Hi there,

How feasible is it to decouple the camera/render viewpoints in artoolkit ?
E.g., in 2D:


CAM0        CAM1
    \       /
     MARKER
        |
        |
       EYE

I was thinking of a scenario like this:
-Both camera's determine relative pos/ori of marker
-If both find the marker: average absolute pos/ori
-Render for viewpoint at EYE
 (Assuming that relative position of EYE wrt CAM0 and CAM1 is known)

What I'm trying to find out, is wheter artoolkit is usable for
a device as this one:
http://homepages.cwi.nl/~robertl/pss/pss.html

  Bram

-- 
Bram Stolk, VR Engineer SARA, Amsterdam.   tel +31 20 592 3000

"Windows is a 32-bit extension to a 16-bit graphical shell for an 8-bit
 operating system originally coded for a 4-bit microprocessor by a 2-bit
 company that can't stand 1 bit of competition."

Bram Stolk wrote:

>How feasible is it to decouple the camera/render viewpoints in artoolkit ?
>
this depends on your setup.

if you have a single camera and the render view is showing that camera's
video as background (as most basic ARToolKit setups do) then it probably
makes no sense to put the renderer's (virtual) camera to another
position than the position determined by ARToolKit since your rendered
content would simply not match your video background.

if you want to use optical see-through HMDs or shutter glasses (as on
that webpage) then this is the way to go. all you need to do is to
offset the camera's position a little bit for the left and right eye...

bye,
Daniel

Hi Bram,

I don't beleive you'd need two cameras to do this. I did something
similar in concept recently - an art installation in which the user can
use big paper cubes with the marker printed on every side, to paint on a
virtual canvass (projected on the wall). This required the camera to be
facing back from the projection surface to the user who was standing
near the projector, and also required the camera to be offset slightly
from the projector. If you look at the code here you can see what I mean:

http://mccormick.cx/dev/argraf/arPaint/

The GL projection matrix had to be inverted and moved around because of
the position and orientation of the user and the projection surface. It's
all hardcoded into the drawing functions because this was a one-off art
installation, but you could modularise it and do something similar.

So in short, I think that you could use ARToolkit to build something
similar to the link that you sent.

Best regards,

Chris.

On Mon, Dec 05, 2005 at 05:34:30PM +0100, Bram Stolk wrote:
> Hi there,
> 
> How feasible is it to decouple the camera/render viewpoints in artoolkit ?
> E.g., in 2D:
> 
> 
> CAM0        CAM1
>     \       /
>      MARKER
>         |
>         |
>        EYE
> 
> I was thinking of a scenario like this:
> -Both camera's determine relative pos/ori of marker
> -If both find the marker: average absolute pos/ori
> -Render for viewpoint at EYE
>  (Assuming that relative position of EYE wrt CAM0 and CAM1 is known)
> 
> What I'm trying to find out, is wheter artoolkit is usable for
> a device as this one:
> http://homepages.cwi.nl/~robertl/pss/pss.html
> 
>   Bram
> 
> -- 
> Bram Stolk, VR Engineer SARA, Amsterdam.   tel +31 20 592 3000
> 
> "Windows is a 32-bit extension to a 16-bit graphical shell for an 8-bit
>  operating system originally coded for a 4-bit microprocessor by a 2-bit
>  company that can't stand 1 bit of competition."
-------------------
chris@m ...........
http://mccormick.cx