Brent Fraser wrote:
Frank,

The short story:

I've got some Single Look Complex EnviSat data I need to process. Will gdalwarp do a piecewise (triangulated) warp using GCPs or only(!) use them to calculate an image-wide polynomial?

Brent,

There are two choices from GCPs. Either polynomial (1st to 3rd order)
or thin plate splines.  Thin Plate Splines are nice in that the
transformation is exact at all GCPs, and it will give similar, but
more continuous results to piecewise triangulated warps which I
don't support.  But TPS is likely to be slow on a large dataset with
a lot of GCPs.

The longer story:
ESA's Best (http://earth.esa.int/services/best/) does an OK job extracting the amplitude from the data, but because it uses tiff files for intermediate storage, and their implementation is limited to 2gb tiffs, it's necessary to tile an input image. Best assigns a different affine transformation to each output GeoTiff tile, which causes a mis-alignment of the tiles when re-assembling them (using gdalwarp etc). Best does output an extensive metadata text file with each tile including a geolocation "array" of latitude and longitude. Unfortunately the "array" is not regular (eg, my test tile has values for lines 1, 811, 812, 1622, 1623, 2240; every 165 pixels except the last interval which is 195 pixels). If gdalwarp would use this semi-regular array to triangulate the pixel locations, that would be great, otherwise I'll have to use something else...

There is no mechanism to use geolocation arrays that are irregular in
this way, and even for regular geolocation arrays the existing code is
a bit questionable.

Best regards,
--
---------------------------------------+--------------------------------------
I set the clouds in motion - turn up   | Frank Warmerdam, [EMAIL PROTECTED]
light and sound - activate the windows | http://pobox.com/~warmerdam
and watch the world go round - Rush    | Geospatial Programmer for Rent

_______________________________________________
gdal-dev mailing list
gdal-dev@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/gdal-dev

Reply via email to