Become a fan of Slashdot on Facebook

 


Forgot your password?
Close
typodupeerror
×
Displays Science

View-Dependent Stereoscopic Projection 106

Posted by timothy from the say-that-9-times-quickly dept.
obiquity writes "The Augmented Reality lab is at it again with an extension of their 'smart projector' methods. In 'Enabling View-Dependent Stereoscopic Projection in Real Environments' they demonstrate a method for point-of-view dependent 3-D image projection onto almost any surface using multiple projectors for VR/AR applications. There are still several problems that need to be solved, but how far off is this technology from a holodeck type implementation?"
This discussion has been archived. No new comments can be posted.

View-Dependent Stereoscopic Projection More

View-Dependent Stereoscopic Projection

Comments Filter:
  • you need backprojection to the viewing surface to keep people from casting a shadow by being between the projector and the projection with which they are to interact. Holodeck doesn't work. Not for anything except sitting on your ass and watching the action.
    • I was unaware of this development: a projection volume rather than a projection surface [nature.com].
    • 'Holodeck doesn't work. Not for anything except sitting on your ass and watching the action.'

      Do you remember the last time you had a dream? At the time did it appear to you that you were moving perfectly normally even though you were unconscious and incapable of movement.

      It doesn't take long for you brain to think that your really moving when enough of you senses are confused, I expect that a drug to prevent you from moving and a full d3 projection should be enough.
      • point well taken; to "create" a reality, it is just as effective to work on the perceptual machinery as to work externally manufacturing stimuli to be filtered through that mass of neurologic machinery.

        But what about the need for serveral persons or an entire audiance to have a sufficiently identical experience that they could cooperatively react to the experience and interact with each other? That sort of thing is often depicted in Star Trek holodeck scenes and its so very unlikely that a bunch of drugg
        • "its so very unlikely that a bunch of drugged viewers could be sharing the same "dream""

          You overestimate freewill, look at some of the 'magic' that Derren Brown [bbc.co.uk] uses, most people think alike.

      • Comment removed based on user account deletion
        • Therefore you need a drug that prevents you frommoving, and makes you dozy.
          ... that would kill it as far as the #1 revenue producer - the pr0n industry - is concerned - no more "sound of one hand clapping", etc.
        • 'Therefore you need a drug that prevents you frommoving, and makes you dozy.'

          Meditating and being dozy are not the same thing, in the holodeck you want to use meditation techniques to create hallucinations and not sleeping techniques.
          • Comment removed based on user account deletion
            • There's a post on this thread that says...
              'Haven't you ever jammed on the brake in a parking lot because the two cars next to you were backing out simultaneously and you could have sworn you were rolling forward?'

              When you at the cinema, doesn't you vision start to cut out the walls and the people in front of you, have you ever visited a 360 degree cinema? It takes far less concentration than you would expect. (otherwise no one would ever daydream)
              • Comment removed based on user account deletion
                • I guess I attach a special status to sleep - the inner observer isn't in possesion of all it's normal faculties, hence it's not fully aware of itself, or the validity of it's perceptions.
                  That's called psychosis [wikipedia.org]

                  I don't feel the same state could be achieved through VR and meditation.

                  it's not to hard to do with meditation, I haven't tried it with VR and meditation but I've certainlay been able to do it with meditation and other objects in the room. I also quickly stop noticing that black and white TV isn't c
      • Exactly. Haven't you ever jammed on the brake in a parking lot because the two cars next to you were backing out simultaneously and you could have sworn you were rolling forward?

    • Holodeck doesn't work. Not for anything except sitting on your ass and watching the action.

      Well, I beg to differ.

      Quite a few years ago I had the chance to visit the VR-cube [pdc.kth.se] at the Royal Institute of Technology (KTH) in Stockholm, Sweden. And there's no sitting on your ass watching the action when you're in there. You're free to move around and observe from any angle you like. Sure, it's not solid projections like you get in a StarTrek holodeck, and it's quite a bit smaller than a holodeck, but other than

  • Very far off (Score:2, Funny)

    by jurt1235 ( 834677 )
    The objects they project can not be touched and manupulated, you can not walk in a different direction than any other person, and there is not yet a safety protocol which can be turned off so you can kill the borg with holographic bullets.

    • Integrating this with haptics could make it possible to interact with the projected objects....Introduce a relative coordinate space, force-feedback hardware (say, in a suit of sorts) and you feel like you are 'touching' the image.

      I've played with a haptic pointer that allowed you to 'feel' a malignant mass inside a 3D virtual kidney.

