[gdal-dev] PDF gets skewed and shifts in geo location raster GeoTIFF

Even Rouault even.rouault at spatialys.com
Wed Sep 23 04:41:58 PDT 2015


Le mercredi 23 septembre 2015 12:54:21, Gane R a écrit :
> When I convert the above mentioned PDF to a GeoTIFF in EPSG:3857 I get the
> raster bounds shifted down from the actual raster location, where as when I
> convert the PDF to GeoTIFF in EPSG:4326 then to EPSG:3857 using gdalwarp
> and I get the raster  bounds to the actual location.  any reasons for this
> shift ?

Please let the list CC'ed

Could you provide data + command line that demontrates this ?

> 
> On Wed, Sep 23, 2015 at 1:42 PM, Even Rouault <even.rouault at spatialys.com>
> 
> wrote:
> > Gane,
> > 
> > I don't really understand what you mean. Is there an error in the PDF
> > georeferencing (I thought we had addressed that a few months ago), or are
> > you
> > surprised that gdalwarp transforms the image to be north-up ? If it is
> > the later, then it is an intended behaviour of gdalwarp.
> > 
> > Best regards,
> > 
> > Even
> > 
> > > There are cases for handling 90 -90 rotation in pdfdataset.cpp
> > > I am using poppler with pdfdataset.cpp as in Revision: 28978 of trunk
> > > 
> > > PDF gets skewed and shifts in geo location raster GeoTIFF, Is there any
> > > issue in computing the geotransform ?
> > > 
> > > GeoTransform values are for an pdf at EPSG:4326
> > > 10776
> > > 6682
> > > -87.844995622259475
> > > 0.000001374506717
> > > 0.000000429963146
> > > 42.847234964542217
> > > 0.000000580102585
> > > -0.000001012112732
> > > 
> > > Corresponding GeoTransform for GeoTiff in EPSG:3857 after gdalwarp
> > > 12529
> > > 10422
> > > GeoTransform values
> > > -9779291.683092894000000
> > > 0.166040058992803
> > > 0.000000000000000
> > > 5290478.766616309100000
> > > 0.000000000000000
> > > -0.166040058992803
> > 
> > --
> > Spatialys - Geospatial professional services
> > http://www.spatialys.com

-- 
Spatialys - Geospatial professional services
http://www.spatialys.com


More information about the gdal-dev mailing list