JM,
Looks to me that GDAL has clipped to the nearest pixel boundaries and
not to the window you have provided. This is probably a feature. Maybe
gdalwarp would do the kind of resampling you wish, but I'm not sure.
Regards,
Ari
On 06/03/2013 03:37 PM, JM wrote:
Hello List.
At first i wanna thanks for the great work you are doing with the gdal
library. It is a source of great tools and opportunities for spatial
issues which whom i involed in.
Among others i am using the commandline tool gdal_translate for
clipping raster files via the following example command:
gdal_translate -projwin 13.6666664938 53.2000002428
13.8333334784 53.1000017686 source.tif target.tif
The source tif is already georeferenced (EPSG:4326). I would now
assume that the boundingbox of the following GTiff is somethink like this:
Lower Left: 13.6666664938 53.1000017686
Upper Right: 13.8333334784 53.2000002428
Instead it is something like this (the full gdalinfo dialog you can
see in the attach to this mail):
Lower Left: 13.6666610463 53.1000244672
Upper Right: 13.8333191458 53.2000151873
So there is a small imprecision in the clipping result. My question is
now if this imprecision is normal (maybe a result of rounding
operations within the clipping process) or if i do something wrong (do
i miss something in my command).
I am using libgdal1 version 1.9.0-3.1ubunut1. Looking forward to
hearing from you.
Kind regards,
JM
_______________________________________________
gdal-dev mailing list
[email protected]
http://lists.osgeo.org/mailman/listinfo/gdal-dev
_______________________________________________
gdal-dev mailing list
[email protected]
http://lists.osgeo.org/mailman/listinfo/gdal-dev
