Home »
Public Forums »
archive »
PV-WAVE: shade_surf question
PV-WAVE: shade_surf question [message #1161] |
Tue, 13 July 1993 14:52 |
Todd M. Kulick
Messages: 1 Registered: July 1993
|
Junior Member |
|
|
My goal is to render shaded surface images of actual
landscapes. I have elevation and color data (LANDSAT). The color
data is 24-bit (RGB) color data. I am trying to use the 'shade_surf'
procedure with the 'shades=' option to accomplish my goal. I best-fit
the 24-bit color data so that my 'shades' data is of N colors, not
256**3. It is my understanding that the 'shade_surf' procedure will
only take byte arrays for the 'shades=' option. If anyone knows of a
neat (or even non-neat) way of overlaying 24-bit color data onto my
elevation data I would be very interested. Still, this is not my only
problem. Even once I concede to best-color-fit my data (which is fine
since I may end up displaying the final image on a PsuedoColor device
anyway) I cannot achieve my goals. :(
The procedure does some form of Gourad shading. It seems that
if two adjacent points are shaded with colormap indices '1' and '10'
then any internal interpolated points do use colormap indices '2'
through '9'. This is not appropriate for my circumstances. Since I
have a full color image that has been best-fit to N(<=256) colormap
entries the adjacency of colormap entries has no real meaning. What I
get is two points that are light green and dark green with orange,
purple and white in between them. Obviously the resultant image is
not a good representation of the actual landscape. Again, I request
help from any of you netters who might know either how PVWAVE is
actually shading my images or how I might convince it to do what I
want.
Pertinent statistics:
I work on Sun Sparc 10s and/or Decstations running SunOS 4.1.3
and/or Ultrix 4.2. PV WAVE version is 4.00. All replies to me
please; I will post a summary if requested.
-t
--
kulick+@cmu.edu | `She had a sorrowful, happy-type of frowning smile . . .'
| -- Satan and Adam (Down Home Blues)
|
|
|
Current Time: Sat Oct 11 14:43:34 PDT 2025
Total time taken to generate the page: 1.35883 seconds