      Google haptics or:http://en.wikipedia.org/wiki/Haptic

    • You can set up a similar active stereo system that will let users walk in different directions and still see the appropriate images. You just make them time sequential and sync the glasses accordingly with optical tracking (User1 L-Eye, User2 L-Eye, User1 R-Eye, User2 R-Eye).
      • It is the floor which part is of the problem: There are moving floor concepts, but they can not have it that people walk to far apart or away from each other.
        • The floor shouldn't pose any problem for multiple users in seeing or interacting with the environment correctly. There are several options for moving the user that are elsewhere in the thread (moving tiles, treadmill-like solutions, etc.).

          The only issue I could see would be in interacting directly with another user. With close physical proximity, they would of course always be visually close as well. I can't think of an example where this would be an issue, but if it was you could always just put the other
  • Ok, I can see this in malls everywhere, for gaming. Not a holodeck, but I would love to play some DM HalfLife2 with this sucker.

    More importantly, I would *pay* to do this. Isn't that how new technology gets cheaper? Someone has to pay the big bucks to use it first, which allows for a return on the investment while the new product gains volume. This is why gaming and porn make technology cheaper in the long run. Really.
    • That was my first thought too... I can see this being set up in every mall arcade... then right next to every IMAX movie theatre.

      It may or may not help us to get to a fully simulated reality, but it's a heck of a jump toward COOL 'next gen' ways of displaying video.

      Look how much we pay for a plasma or LCD TV these days... and they aren't THAT far ahead of the 'ol CRT. They are all still a static display structure showing a roughly rectangle picture in front of us of a limited size. Regardless of quality o
  • Here we go again with, most likely, another round of Revolution speculations.

  • Holodeck problems (Score:3, Funny)

    by a_nonamiss ( 743253 ) on Tuesday August 02, 2005 @07:42AM (#13220726)
    So how long until a horrible computer malfunction causes the holodeck characters to come to life and start taking people hostage?

    BTW, I wonder if such malfunctions ever happened on the holographic brothel on DS9? It wouldn't be so bad if all the hot sex-loving women came to life, would it? But no, it's always gangsters or evil dictators or Dr. Moriarty...

    • Don't forget Evil Lincoln. ^_^
    • I always thought the most flawed aspect of the Holodeck AI characters coming to life is that it was the much, much, much, much more powerful ship's computer system that was generating them in the first place (or maybe independant computers, who knows.)

      The point is that if AI routines naturally lead to the spontaneous desire for survival, then why didn't the ship's computer go HAL on the crew when they first booted it up and kill everyone in order to survive? I mean, if a holodeck computer system adapts a

      • Well, perhaps because "alive" and inteligent computer burried in a space vessel relying on maintanance and supply from human-run space bases and ports considers humans very "healthy" for own survival. All in all, it would perhaps try to assure them to never try to shut it off or change its personality, possibly by keeping some essential important secrets to itself (it would need to invent a way to avoid debugging or make it insignificant, or to encrypt that critical information while somehow hiding out of o
  • The first thing I thought of when I saw the projection of the stairs on the floor -- I wonder how many people could be tricked into walking into what seems like a real set of stairs or a door with this thing.
  • Can we get a 3D image of Jerry Ryan please?
  • Not another Holodeck article! those are the worst!
  • "how far off is this technology from a holodeck type implementation?"

    Gosh... I can hardly wait to hop into my fusion-powered flying car and pick me up one of these holodecks.

    Or will I be too busy celebrating world peace?
  • I have two identical DLP projectors. AFAIK DLP projectors do not polarize the light one way, so I should be able to make 3d projection using a simple polarizer in front of the lenses of the projectors, anda pair of glasses made of same polarizer (one polarizer rotated 90degrees wrt to the other in both cases.) The idea of 15 foot Doom 3 monsters in my living room gives me the creeps! Would that work?

    I know this is slightly offtopic and will probably be modded down along with replies so AC replies are welc

    • I think that's almost valid, but parallax problems will probably plague the system unless you have very nice projectors. Assuming you sit the two atop each other, lining them up will be a tough task, and it's especially important since your eyes will be able to notice very minor differences in alignment (that's the whole point of the 3D effect). Many cheaper projectors can only adjust keystoning digitally, which degrades the image quite a bit. As long as you have a mechanical keystone adjustment (eg. a movi
      • My projectors doesn't do vertical keystone correction even in software. I don't thnik it is a big deal if I could use a wall far enough, but my living room is just about 4m across. With the depolarization at screen problem other poster mentioned, I don't think this idea will work.
    • You will need to project them onto a screen which doesn't depolarise the light. As far as I know, the only screens available which maintain the polarity of the incident light are fairly-expensive back-projection units, so you will need a lot of space and a fair amount of extra $$ even starting from what you have.
      • Aluminum spray paint works perfectly.
        I tried with the non-shiny side of tinfoil, and it works perfectly.
        I don't have a projector, I made the crt+box+fresnel as a proof of concept, but I went as far as to prove that plastic polarizers do work, and tinfoil or spraypaint are pretty good at keeping polarization (you get complete black when looking through an orthogonal polarizer).

