Armin, This can be done in either OGR or PostGIS.
For OGR, refer to the OGRGeometry class reference, especially the ConvexHull(), Union() and UnionCascaded() functions. For PostGIS, ST_Boundary() and ST_Union(). On Wed, Mar 16, 2011 at 6:23 AM, Armin Burger <[email protected]> wrote: > Marius, > > thanks for the suggestion. I don't know if shapely supports re-projection > of geometries or has a simplify algorithm, which I will both need. There > could be some other functions of GDAL/OGR that I might need in the future as > well. So in principal I'd prefer to stick to GDAL. > > armin > > > On 16/03/2011 01:35, Marius Jigmond wrote: > >> Armin, >> >> Not sure if this would be straightforward in GDAL but you might want to >> consider a combination of GDAL + Shapely >> (http://gispython.org/shapely/docs/1.2/manual.html#cascading-unions). >> What you're trying to do is spatial analysis not directly implemented in >> GDAL. Shapely requires that you're working with Cartesian coordinates. >> The upside is that it has WKT/WKB support for direct loading into >> PostGIS. >> >> -marius >> >> On Wed, 2011-03-16 at 01:08 +0100, Armin Burger wrote: >> >>> Hi all >>> >>> I would like to catalogue shapefiles scattered over lots of directories >>> of the file system and store retrievable information of the shapefiles >>> in a PostGIS layer. Extracting parameters like extent, projection, >>> fields, etc works very fine with GDAL's Python bindings. >>> >>> But I would also like to store a sort of "footprint" of the whole >>> shapefile as a geometry object since the extent is a bit coarse >>> geographic representation of the shapefile. >>> >>> So far I have no better idea than eg. for polygon shapefiles looping >>> through all features, applying a Union function on them. And at the end >>> trying to use the Simplify method on the resulting polygon that will be >>> used as the footprint. >>> >>> This is for sure not very efficient for larger shapefiles with lots of >>> records. And for line and point shapefiles I still don't have a clue how >>> their records could be represented by an enclosing polygon (maybe the >>> Boundary functions does something like this...). >>> >>> Any ideas how this footprint generation could be achieved in a feasible >>> way using GDAL/OGR Python? >>> >>> Cheers, Armin >>> _______________________________________________ >>> 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 > -- Best regards, Chaitanya kumar CH. /tʃaɪθənjə/ /kʊmɑr/ +91-9494447584 17.2416N 80.1426E
_______________________________________________ gdal-dev mailing list [email protected] http://lists.osgeo.org/mailman/listinfo/gdal-dev
