Thomas, I would start by creating a colour table with each group containing the same shade of grey. This will ease the polygonization by avoiding duplicates. There will be three or four (RGB/RGBA) bands from the colour table but you can just consider one band during polygonization.
On Tue, May 4, 2010 at 4:52 PM, Thomas Unternaehrer < thomas.unternaeh...@gmail.com> wrote: > Hello, > > I try to get a dissolved shapefile out of a large raster file (Erdas > img) using Python. The raster file contains values from lets say 1 to > 100. What I need is a shapefile that contains just five groups where > 1-20 get the value 1, 21-40 the value 2 and so on. > > What I've done so far are the following steps: > - split the raster file into several pieces using gdal_translate > - run gdal.Polygonize on each part to get shapefiles (dissolved based > on the raster values) > - manipulate the resulting dbf-files to get my described groups (gives > me a new field: fieldGrouped) > > Now I need to dissolve the resulting shapefiles based on the new field > fieldGrouped to get smaller shapefiles. > > How can I do this using gdal/ogr or any other library? Can I > manipulate the original values in my raster file? GRASS has this > functionality but if I'm right, I need to start GRASS and import the > shapefiles to it (it should be easy to run it on a different machine). > Nice would be a function like gdal.Polygonize. I'm more or less > completely new to GIS at all. > My main goal is to produce tiles for Google Maps. > > Any hint or help would be appreciated > > Thomas > _______________________________________________ > gdal-dev mailing list > gdal-dev@lists.osgeo.org > http://lists.osgeo.org/mailman/listinfo/gdal-dev > -- Best regards, Chaitanya kumar CH. /tʃaɪθənjə/ /kʊmɑr/ +91-9848167848 17.2416N 80.1426E
_______________________________________________ gdal-dev mailing list gdal-dev@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/gdal-dev