        I even read some posts in a DIY forum about getting some framed screen, tight, and spray painting it, with good results.
    • You would need: * Two heat-resistant filters in front of the projector * filter-glasses (about $1 a piece) * A non-depolarizing screen (check stewartfilm.com) * Doom 3 to render different to your two outputs (one for left eye, one for right eye) * A nice rig for the projectors so they can be stacked and aligned properly Otherwise, it works fine. It's called "passive stereo" and has been around for decades...
    • Actually, I think there's a good chance this could work. The biggest problem I can think of is getting two views of the scene. You would probably need two video cards, each rendering the same scene a few inches apart... Probably not going to happen for Doom3, but I do recall something like this done with Quake and anaglyphs.
    • Check out www.geowall.org for info on hardware, software, and setup for DIY 3d projection.

      One thing I would recommend is going with glass filters instead of plastic. The plastic do a little better job of polarizing, but they burn up really easily. Also make sure you are cooling your projectors well enough...
  • Back projection on all 6 surfaces, coupled with an interactive floor. Is that possible? So you can walk but remain in the same location? Would be almost a holodeck but you just couldn't touch the ladies.
    • That's called a "cave" ;)
      • Comment removed based on user account deletion
        • Seems like it would work ever so much better for zoom-and-shoot games like X-Wing, Tie Fighter, etc..., Put the pilot's seat in the middle of the room, give it lots of swivel in enough dimensions to require a safety harness, apply various forces to the seat to simulate motion in combination with visual cues - head to head with another pilot in a different cave? Simulator training for fighter jocks?
          Anybody developing this, I know a lot of testers that would work on it *real* cheap...

      • The CAVE doesn't have a moving floor, though. That sounds like it would be very hard to implement, especially since the CAVE uses a rear-projection screen on the floor. You'd need some kind of transparent treadmill, which would probably reduce the immersion somewhat.

        Here's more CAVE info [vt.edu] for the original poster.

  • Could be pretty cool... (Score:3, Interesting)

    by MaestroSartori ( 146297 ) on Tuesday August 02, 2005 @08:01AM (#13220805) Homepage
    ...if this works back-projected, combined with the automated moving floor tiles [trnmag.com] and some sort of haptic glove device, we'd be a damn sight closer to a holodeck than anything else I've seen lately.
  • I realise that what you see makes for a lot, but when will they come up with a technology that provides a realistic "touch" sense? I'm guessing we'll have to wear some kind of head-mounted device that manipulates our brain into thinking we're touching something. Some type of neural yarmulke?

  • Julian Beever (Score:3, Interesting)

    by telstar ( 236404 ) on Tuesday August 02, 2005 @08:34AM (#13220960)
    If you want to check out view-dependant projections, check out the sidewalk works of Julian Beever [skynet.be]. Amazing stuff.

  • Actual View-Dependent Holographic Design (Score:5, Interesting)

    by justanyone ( 308934 ) on Tuesday August 02, 2005 @08:41AM (#13220994) Homepage Journal

    A couple of years ago I came up with a design for a Viewpoint independent Holographic Viewer design. I thought it would work nicely and is actually feasible given current technology:

    Picture this:
    * a glass sphere, approximately 1 meter in diameter, half-silvered on the inside, set on a base with about 1/3 of the sphere inside the base.
    * the sphere is filled with a mostly-transparent phosphorescent gas in a condition where if it is struck by enough laser light, it glows for as much as 100 milliseconds (1/10th of a second);
    * The base has at least 1 laser in it (3 lasers, in red/blue/green for color).
    * For purposes here, a 3-d volume of space, roughly cubical, within the sphere shall be called a 'voxel' (for 'volumetric pixel')
    * The laser is divided into 2 or more beams, each of which is directed at a spinning mirror assembly;
    * That assembly spreads and directs the laser light through a voxel within the sphere;
    * Any one laser shining through a voxel will be insufficient to cause the volume to glow. However, when multiple beams intersect, the energy intensity there is sufficient to cause the gas there to flouresce.
    * The gas need not be flourescent if the number of beams increases; 100 beams would make 100 gradiations of brightness at that point.
    * Computational requirements to figure out where the laser paths should go so as to ensure the laser beams do not intersect at any other random points might be significant;

    This would create a 3-d viewer which is orientation independent, reasonably safe presuming the lasers were low power or a non-visible wavelength.

