[gdal-dev] Georeference without rotation of output image
Ivan Castellucci Vidal
ivanc.vidal at gmail.com
Fri Aug 14 10:29:53 PDT 2015
On Fri, Aug 14, 2015 at 2:23 PM, Even Rouault <even.rouault at spatialys.com>
wrote:
> Le vendredi 14 août 2015 19:15:44, Ivan Castellucci Vidal a écrit :
> > Thanks Even!
> >
> > That's exactly what I was looking for and I just have one question about
> > it. GDAL documentation states: "Due to imprecision in the calculations
> the
> > fit algorithm will often return non-zero rotational coefficients even if
> > given perfectly non-rotated inputs"
> >
> > Is there a way to use a more precise algorithm?
>
> The algorithm is mostly a least square regression. There are also a few
> heuristics to force the rotational coefficients to 0 in a few situations,
> like
> when there are 4 GCPs that form a rectangle. But since you mentionned that
> your images would have rotation, that shouldn't be much of a concern I
> think.
> Anyway in cases where the theory should be perfect 0, the values got in
> practice should be sufficiently close to zero for most practical purposes.
>
>
Thanks for your help, Even
Your help was really valuable
>
> > It's not a requirement by
> > now since this program is in development but at some point I'll need to
> > count with the maximum possible precision to make it work in a production
> > release.
> >
> >
> > On Fri, Aug 14, 2015 at 1:22 PM, Even Rouault <
> even.rouault at spatialys.com>
> >
> > wrote:
> > > > Instead of warping the image, you could compute a geotransformation
> > >
> > > matrix
> > >
> > > > that would include rotational terms with GDALGCPsToGeoTransform()
> > > >
> > > > http://www.gdal.org/gdal_8h.html#ae44b32f08d61072dddbdd30ac8ca962b
> > > >
> > > > and set it to your geotiff with GDALSetGCPs()
> > > >
> > > > http://www.gdal.org/gdal_8h.html#acd47aa85ca5751fbe9ea09c1a2341ef8
> > >
> > > Arg, sorry I meant with GDALSetGeoTransform() of course :
> > > http://www.gdal.org/gdal_8h.html#ae93448112c1a7e69f2764c1aa3c6c8b5()
> > >
> > > GDALSetGCPs() is to set the "raw" GCPs.
> > >
> > > --
> > > Spatialys - Geospatial professional services
> > > http://www.spatialys.com
> > > _______________________________________________
> > > gdal-dev mailing list
> > > gdal-dev at lists.osgeo.org
> > > http://lists.osgeo.org/mailman/listinfo/gdal-dev
>
> --
> Spatialys - Geospatial professional services
> http://www.spatialys.com
>
--
Iván Castellucci Vidal
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/gdal-dev/attachments/20150814/9131d0ca/attachment.html>
More information about the gdal-dev
mailing list