[Gvsig_english] Raster georeferencing

Benjamin Ducke benjamin.ducke at oxfordarch.co.uk
Tue Oct 27 12:35:53 CET 2009


Wow, that's interesting to know! I was quite confused by this
myself. The new georeferencing tool is wonderful, but my
enjoyment of it is a little spoilt by those inconsistencies
in the UI that Ernesto described.

Seeing that the georeferencer is such an essential tool
for daily work, would it be possible have his suggestions
implemented in 1.9? They don't sound too complicated but
would make a lot of sense.

In addition, I would support getting rid of the selection
of the raster to be georeferenced (Ernesto's point #2)
completely. It is much simpler to have the currently
active raster layer selected automatically. Also, this
would resolve situations where no layers have yet been
added to a view and the Raster Tools are not available.
With the current design, users have to add a "dummy" layer
to the View to get the Raster Tools, then specify the
raster to georeference separately. Better workflow would
be: add raster to view -> start georeferencer -> raster
gets selected automatically -> done.

Best,

Ben

----- Original Message -----
From: "ernesto sferlazza" <e.sferlazza at gmail.com>
To: "gvsig internacional" <gvsig_internacional at listserv.gva.es>
Sent: Tuesday, October 27, 2009 11:08:43 AM GMT +01:00 Amsterdam / Berlin / Bern / Rome / Stockholm / Vienna
Subject: Re: [Gvsig_english] Raster georeferencing

Thanks to Juan Lucas Dominguez Rubio, I have solved the "mistery" (in
my mind) regarding the difference between "affine" and "polinomial 1st
order"  transformation.
The transformation is mathematically the same thing, but in the first
case ("affine") the transformation is performed "on the fly" by the
GIS sw that interprets the .wld file (containing skew axes parameters
not null) created with the georeferencing tool, while in the latter
case ("polinomial 1st order") the original raster is resampled in a
new georeferenced raster.
I have some little remarks regarding the procedure.
1) If  "affine transformation" option is selected there is not any
output raster to be created, so the output file input box should be
disabled;
2) the selection of the raster to be georeferenced in the window is
redundant, since the raster is the same selected in the "generic
ToolBar Panel" of the view;
3) the default size of output pixel should be proposed after the input
of Ground Control Points and not fixed antecedently;
4) once starded the "affine" transformation "test" it should be useful
(expecially in the case of georeferencing without target georeferenced
cartography) a message window informing that the .wld file has been
correctly created.

Best regards

Previous message:
> De: gvsig_internacional-bounces at listserv.gva.es en nombre de ernesto sferlazza
> Enviado el: mi? 14/10/2009 9:53
> Para: gvsig_internacional at listserv.gva.es
> Asunto: [Gvsig_english] Raster georeferencing
>
> Hello list and developers
> The algorithms at the moment available for raster georeferencing are
> "affine" and polinomial (1st, 2nd and 3rd order). The question I
> retrieve is: aren't "affine" and "polinomial 1st order" the same
> transformation?
> Anyway, nothing happens when I select "affine" transformation and try
> to start the process.
> [...]
> Ernesto Sferlazza

-- 
ing. Ernesto Sferlazza
responsabile nodo provinciale di Agrigento del SITR Sicilia
resp. gruppo Sistema Informativo Territoriale
Provincia regionale di Agrigento
sede istituzionale: piazza Aldo Moro, 1 - 92100 AGRIGENTO
sede operativa: via Acrone, 27 - 92100 AGRIGENTO
cell. 3392718463 (personale)
       3316874296 (aziendale)
tel. uff 0922593752
_______________________________________________
Gvsig_internacional mailing list
Gvsig_internacional at listserv.gva.es
http://listserv.gva.es/cgi-bin/mailman/listinfo/gvsig_internacional



------
Files attached to this email may be in ISO 26300 format (OASIS Open Document Format). If you have difficulty opening them, please visit http://iso26300.info for more information.



More information about the Gvsig_internacional mailing list