[gdal-dev] Polygonize via Python on Windows

Even Rouault even.rouault at mines-paris.org
Mon Apr 25 15:22:34 EDT 2011


Le lundi 25 avril 2011 21:17:48, Jay Jennings a écrit :

If you have VS2008, I tunk you need to upgrade your Python version to Python 
2.6. Because when building Python native modules, you must use a version of 
Visual Studio compatible with the version that was used to build the Python 
interpreter.

> Thanks Elijah, that thread does look promising.  But it involves using
> gdal_polygonize.py, which is giving me trouble (we haven't tried
> GDAL/Python options before).  Initial attempt at gdal_polygonize.py gives
> the error "ImportError: No module named gdal".  After researching the
> archives, I found some advice from Howard Butler in October 2007 to go to
> the 'swig' directory and do 'nmake -f makefile.vc python'.  Tried that,
> and it processed for awhile then quit with this error msg:
> 
> "error: Python was built with Visual Studio 2003;
> extensions must be built with a compiler than can generate compatible
> binaries. Visual Studio 2003 was not found on this system. If you have
> Cygwin installed, you can try compiling with MingW32, by passing "-c
> mingw32" to setup.py. NMAKE : fatal error U1077: 'C:\Python25\python.exe'
> : return code '0x1' Stop. "
> 
> It's true, I don't have VS 2003, but I do have VS 2008 and VS 2010.
> 
> Does anyone have setup advice on using gdal_polygonize.py on Windows 7
> 64-bit with GDAL 1.8.0 ?  I have a feeling I'm missing something simple. 
> We have GDAL 1.8.0 built from source, which works well, but we've never
> used the Python options.  Most of the GDAL/Python discussions in the
> archives seem to be about Linux or Apple OS.  Thanks for any guidance. PS
> I have Python 2.5 installed.
> --JJ
> 
> -----Original Message-----
> From: Elijah Robison [mailto:elijah at villagis.com]
> Sent: Friday, April 22, 2011 3:05 PM
> To: Jay Jennings
> Cc: gdal-dev at lists.osgeo.org
> Subject: Re: [gdal-dev] Polygonize from C# ?
> 
> Hi Jay, you might want to review the following thread (Get extent from
> rasters).  Based on a cursory read of his advice, I believe Trent offers
> a solution that might accommodate your task:
> http://lists.maptools.org/pipermail/fwtools/2010-October/001886.html
> 
> /Elijah
> 
> Jay Jennings wrote:
> > Correct, we just need the footprint of the "good" (non-NODATA) area,
> > which will often approximate a simple 4-point polygon with no holes.
> > 
> > -----Original Message-----
> > From: Elijah Robison [mailto:elijah at villagis.com]
> > Sent: Friday, April 22, 2011 2:07 PM
> > To: Jay Jennings
> > Cc: gdal-dev at lists.osgeo.org
> > Subject: Re: [gdal-dev] Polygonize from C# ?
> > 
> > Jay, do you basically want a footprint shape of the useful data in the
> > Sat images (say, without the collar), or is the problem more complex,
> > where perhaps you have a mosaic of several images and, like swiss
> > cheese, there are no-data holes within the image?
> > 
> > It sounds to me like you just want a footprint of the collarless area;
> > is this correct?
> > 
> > /Elijah
> > 
> > Jay Jennings wrote:
> >> I see, thanks.  But since I have satellite images, this part of the doc
> >> is alarming: "...For non-thematic raster data (such as satellite
> >> images) the result will essentially be one small polygon per pixel, and
> >> memory and output layer sizes will be substantial. The algorithm is
> >> primarily intended for relatively simple thematic imagery, masks, and
> >> classification results".
> >> 
> >> I just want one polygon enclosing the non-zero pixels (or alternatively,
> >> enclosing the zero pixels, then I could do a geometric difference).  Is
> >> there any way to trick the method into doing that, to maybe avoid the
> >> heavy resource needs ?
> >> 
> >> 
> >> -----Original Message-----
> >> From: Even Rouault [mailto:even.rouault at mines-paris.org]
> >> Sent: Friday, April 22, 2011 11:45 AM
> >> To: Jay Jennings
> >> Cc: gdal-dev at lists.osgeo.org
> >> Subject: Re: [gdal-dev] Polygonize from C# ?
> >> 
> >> Selon Jay Jennings <Jennings.Jay at geoeye.com>:
> >> 
> >> Generally, refering to the C/C++ doc is a good way of finding the info :
> >> 
> >> See http://gdal.org/gdal__alg_8h.html#3f522a9035d3512b5d414fb4752671b1
> >> 
> >>> Hi list,
> >>> Can anyone point to any guidance or (preferably) examples on using
> >>> GDAL.Polygonize() with C# bindings ?  I see what the interface is, but
> >>> there are some unexplained parameters, e.g. "string[] options".
> >>> 
> >>> My eventual goal is to open a raster and produce a WKT POLYGON string
> >>> enclosing the "good" (non-NoData) area of the raster.  Thanks.
> >>> 
> >>> .........................................................
> >>> Jay Jennings
> >>> jennings.jay at geoeye.com
> >> 
> >> _______________________________________________
> >> gdal-dev mailing list
> >> gdal-dev at lists.osgeo.org
> >> http://lists.osgeo.org/mailman/listinfo/gdal-dev
> 
> _______________________________________________
> gdal-dev mailing list
> gdal-dev at lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/gdal-dev


More information about the gdal-dev mailing list