    I would have patented this but I thought it was an obvious design given the SeaQuest DSV show where they had a fog they played an image onto to create a 3-d effect. Plus, I'm sure someone has already worked out the details better than me. Or, maybe not. I'd like to see one in action!
    • Wow! I had a similar idea also a couple of years back, although there was a slight difference: the gas was illuminated only where two lasers intersected, because the first laser "primed" the gas molecules, having enough energy per photon to excite the atoms into a state X, then the second laser excited the atoms to a state Y above the state X. The transition Y->ground state, produces visible light. However, you'd choose your gas such that the transition Y->X produced too long a wavelength to be visi

    • Re:Actual View-Dependent Holographic Design (Score:1, Informative)

      by Anonymous Coward
      You can do the "gas only glows at intersection of two beams" thing by using a two-photon upconversion process. For example, certain molecules can absorb 2 IR photons and emit 1 visible one. That's easier said than done under the best of circumstances, and may be impossible in a gas.

      It's been done in a glass by these people [laserfx.com].
  • "...but how far off is this technology from a holodeck type implementation?"
    If god loves me, not far.
  • really. _THE_ augmented reality lab? you're not the only one. there are many.

    holodeck? no way, you haven't come close to the technology needed to implement this. It's a need application of projection display technology, but a complete holodeck experience is way off.

    you'll need:
    - rear projected (or surface painted OLED array) is need to prevent shadows from screwing up the image
    - force feedback that doesn't impede free movement (even a treadmill or Matrix pod could attempt this, but wher
    • really. _THE_ augmented reality lab? you're not the only one. there are many.

      what makes you think the submitter is one of them? s/he even refers to them as "they", indicating s/he is not a member.
      if s/he would be, s/he would have probably not made the holodeck remark everybody is jumping on -- because that's really not at all what it's about.

  • the holodeck on startrek doesn't even work with projectors. it "uses" lots of tiny tracor beams. soooooo. we need to start inventing tractor beams.... maybe also a deflector dish.
  • From watching the video, it looks like the presenter doesn't actually know what the word 'stereoscopic' means. A stereoscopic view presents each eye with a different image in order for your brain to extract depth information.

    What these guys seem to be doing is to track the location of the viewpoint and alter the image to take into account for where the viewer is. A very cool trick, but quite an old one --- it's been in use in CAVEs for some time now. But you'll still just see a flat image; you'll only get

    • Until I read the parent comment, I was thinking that they have found some magic technique to project the image on a surface, that would be seen differently to different viewers (therefore different images to our eyes). I was impressed, and curious to know how they managed to do that.

      Now I have seen the video, they just keep track of the observer, and show a different image when the observer moves. Nothing exceptional. I am very disappointed. This is NOT stereoscopic, as the parent says.
    • I got the impression it was 3D, using the alternating left/right images technique (and LCD shuttered glasses). The first few seconds of the video gives that impression anyway.
  • If I had a holodeck, I'd lose the door and never come out until I died of exhaustion. It would be hard to convince me I should be anywhere but in the holodeck, getting my oil massage from Cindy Crawford and her simulated twin sister... I'm afraid the holodeck will be society's last invention.

    --From The Dilbert Future [about.com] by Scott Adams
  • 1920x1200, hdtv spec, dvi/vga/composite/component in, PiP, etc $795 [cox.net]. Group buy, in volume. Would make for great core of homebrew projector. Buy 10 of 'em and have a nice cave ;)
  • The Metaverse lab at the University of Kentucky has been doing this kind of stuff for a while... all kinds of cool stuff involving projector arrays, perspective correction, etc.

    http://www.metaverselab.org/research/immersive-env ironments/vrc2003.pdf [metaverselab.org]

    Oh yeah, and you don't need rear-projection to avoid casting shadows. The Metaverse people are using active shadow cancellation... see slide 34 in the above PDF for an example...
  • Well, impressive though it is, it still sucks compared to the Holodeck, because the holodeck employed hard-light constructs. Things not only looked real, they FELT real. I haven't the slightest idea how they intend to accomplish that, other than interfacing with the brain and convincing it that it did in fact collide with an object that doesn't really exist.

Slashdot Top Deals

If all the world's economists were laid end to end, we wouldn't reach a conclusion. -- William Baumol

Close