[GRASS-user] best dem resolution for orthorectification

Maning Sambale emmanuel.sambale at gmail.com
Sun Jan 6 11:04:25 EST 2008


Hi,

I'm trying to interpolate a DEM to match my imagery.  The images are
approximately 60 cm resolution of about 829,490,277 cells!

GRASS 6.2.2cvs (phil_geog):~ > g.region rast=infanta_qb.rgb -pm
projection: 3 (Latitude-Longitude)
zone:       0
datum:      wgs84
ellipsoid:  a=6378137 es=0.006694379990141316
north:      14:47:20.445994N
south:      14:38:29.85632N
west:       121:33:58.254452E
east:       121:44:02.028909E
nsres:      0.60401006
ewres:      0.58788111
rows:       26999
cols:       30723
cells:      829490277


My input dem points contain 20800 points.  I tried interpolation in
smaller region and it seems to work OK.

This time trying at full region setting, I am getting segmentation fault
using v.surf.rst.

v.surf.rst input=dem_pt_fin layer=0 dmax=0.000014 dmin=0.000003
elev=dem_surf.rst maskmap=MASK zmult=1.0 tension=40. segmax=40
npmin=300 

A few questions:
1. Is the segfaults caused by my limited hardware (1 GB ram and 2 GB
swap)?  Does v.surf.rst temporarily writes to hardrive? My PC reports a
limited hardisk space while running v.surf.rst (drive freespace 10 GB).
Probable solution: do v.surf.rst with smaller regions the r.patch

2. Are there size limits to the number of vector points and output
raster?

http://www.dinama.gub.uy/grass/gdp/html_grass63/r.in.xyz.html
reports:
"Interpolation into a DEM
The vector engine's topographic abilities introduce a finite memory
overhead per vector point which will typically limit a vector map to
approximately 3 million points (~ 1750^2 cells)"

3. Would bigger DEM pixel resolution say 10-30 meters be OK for
orthorectification?

cheers,

maning





More information about the grass-user mailing list