On 08/05/2011 03:57 AM, Etienne wrote:
Hi all,
I am trying to resample a raster with categorical data (land surface vegetation/cover) from a high resolution (500m) to a coarser resolution.
I would like to do something similar to what the GRASS operator r.resamp.stats does, using the modal category present in the higher resolution dataset to pick the value in the lower-resolution map.
It seems that my only option using gdal_warp is nearest neighbor, which will pick the value nearest to the coarser pixel center (regardless of the most present category within the region).
I would also like have a low-resolution map with the proportions of all land
types (from the high resolution map).
Any idea on how to accomplish what I want using gdal? I am not familiar with
GRASS, and would prefer to stick with GDAL if I can. However, I am open to any
suggestions.
Etienne,
there is a workaround for this: gdaladdo, the tool to add overviews
('pyramids') to a file, does support the mode operator. So what you
could do is add overviews to your file such that the overviews cover the
output resolution you want, and then you can resample with
gdal_translate (when no projection change is necessary, just give it
your desired output resolution, see the gdal_translate docs), or with
gdalwarp (and use nearest). They should use the overview level that is
closest to your output resolution for resampling.
Best,
Vincent.
Many Thanks
Etienne
_______________________________________________
gdal-dev mailing list
gdal-dev@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/gdal-dev
_______________________________________________
gdal-dev mailing list
gdal-dev@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/gdal-dev
