On Tue, 2009-02-17 at 02:01 +0100, Uwe Stöhr wrote:
> Nikos Alexandris schrieb:
> 
> > I am interested in this statement...
> > Could you please extent a bit or give some pointers?
> 
> This is now a bit off-topic.

Once and a while it doesn't hurt :-)


> I maintain at work a scanning electron microscope (SEM). Due to technical 
> reasons the distance 
> between the pixels in x-direction is different from the y-direction. So you 
> usually take SEM images 
> with 512x512 pixels, but get a rectangular, non-quadratic image.

This is, more or less, also the case with images referenced in a
geographic coordinate system (think degrees, minutes, seconds). X and Y
are in reality different since latitude varies greatly when you measure
it near the equator against higher latitudes. 

The maps (images) look "distorted" but they are still, on-display,
rectangulars.

Anyhow, the scales (geo-maps vs. SEM's) have nothing to do :-)

I wonder if GRASS-GIS, which is a power-full raster engine, could be of
any use for SEM images? Why not referencing to a micro-scale reference
system images and play around? :D You can also play around with voxels
and 3D-stuff.

For the fun of the game I would like to try it out some day. I might
even post to folks over in GRASS-user mailing list about this (foolish
perhaps) idea.



> The image format is TIFF and the pixel distance is specified in the TIFF 
> header.

I understand here "distance" as "resolution". Something like an image in
which the x-dimension of a pixel is 1 metre and the y-dimension of the
pixel is set to be 1.5 metre. 
 

>  Unfortunately the 
> only program that checks for the pixel distance is Windows' built in image 
> viewer. In e.g. Adobe 
> Acrobat you have to specify the pixel distance manually to get a 
> non-distorted image.

So you mean you have to set manually the resolution of x-y if I get it
right.

Just for the records: under Linux you have for example imagemagick (a
"great", in number and in abilities, collection of image manipulation
tools).

I am pretty sure you could do neat stuff with your images.

> 
> regards Uwe

Kind regards and thank you for your explanation.
Nikos

Reply via email to