On Wed, Apr 23, 2008 at 1:07 AM, Markus Neteler <neteler@osgeo.org> wrote:
On Fri, Feb 1, 2008 at 2:16 PM, Benjamin Ducke wrote:
...
The output vector map will now also preserve the 3D coordinates of
the original input points. However, the centroids for the triangles
are all wrong (placed at z=0.0). I don't see a quick way of fixing
this right now.
I have fixed it using Vect_tin_get_z() for delaunay (someone needs to
do it for voronoi).
Is there anyway to display a TIN in 3D with GRASS? In NVIZ, it just looks
like a Delaunay triangulation network unless there is something I'm not
doing.
Michael
I wonder if I could colorize the 3D TIN somehow using v.colors.
So far it does not have a DB connected.
__________________________________________
Michael Barton, Professor of Anthropology
Director of Graduate Studies
School of Human Evolution & Social Change
Center for Social Dynamics & Complexity
Arizona State University
Is there anyway to display a TIN in 3D with GRASS? In NVIZ, it just
looks like a Delaunay triangulation network unless there is something
I'm not doing.
Make sure your 3D triangulation output "Map is 3D".
AFAIK faces will always be based on a single color (some todo).
off on a tangent:
As I've ranted on about before, TINs are really just the point data that
define them, so in general you'd be much better off extracting the points
and using a real interpolation module like v.surf.rst to make a surface
from those points. The exception is from e.g. the poster the other day
who's point data was not uniformly distributed, density increased by
orders of magnitude as you get nearer to an object of interest (dynamic
computational mesh size)
e.g. http://gfs.sourceforge.net/examples/examples/tangaroa.html
v.surf.rst does the same sort of thing during processing for no-data
search areas, but the final product there is at the region resolution.
The problem the poster the other day had was that the overall grid at a
resolution appropriate for the coastal areas would make a massive array
when extended to include the oceanic data, and you probably have to make
a number of maps each as a different scale.
I remember in the past when talking about the GRASS 7 raster format it
was mentioned that perhaps Adaptive JPEG-like blocks could be used to
save the data instead of row-by-row. This is slightly a perversion of
that, but an interesting problem to think about.
Is there anyway to display a TIN in 3D with GRASS? In NVIZ, it just looks
like a Delaunay triangulation network unless there is something I'm not
doing.
Attached a v.delaunay trangulation from 3D Lidar points
as visualized in NVIZ.
On Wed, Apr 23, 2008 at 1:12 PM, Hamish <hamish_b@yahoo.com> wrote:
Markus:
> Attached a v.delaunay trangulation from 3D Lidar points
> as visualized in NVIZ.
It would be interesting to create a nviz screenshot comparing that with a
lidar -> v.in.ascii -> v.surf.rst surface placed below instead of the
plane.
Here are the Lidar data:
North = 5403850 m; South = 5403547 m
West = 512700 m; Est = 512835 m
UTM Zone 32U, EPSG code 32632 (Germany, Stuttgart, Vaihingen)
Available data: last pulses
Is there anyway to display a TIN in 3D with GRASS? In NVIZ, it just looks
like a Delaunay triangulation network unless there is something I'm not
doing.
Attached a v.delaunay trangulation from 3D Lidar points
as visualized in NVIZ.
Markus
__________________________________________
Michael Barton, Professor of Anthropology
Director of Graduate Studies
School of Human Evolution & Social Change
Center for Social Dynamics & Complexity
Arizona State University
