[Gdal-dev] need tools for TIF image reprojection

Fri Jul 14 09:32:22 EDT 2006

Hi Zhonghai;

Doing some quick calculations, the Latitude error is about 200 meters and the longitude error is about 100 meters. These errors look like the type and magnitude of errors you'd get if you didn't correctly convert from one datum to another.

So, make sure when you go from the Swiss National System to Geographic (lat, lon in Bessel 1841), you must then convert (transform) those lat, lons to ETRS89 lat, lons then project those lat, lons to UTM Zone 32N projection coord.s

Chuck Wivell

>>> "Zhonghai Wang" <zhonghaiw at gmail.com> 7/14/2006 8:12 AM >>>
Hi Petteri,

unfortunately not, the reprojected tif image has always the same
displacement. I think the problem occurs between the transformation from
Swiss National System (CH1903) to WGS84 system.

thanks you for your advice

Best Regards

zhonghai

On 7/14/06, Zhonghai Wang <zhonghaiw at gmail.com> wrote:
>
>  Hi Petteri,
>
> I hope the old version will work fine for the reprojection. I am going to
> try it now. Thanks for this important info.
>
> kind regards
>
> zhonghai
>
>
> On 7/14/06, Petteri Packalen <packalen at cs.joensuu.fi> wrote:
> >
> > Hi,
> >
> > http://iweb.gdal.org/dl/fwtools/ 
> >
> > and then download a6 version according your OS. I used Windows version.
> >
> > Best regards,
> > Petteri
> >
> > On Fri, 14 Jul 2006, Zhonghai Wang wrote:
> >
> > > Hi Petteri,
> > >
> > > I am using the latest version FWTools1.0.5, which version do you mean?
> > > FWTools a6? never heard.
> > >
> > > if possible, can you give me a link for the version you mentioned?
> > >
> > > thanks a lot
> > >
> > > zhonghai
> > >
> > > On 7/14/06, Petteri Packalen < packalen at cs.joensuu.fi> wrote:
> > >>
> > >> Hi,
> > >>
> > >> This just a guess: what version of GDAL are you using? I had similar
> > kind
> > >> of problem about one month ago before I left to summer holiday. I
> > >> downloaded the newest FWTools version and when I tried to transform
> > some
> > >> images from Finnish kkj to EUREF FIN coordinate system datum shift
> > was not
> > >> applied. EUREF FIN is basically ETRS89 / UTM 35N. When I revert back
> > to
> > >> FTTools alpha 6 everything worked propely. I don't remember what was
> > >> actual gdal versions in these FWTools releases. Please try FWTools a6
> > if
> > >> possible.
> > >>
> > >> I didn't remember to fill a bug report. I will look at it later when
> > I
> > >> come back to office.
> > >>
> > >> Best regards,
> > >> Petteri
> > >>
> > >> On Fri, 14 Jul 2006, Zhonghai Wang wrote:
> > >>
> > >> > Hi Andrew,
> > >> >
> > >> > you are right, in the most cases the GDAL utilities work well, and
> > I've
> > >> also
> > >> > use the GDAL utilities for a long time, especially the FWTools made
> > by
> > >> > Frank, we should all thank Frank for such a wonderful GIS toolkit.
> > >> >
> > >> > I've check again and again with the GDALWARP, but things are not
> > going
> > >> in
> > >> > the right way. I am trying to translate the TIF images from the
> > >> coordinate
> > >> > system CH1903 (Swiss National System) to ETRS89 / UTM ZONE 32N, and
> > >> there
> > >> > are always  some displacement with the base maps, I am sure that
> > the
> > >> base
> > >> > map and the original coordinate info for the TIF is correct. and
> > here is
> > >> an
> > >> > example:
> > >> >
> > >> > (the transforming process is : CH1903 -> WGS84 -> ETRS89 / UTM ZONE
> > 32N)
> > >> >
> > >> > the TIF image coordinate Info (*.tif): (Swiss National System)
> > >> > 0.500000
> > >> > 0.000000
> > >> > 0.000000
> > >> > -0.500000
> > >> > 598125.250000
> > >> > 199999.750000
> > >> >
> > >> > If I transform the coordinate to WGS84 with GDAL, the value is:
> > >> >
> > >> > 0.0000059839
> > >> > 0.0000000000
> > >> > 0.0000000000
> > >> > -0.0000059839
> > >> > 7.4149523128
> > >> > 46.9529994915
> > >> >
> > >> > but the lon/lat of the upper left corner should be:
> > >> >
> > >> > 7.41401039348263
> > >> > 46.9510762167473
> > >> >
> > >> > I would say that this is already big difference, because If I use
> > this
> > >> > lon/lat to get the UTM coordinate, this small difference will be
> > >> propagated,
> > >> > and will cause a displacement about hundreds meters.
> > >> >
> > >> > I am trying to find the source code, which does this coordinate
> > >> > transformation, so that I can check if the transformation
> > parameters are
> > >> > correct, but unfortunately I can not find the source for this.
> > >> >
> > >> >
> > >> > If someone has done something similar, please point me out how to
> > fix
> > >> this
> > >> > problem, especially the GIS people in Switzerland. I would really
> > >> appreciate
> > >> > it.
> > >> >
> > >> > zhonghai
> > >> >
> > >> > On 7/14/06, Andrew Williams < awilliams at rapidmap.com.au> wrote:
> > >> >>
> > >> >> zhonghai,
> > >> >> I've litterally translated hundereds of TIF's from all over the
> > world
> > >> from
> > >> >> multiple projections to multiple projections. I have also had
> > >> "failures".
> > >> >> However, in every case, it was either my fault in specifying the
> > -s_srs
> > >> or
> > >> >> -t_srs. In some cases the fault lay in the original image. The
> > TIFF
> > >> itself
> > >> >> was supplied with incorrect georefernceing information and so it
> > was
> > >> never
> > >> >> going to work.
> > >> >>
> > >> >> Have you got a dump from GDALINFO that we can see and the GDALwarp
> > >> command
> > >> >> line you've used?
> > >> >>
> > >> >> regards
> > >> >> Andrew
> > >> >>
> > >> >>        -----Original Message-----
> > >> >>        From: Zhonghai Wang [mailto: zhonghaiw at gmail.com] 
> > >> >>        Sent: Fri 7/14/2006 12:14 AM
> > >> >>        To: gdal-dev at lists.maptools.org 
> > >> >>        Cc:
> > >> >>        Subject: [Gdal-dev] need tools for TIF image reprojection
> > >> >>
> > >> >>
> > >> >>        Hi guys,
> > >> >>
> > >> >>        I need a tool to reproject my TIF images, either free or
> > >> non-free
> > >> >> software.
> > >> >>        I've tried with the GDAL utilities, sure I can do the
> > >> reprojection,
> > >> >> but the coordinates seems not correct, there are serveral hundert
> > >> meters
> > >> >> displacement according to the my base maps.
> > >> >>
> > >> >>        thanks for your info.
> > >> >>
> > >> >>        zhonghai
> > >> >>
> > >> >>
> > >> >>
> > >> >>
> > >> >
> > >>
> > >
> >
>
>