Global Mapper v25.0

Equirectangular projection

Trailfinder
Trailfinder Global Mapper UserTrusted User
edited February 2011 in Projection Questions
Hi!

I'm trying to load a map tile (probe.jpg) for which I have created a world file (probe.jgw).
The map's projection is equirectangular, map datum RGF-93. True scale latitude is 46.5°.
Loading this tile in GM results in wrong position (about 1000 m too far east, and 9 km too far north).
Same procedure done in TTQV succeeds, the resulting cal file (probe_jpg.cal.txt) then used in GM works fine.
What's going wrong here?

I'm using GM 11.02 and TTQV 4.0.133 PU

Regards
Helmut

Comments

  • global_mapper
    global_mapper Administrator
    edited February 2011
    Helmut,

    The problem is that the coordinates in the world file are not correct (or at least don't match the .cal file). If you delete the world file and name the .cal file something like just proj_jpg.cal, then the .cal file coordinates are used and the data is correctly placed.

    Let me know if I can be of further assistance.

    Thanks,

    Mike
    Global Mapper Support
    support@globalmapper.com
  • Trailfinder
    Trailfinder Global Mapper User Trusted User
    edited February 2011
    Hi Mike,

    but that world file is exactly the one I used in TTQV to calibrate the map.
    The original bbox parameters from the wms-c source are:
    BBOX=584704,5449728,585216,5450240


    I want to use the self produced world files directly in GM. TTQV I used just to validate the correctness of the world file!

    Helmut


    Helmut,

    The problem is that the coordinates in the world file are not correct (or at least don't match the .cal file). If you delete the world file and name the .cal file something like just proj_jpg.cal, then the .cal file coordinates are used and the data is correctly placed.

    Let me know if I can be of further assistance.

    Thanks,

    Mike
    Global Mapper Support
    support@globalmapper.com
  • global_mapper
    global_mapper Administrator
    edited February 2011
    Helmut,

    Are you sure that the WMS-C bounding box coordinates are using the Equirectangular projection with the RGF-93 datum and the same projection parameters? I'm guessing that maybe the coordinates are using a different datum or something, like perhaps the Google Maps datum.

    Thanks,

    Mike
    Global Mapper Support
    support@globalmapper.com
  • Trailfinder
    Trailfinder Global Mapper User Trusted User
    edited February 2011
    Mike,

    I'm sure. In the TTQV cal file you also see those projection and datum parameters:
    projparams = 10 = proj=eqc lon_0=0 x_0=0 y_0=0 lat_ts=46.5
    datum1 = 10 = RGF93# 6378137# 298.257222101# 0# 0# 0#


    Helmut
  • global_mapper
    global_mapper Administrator
    edited February 2011
    Helmut,


    I see that in the .cal file, but I'm not sure where you are getting the projected world file coordinates from. The .cal file lists the lat/lon coordinates of the corners, not the projected coordinates. When I manually project the top-left lat/lon coordinate from the .cal file to the Equirectangular projection specified in the .cal file I get the following coordinates:

    Easting: 583724.3
    Northing: 5441102.6

    These are quite a bit different than what you have in your world file.

    Thanks,

    Mike
    Global Mapper Support
    support@globalmapper.com
  • Trailfinder
    Trailfinder Global Mapper User Trusted User
    edited February 2011
    Mike,

    perhaps a sample transformation of an authority like French IGN can help here.
    Can you please check the values given by IGN in the first link (it's the first example). Do you get the same results? TTQV does.
    In the second the used projection parameters are defined.

    http://api.ign.fr/geoportail/api/doc/developpeur/wmsc.html
    http://spatialreference.org/ref/sr-org/6882/html/

    Helmut
  • global_mapper
    global_mapper Administrator
    edited February 2011
    Helmut,

    Ah the problem is due to a difference in how the Equirectangular projection is done between the USGS GCTPC library (which Global Mapper uses for the projection) and the PROJ.4 library which IGN uses. The USGS library uses the radius of the spheroid for the conversion, whereas PROJ.4 uses the semi-major axis of the ellipsoid. It seems to me that the USGS approach would be better absent a true ellipsoidal implementation of the projection, although really it's academic. It just causes differences depending on which method is used as there isn't really a standard (see http://www.remotesensing.org/geotiff/proj_list/equirectangular.html, which points this out and the confusion resulting from it). I'm not really sure a proper solution for this since changing to use what IGN/Proj.4 uses will break things for those applications using the USGS approach.

    Thanks,

    Mike
    Global Mapper Support
    support@globalmapper.com