[Gdal-dev] convert HDF ASTER to geotiff

Jon Saints saintsjd at yahoo.com
Mon May 2 10:53:05 EDT 2005


Would you mind providing the list with an example of
your gdal_warp command? 

I have been doing the following:
1) use gdal_translate -sds to extract various ASTER
subsets into non-georeferenced tifs
2) use gdal_merge to merge the first three subsets
into a non-georeferenced multi-band tiff.

*** here I have not been able to use gdal_warp because
the image has no "original coord sys" associated with
it. so I use gdal_translate to apply lat/long coordsys
*** 

3) finally apply lat/long coord sys using UPPERLEFT
LOWERRIGHT values provided by gdalinfo on the original
HDF.

Does this look ok? could you provide an example of the
gdal_warp command?

Thanks
Jon


--- "vincent at ecovla.nl" <vincent at ecovla.nl> wrote:
> Jon Saints wrote:
> 
> > Ok, I have narrowed the problem.
> >
> > Yes, I agree that ASTER's come with ground control
> points, however,
> > when using gdal_translate it seems that the
> utility doesnt apply the
> > ground points correctly.
> >
> > My observation is this: In the raw aster image the
> corners (ul and
> > lr) of the raw image data are not the "corners" of
> the data. Along
> > both edges of the actual data there are no data
> values hold space as
> > padding. The actual image data is a slanted
> trazezoid padded on both
> > sides with no data values within the rectangle
> that is the raw aster
> > image. gdal_translate can apply the GCP points to
> the raw image, but
> > does so incorrectly. It applies the ul and lr
> values to the no data
> > padding of the raw image, rather than applying the
> point to the
> > actual corner where the actual data begins.
> >
> > I hope this makes some sense. I found it difficult
> to explain without
> > drawing pictures.
> >
> > Is there any workaround for this?
> >
> > Thanks Jon
> 
> 
> Jon,
> 
> Are you sure that for your ASTER images, the GCP's
> don't apply to the
> nodata areas too? In the ASTER images that I have
> used so far, the
> opposite is true. GCP's range from (0.5, 0.5) to the
> full pixel extent
> of the ASTER image (that is: including the nodata
> areas) , i.e. the last
> GCP maps pixel coordinate (2490.5,2100.5). This of
> course would never be
> so if the nodata parts were not covered by the
> GCP's. Moreover, the
> georeferencing that I've done using gdalwarp (note:
> NOT gdal_translate)
> looks very accurate; I'm sure I would have noticed
> it all my ASTER
> images were rotated and slanted.
> 
> You can check the values of the GCP's using gdalinfo
> on one of the
> subdatasets of your ASTER HDF files.
> 
> just my 2c.
> 
> regards,
> Vincent.
> 
> >
> >
> > --- Kenneth Melero <kmelero at sanz.com> wrote:
> >
> >> Andrey,
> >>
> >> Can you provide an example of this?
> >>
> >> Thanks! -Ken
> >>
> >> -----Original Message----- From:
> gdal-dev-bounces at xserve.flids.com
> >> [mailto:gdal-dev-bounces at xserve.flids.com] On
> Behalf Of Andrey
> >> Kiselev Sent: Saturday, April 30, 2005 2:42 AM
> To:
> >> gdal-dev at remotesensing.org Subject: Re:
> [Gdal-dev] convert HDF
> >> ASTER to geotiff
> >>
> >>
> >> On Fri, Apr 29, 2005 at 01:18:15PM -0700, Jon
> Saints wrote:
> >>
> >>> Hi, I am new to GDAL. I have had some success
> >>
> >> converting hdf to tif
> >>
> >>> using gdal_translate.
> >>>
> >>> i have not been able to apply a coorsys to the
> new
> >>
> >> tif image. has
> >>
> >>> anyone any expirience with this sort of thing?
> >>
> >> ASTER scenes are georeferenced with GCP points,
> that means you
> >> should use gdalwarp to transform the scene in
> desirable projection
> >> and extent. And you definitely need the latest
> CVS snapshot of GDAL
> >> sources where the proper functionality has been
> implemented.
> >>
> >> Regards, Andrey
> >>
> >> -- Andrey V. Kiselev Home phone: +7 812 5970603
> ICQ# 26871517
> >> _______________________________________________
> Gdal-dev mailing
> >> list Gdal-dev at xserve.flids.com
> >> http://xserve.flids.com/mailman/listinfo/gdal-dev
> >>
> >> _______________________________________________
> Gdal-dev mailing
> >> list Gdal-dev at xserve.flids.com
> >> http://xserve.flids.com/mailman/listinfo/gdal-dev
> >>
> > _______________________________________________
> Gdal-dev mailing list
> > Gdal-dev at xserve.flids.com
> > http://xserve.flids.com/mailman/listinfo/gdal-dev
> 
> 
> 



More information about the Gdal-dev mailing list