[Gdal-dev] gdalwarp reprojection not exact

lBert l1brtine at gmx.net
Fri May 16 12:59:42 EDT 2008


Hi,

I've been trying to reproject an orthophoto (1-m resolution) using gdalwarp
(gdal 1.5), but the results are not very satisfying (see below). I overlaid
the reprojected image with the input image and at the first look the results
seemed fine, but when I zoomed in, I realized that the reprojected image is
quite distorted. Since the distortion occurs in these blocky patterns
(blocks of ~30x30 pixels) I assume that this is the result of gdal using
some linear approximation. Now, I have tried changing several options, also
setting the -te switch to 0.0, but the results are always the same. Of
course when I use cubic convolution the effect is not as pronounced, but it
is still there. I played around with this the whole day and I can't find a
solution. ArcGIS and Erdas Imagine produce good results even with polynomial
approximation. 
Please help!

Here is an animated gif comparing the input and reprojected image:
http://www.nabble.com/file/p17267217/gdalwarp_input_output.gif
gdalwarp_input_output.gif  

Here is the command I used:
gdalwarp.exe -t_srs
PROJCS["unnamed",GEOGCS["NAD83",DATUM["North_American_Datum_1983",SPHEROID["GRS
1980",6378137,298.257222101,AUTHORITY["EPSG","7019"]],TOWGS84[0,0,0,0,0,0,0],AUTHORITY["EPSG","6269"]],PRIMEM["Greenwich",0,AUTHORITY["EPSG","8901"]],UNIT["degree",0.0174532925199433,AUTHORITY["EPSG","9108"]],AXIS["Lat",NORTH],AXIS["Long",EAST],AUTHORITY["EPSG","4269"]],PROJECTION["Lambert_Conformal_Conic_2SP"],PARAMETER["standard_parallel_1",43],PARAMETER["standard_parallel_2",45.5],PARAMETER["latitude_of_origin",41.75],PARAMETER["central_meridian",-120.5],PARAMETER["false_easting",1312335.958005249],PARAMETER["false_northing",0],UNIT["user-defined",0.3048]]
-et 0.0 -ot Byte -of hfa -tr 3.28 3.28 -te 478085.70 1363295.70 511509.30
1383735.30 naip_utm.img naip_lam.img

Thanks a bunch,
Bert

-- 
View this message in context: http://www.nabble.com/gdalwarp-reprojection-not-exact-tp17267217p17267217.html
Sent from the GDAL - Dev mailing list archive at Nabble.com.



More information about the gdal-dev mailing list