[Fluxus] kinetic
evan.raskob [lists]
lists at lowfrequency.org
Sun Dec 12 07:21:28 PST 2010
That's really cool, what a great idea to use a photo camera... thanks for sharing that!
Best
Evan
On 2 Dec 2010, at 15:48, Scott wrote:
> You are totally correct, after a little research (done by a friend, who's been deconstructing a kinect and reading the patents, not by me!).
>
> The original "zcam" was a time of flight system.
>
> The Primesense system is a structured light system, but it does seem to take advantage of the construct/deconstructive interference of the IR LASER it uses. It looks like the laser passes through an HOE (holographic optical element) to create the structure portion.
>
> Here's my friends blog post on his explorations:
>
> http://www.danreetz.com/blog/
>
>
> On Thu, Dec 2, 2010 at 1:13 AM, David Griffiths <dave at pawfal.org> wrote:
> On Wed, 2010-11-17 at 08:10 -0800, Scott wrote:
> > Here's a link to the company that makes the underlying technology...
> > or one very similar to it: http://www.primesense.com/?p=487
> >
> > What's interesting about the zcam is that it actually uses
> > TIME-OF-FLIGHT to compute depth, so IR reflectivity shouldn't really
> > matter!
> >
>
> It does if the pulse they are sending out to measure is IR. IIRC the
> picture image is captured at a different time to the depth. Sunlight and
> reflective materials were a problem with all the cameras we were looking
> at.
>
> In any case I thought the kinetic camera was using structured light - as
> it has a projector which projects a pattern on the scene, ir camera to
> measure it and a normal rgb camera.
>
> The ir camera and rgb camera are seperate, so you get the shadow effect
> you can see in the depth images - time of flight cameras use the same
> lens so are a little better.
>
> cheers,
>
> dave
>
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.pawfal.org/pipermail/fluxus-pawfal.org/attachments/20101212/7059d63e/attachment-0001.html>
More information about the Fluxus
mailing list