No subject

for transforming image coordinates to projected coordinates) is a property
of the raster image, not of the projection itself. (Am I right?)
You can generate a simple six line text world file containing the A-F
coefficients using the coefficients:

A
D
-B
-E
C
F


You have to check the sings and probably the C and F values has to be
modified because in the world file the image coordinates are counted from
top left corner, what is probably not the case of CoordSys.

Best Regards,
Gabor


> Hi Mikael,
>
> Thanks very much for the answer.
> At least the question of whether or not the Affine process is
> supported has been resolved. :)
>
> Cheers,
> James
>
> On Tue, Jan 29, 2013 at 5:02 PM, Mikael Rittri
> <Mikael.Rittri at carmenta.com> wrote:
>> Hello James.
>>
>> Just a partial answer, I am afraid:
>>
>>
>>
>> I don’t think you can add an affine post-processing to a Proj.4 CRS
>> definition.
>>
>>
>>
>> For your example, the interpretation of the MapInfo CoordSys clause
>> should
>> be
>>
>>
>>
>> “CoordSys Earth Projection”, // Just a name (probably the wrong one in
>> this
>> case)
>>
>> 8,                   // The projection type, in this case Transverse
>> Mercator
>>
>> 33,                 // The geodetic datum, in this case GRS80 (treat it
>> like
>> WGS84)
>>
>> 7,                   // length unit, in this case meter
>>
>> 117,                // central longitude
>>
>> 0,                   // central latitude
>>
>> 0.9996,           // scale factor
>>
>> 500000,          // false easting
>>
>> 10000000,       // false northing
>>
>> 7,                   // length unit for affine transformation, also
>> meters,
>>
>> 0.890953, -0.455062, 2903977.24, 0.455063, 0.890954, -6919253.68,
>>  //
>> affine parameters A, B, C, D, E, F
>>
>> -100000, -100000, 100000, 100000               // bounds
>>
>>
>>
>> I think the basic CRS in this example can be expressed in Proj.4 as
>>
>>
>>
>> +init=epsg:28350
>>
>>
>>
>> which gives GDA94 / MGA zone 50, but I think you must implement the
>> affine
>> transformation yourself.
>>
>>
>>
>> I got the information from
>> http://www.sgsi.com/dnld/MapInfoPro_UserGuide_Current.pdf
>>
>> but I didn’t understand the meaning of the affine parameters.
>>
>>
>>
>> And I think the 8 ought to have been 3008, maybe…
>>
>>
>>
>> Best regards,
>>
>>
>>
>> Mikael Rittri
>>
>> Carmenta
>>
>> Sweden
>>
>> http://www.carmenta.com
> _______________________________________________
> Proj mailing list
> Proj at lists.maptools.org
> http://lists.maptools.org/mailman/listinfo/proj
>