and so on abandon it for a place where they can build a modern city, which is of course as bland and uninteresting as the rest. The only difference is that the "old city" is also rebuilt, but it's probably primarily for identity, culture and tourism. I haven't got the impression Poles in general are clinging more to the past
    • by Skim123 ( 3322 )
      Wouldn't you say this is the case for most "old cities?" The old parts deteriorate over time, and need to be renovated or rebuilt, but the government requires that they be renovated or rebuilt in the "classic" style for purpose of culture, tourism, history, etc.?
  • Photogrametry still seems extremely labor intensive. U can't just throw a bunch of photos at the computer and get 3D worlds out. The difference seems to be much cheaper labor than 2002. U can throw a bunch of photos at a legion of Indian artists and get 3D worlds for free.

  • There's a professor in my building who works on constructing spatio-temporal representations of information from 2D images that allow you to see a city evolve through time. Not sure if all their demos are on the website, but the ones I've seen are pretty ridiculous. http://www.cc.gatech.edu/4d-cities/dhtml/index.html [gatech.edu]
  • by nick_davison ( 217681 ) on Sunday November 04, 2007 @11:04PM (#21237351)
    You'll notice the Notre Dame model was of the endlessly photographed front aspect.

    I wonder how much detail such a technique can pull on the rear of the building? Or the back of Rodin's Kiss? How about the top surfaces of the Statue of Liberty?

    Of course that won't stop CSI, next season, from pulling a bunch of pictures from Flickr of the front exterior of a building, reconstructing a complete 3D model, open accurately hinging doors, travelling inside, going down in to the basement and looking at the reflection in someone's eye to identify the killer.
  • "...another team at the University of Washington (UW) is using our vacation photos to create 3D models of world landmarks"
    More like make a 2D image of something you are likely to find in a gift shop on the grounds of these sites. Really... Take a look.
  • by snavely ( 1184409 ) on Monday November 05, 2007 @12:03AM (#21237707)
    I was a co-author on this work -- it's great to see it on Slashdot =)

    I also worked on the Photo Tourism project (which is related to Photosynth). There's a big difference between Photosynth and this new 3D reconstruction work, in that Photosynth takes a photo collection and reconstructs camera positions and a sparse point cloud (a set of disconnected 3D points floating in space), while in this new work we build *dense* 3D models of scenes (in the form of polygon meshes). Dense models are usually much better for use in applications like computer graphics, since they can be used to render scenes with much more photo-realism.

    These two problems require different algorithms to solve---for this dense problem we use a technique called multi-view stereo, but we made it work with images taken by many different people under different conditions.

    - Noah
    • I have a copy of Multiple view geometry in computer vision 2nd edition by Hartley and Zisserman, which I intend to read in my copious spare time. I find this area of computing fascinating and think it will have a huge impact on mapping. I would be grateful if you would recommend any other good references in this area? Thanks
  • Where can I download the software? Is there a open source project on this?

    Over the years I have seen many different attempts from reconstructing scenes from digital pictures, some of them being very impressive, but for some reason I have never been able to find a working piece of software or a open source project dealing with this. Yes, there are several nice frame works available. Or I am I just looking in the wrong places? I did find some commercial packages, but these are beyond my budget.

    Is there s

  • Like, looking at the model of Notre Damn, I was thinking, surely someone has taken pictures of the sculptures over the door. The model looks like a blob. And you are not really plotting entire 3D landscapes, but, from what it looks like, creating 3D models of single structures, and you are not even adding color. I mean, at least the Microsoft project creates 3D landscapes. I think the two projects should merge, then you could have 3D models with actual picture overlays.

    I don't know, I have just seen so many

Maybe you can't buy happiness, but these days you can certainly charge it.