Hi, Function ProcessRecordHeaders in "formats/l1b/l1bdataset.cpp" subsamples the GCPs of the l1b image, but the points are unordered an not equidistant.
I attached to TCK 2403 a patch that modifies this behaviour. I tested the image attached to the ticket and seems to work ok(no error produced). I also test the patched driver with an image of Argentina and worked ok. I'd like to know if I misunderstood the intention of the code and the points in each line shouldn't be equidistant. Thanks Ramiro 2011/6/15 Andrew Brooks <a...@sat.dundee.ac.uk> > Hi > > From memory the l1b driver doesn't use enough points to get proper > geolocation because it sub-samples the 51 which are provided in the l1b > file. The best solution (that recommended by NOAA) is to use lagrange > interpolation across the 51 points. A lagrange algorithm has been provided > in the ticket you referenced but nobody has added it to gdal. I think at the > moment you could consider the l1b driver broken. > > Andrew > > > _______________________________________________ > gdal-dev mailing list > gdal-dev@lists.osgeo.org > http://lists.osgeo.org/mailman/listinfo/gdal-dev >
_______________________________________________ gdal-dev mailing list gdal-dev@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/gdal-dev
