Hello Jonny,
I assume that this is not possible. Are the holes in your
elevation data big?
I would suggest that you interpolate the missing data. This can be
done tools from gdal-library:
http://www.gdal.org/
Ready compiled packages are available for Windows from OSGeo4W
http://trac.osgeo.org/osgeo4w/
(OSGeo4W Installer)
To fill the holes you need to do:
1. copy the elevation matrix to a separate text file ie. dem.txt
2. add a header to the text file that it looks like:
ncols 366
nrows 366
xllcorner 0.000000000000
yllcorner -366.000000000000
cellsize 1.000000000000
471.2 473.3 482.9 489.5 492.3 ...
447.7 449.8 453.8 456.4 457.2 ...
...
you have to adapt the ncols, nrows and yllcorner values to fit to
your matrix size (in my case 366 by 366 values)
3. convert and mark nodata values with gdal_translate. "0" is the
value for nodata
in OSGeo4W Shell:
> gdal_translate -a_nodata 0 dem.txt demnodata.tif
4. fill the nodata regions by
in OSGeo4W Shell:
gdal_fillnodata.py demnodata.tif filled.tif
5. convert back to ascii matrix
in OSGeo4W Shell:
gdal_translate -of AAIGrid filled.tif filled.txt
Remark: the produced *.tif files can not be displayed in every
image viewing software as it contains no the usual RGB data but
instead integer or float valus. One nice software to display (and
even analyze) is ImageJ:
http://imagej.nih.gov/ij/
Best regards
Wolfgang
Am 03.07.2013 21:46, schrieb Jonny Prouty: