I'm trying to import GOES tif images that have no georeferencing. The
georef information is in a separate file which I finaly opened. It
contains the lat and lon for every cell in the GOES images. The
problem is, I just can't seem to get the lat/lon information into the
TIF files. Is there any way to do this? gdalwarp? gdal_translate?
i.points?
I'm trying to import GOES tif images that have no georeferencing. The
georef information is in a separate file which I finaly opened. It
contains the lat and lon for every cell in the GOES images. The
problem is, I just can't seem to get the lat/lon information into the
TIF files. Is there any way to do this? gdalwarp? gdal_translate?
i.points?
add GCPs with gdal_translate, then gdalwarp?
Maciek
--------------------
W polskim Internecie s? setki milion?w stron. My przekazujemy Tobie tylko najlepsze z nich! http://katalog.panoramainternetu.pl/
but my control points are in a huge text list and the switch -gcp in
gdal_translate does not accept that. So, I have to enter the corner
points by hand? Or is there a way to make gdal_translate read from a
text file (all 1200000 points)?
Thanks
Daniel
On 4/5/06, Maciek Sieczka <werchowyna@epf.pl> wrote:
> Hi again,
>
> I'm trying to import GOES tif images that have no georeferencing. The
> georef information is in a separate file which I finaly opened. It
> contains the lat and lon for every cell in the GOES images. The
> problem is, I just can't seem to get the lat/lon information into the
> TIF files. Is there any way to do this? gdalwarp? gdal_translate?
> i.points?
add GCPs with gdal_translate, then gdalwarp?
Maciek
--------------------
W polskim Internecie s? setki milion?w stron. My przekazujemy Tobie tylko najlepsze z nich! http://katalog.panoramainternetu.pl/
> > I'm trying to import GOES tif images that have no georeferencing. The
> > georef information is in a separate file which I finaly opened. It
> > contains the lat and lon for every cell in the GOES images. The
> > problem is, I just can't seem to get the lat/lon information into the
> > TIF files. Is there any way to do this? gdalwarp? gdal_translate?
> > i.points?
>
> add GCPs with gdal_translate, then gdalwarp?
but my control points are in a huge text list and the switch -gcp in
gdal_translate does not accept that. So, I have to enter the corner
points by hand? Or is there a way to make gdal_translate read from a
text file (all 1200000 points)?
I don't think that gdal_translate was designed to handle such large
numbers of control points.
You could import your data as a vector point map then use v.surf.rst.
You would probably need to match the coordinates with the cell values
manually (e.g. exporting the raster with r.out.ascii, importing the
coordinates along with the r.out.ascii output into a DBMS, then
joining the two tables).
Alternatively, you may be able to get adequate results by using a
small sample of the control points and gdalwarp's -tps switch.
What I'm trying to do is to convert the goes coordinates file into a
POINTS file for an imagery group and then use i.rectify.
Do you think that's doable? So far I have only got really bad results
but probably due to the fact that I'm a lousy programer and I'm having
a hard time understanding the way the coordinate file is stored...
Thanks
Daniel
On 4/5/06, Glynn Clements <glynn@gclements.plus.com> wrote:
Daniel Victoria wrote:
> > > I'm trying to import GOES tif images that have no georeferencing. The
> > > georef information is in a separate file which I finaly opened. It
> > > contains the lat and lon for every cell in the GOES images. The
> > > problem is, I just can't seem to get the lat/lon information into the
> > > TIF files. Is there any way to do this? gdalwarp? gdal_translate?
> > > i.points?
> >
> > add GCPs with gdal_translate, then gdalwarp?
>
> but my control points are in a huge text list and the switch -gcp in
> gdal_translate does not accept that. So, I have to enter the corner
> points by hand? Or is there a way to make gdal_translate read from a
> text file (all 1200000 points)?
I don't think that gdal_translate was designed to handle such large
numbers of control points.
You could import your data as a vector point map then use v.surf.rst.
You would probably need to match the coordinates with the cell values
manually (e.g. exporting the raster with r.out.ascii, importing the
coordinates along with the r.out.ascii output into a DBMS, then
joining the two tables).
Alternatively, you may be able to get adequate results by using a
small sample of the control points and gdalwarp's -tps switch.
I'm trying to import GOES tif images that have no georeferencing. The
georef information is in a separate file which I finaly opened. It
contains the lat and lon for every cell in the GOES images. The
problem is, I just can't seem to get the lat/lon information into the
TIF files. Is there any way to do this? gdalwarp? gdal_translate?
i.points?
Create a world fiel for each image, gdal_translate (I think) can then read
the world file to turn the tif into a geotif.
I'm trying to import GOES tif images that have no georeferencing. The
georef information is in a separate file which I finaly opened. It
contains the lat and lon for every cell in the GOES images. The
problem is, I just can't seem to get the lat/lon information into the
TIF files. Is there any way to do this? gdalwarp? gdal_translate?
i.points?
I see you used i.points to do it, but FYI if the georef points were
square relative to each other you could just make a "world file" for the
TIFF image defining the coordiantes at the four corners. GDAL will look
for and use this file if it exists. Alt. is to put that info into the
TIFF file directly with gdal_translate to make a GeoTIFF.
IIRC the -tps rectification of gdalwarp was created to use georef points
at the start and end of satellite scan lines, so maybe some joy adding
lots of -gcp with gdal_translate after all.
but my control points are in a huge text list and the switch -gcp in
gdal_translate does not accept that. So, I have to enter the corner
points by hand? Or is there a way to make gdal_translate read from a
text file (all 1200000 points)?
Use -gcp multiple times, one for each control point. But the shell
probably limits you to 4096 characters per call.
The only thing that I'm worried now is that the rms = 0.31 in
i.points. Is that too high? I'm attaching both the i.points screen
dump and the python script, in case someone needs it
That RMS is 2D over the whole area I belive, so judging from your
attached screen shot I don't think it is applicable (warping).