accuracy of raster maps

[Frank Warmerdam]

Jan Hartmann wrote:
> Hi Frank,
>
> I didn't get your projection solution to work on these unreferenced
> raster images. However, if I round all mapextents to whole numbers, the
> error almost disappears. At least, the data points are always within the
> same pixel, which is all I need.

Jan,

Ah, yes.  If everything does have valid assigned projections the warping
code won't work.  I would like to remove this restriction but it turns out
to be messy to do so.

> This probably works because in this case map units and pixel units are
> the same: one pixel is one map unit I'm not sure how this could be
> generalized for georeferenced maps with pixel sizes other than map
> units. I guess forced warping would be necessary in this more general case.

Right.

> I could imagine that forced warping would make sense only for scales
> where individual pixels are visible, e.g. when map extent is smaller
> than map size. Perhaps this could be a further refinement to the
> PROCESSING directive.

That might be possible.

Best regards,

--
---------------------------------------+--------------------------------------
I set the clouds in motion - turn up   | Frank Warmerdam, warmerdam at pobox.com
light and sound - activate the windows | http://pobox.com/~warmerdam
and watch the world go round - Rush    | Geospatial Programmer for Rent