Hi everyvone,
I have a problem with gdaladdo while preprocessing raster images. For
the output of raster images in a Mapserver application I use gdaladdo to
build overviews of an image that represents Germany (gdaladdo -r
average test.tif 2 4 8 16 32 64 128 256 512 1024). Typically just the
first 4 levels are used in the application. To get better output results
from Mapserver I used the processing method "resample=average". But the
problem for high quality output in the Mapserver application seems to be
the great difference between the levels of the overviews. So I tried to
build overviews with gdaladdo -r average test.tif 2 3 4 5 6 7 ....
20. But now I have the problem that at the same zoomlevels in mapserver
and in the pyramid (e.g. image at 500x500 pixel) the quality of the
image is worse than in former case. Maybe gdaladdo does compute the next
level not from the original image but from the previous level in the
pyramid? Is there a possibility to build the each overview from
original image to get the best qualty in downsampled images?
thanks in advance
Holger Schiebold
_______________________________________________
mapserver-users mailing list
[email protected]
http://lists.osgeo.org/mailman/listinfo/mapserver-users
