Robert Grzeszczuk (rg)
Mon, 21 Jun 1999 14:28:24 -0700 (PDT)
There are a couple of techniques I can think of. First involves using texgen
in the eye reference frame, the second uses the fact that Volumizer slices
volumes in planes of constant depth.
If your volume is not textured (hard to tell from your description), you can
texture map a 1D texture containing depth encoding (e.g., transparency and/or
color coding: semi-transparent blue is far away, white in the middle, opaque
red is close by). Standard glTexGeni() command with GL_EYE_LINEAR argument will
do the trick.
If your volume has textures already, and your hadware supports 3D texture
mapping, you can take advantage of the fact that Volumizer will sample your
volume in planes that are "equi-deep." Instead of using white to draw the
volume slices (see voglDraw.cxx) use the color scheme of your choice to affect
the appearance of the near vs. far slices.
Does any of the above suggestions solve your problem?
-rg
On Jun 21, 9:16pm, Gero von Wagner wrote:
> Subject: Applying distance information to volumizer data
> Hi everybody!
>
> We are working on a simulation of a physical detector system for
> radiation. We are using volumizer to visualize our results.
>
> Because the detector's sensitivity highly depends upon the distance from
> the radiation source (~radius^2), we are now searching for a way to
> apply information from the depth buffer to the accumulation buffer.
>
> The idea is as follows:
>
> The z-coordinate will be weight according to the front and back clipping
> plane. For each texture plane rendered into the volume we should then be
> able to load the texture value with a certain "punishment" factor
> depending on the distance (the value in the z-buffer). It seems to us
> that volumizer doesn't work this way by default (intensity doesn't
> change if you put the volume farther away).
>
> Just to prevent us from re-inventing the wheel: Is there anybody who has
> linked depth and accumulation buffer before? Or can tell us if our
> approach can be successful at all?
>
> Thanks a lot.
>
> Gero von Wagner
> --
> Gero von Wagner
> Institute for Biomedical Engineering
> University of Karlsruhe
>-- End of excerpt from Gero von Wagner
This archive was generated by hypermail 2.0b2 on Mon Nov 01 1999 - 14:02:50 PST