[GRASS-user] sharpening an image based on multiple samples

GRASS GRASS user
1

Hi,

.. if there are any astronomers in the house ..

I have a series of images (all the same, slight jitter between each
one), I'd like to use a sharpening algorithum to combine them into one
very crisp image.

Anyone know the trick? I guess something like i.fusion.brovey, r.series,
or r.mapcalc magic.

Input is heavy-tripod DV of the night sky (very stable).

thanks,
Hamish

2

On Fri, Sep 08, 2006 at 12:44:42AM +1200, Hamish wrote:

Hi,

.. if there are any astronomers in the house ..

I have a series of images (all the same, slight jitter between each
one), I'd like to use a sharpening algorithum to combine them into one
very crisp image.

Anyone know the trick? I guess something like i.fusion.brovey, r.series,
or r.mapcalc magic.

You could try to smooth it with r.series, doing vertical averaging?
Or is it spatial jitter?

Markus

Input is heavy-tripod DV of the night sky (very stable).

thanks,
Hamish

3

> I have a series of images (all the same, slight jitter between each
> one), I'd like to use a sharpening algorithum to combine them into
> one very crisp image.

..

> Input is heavy-tripod DV of the night sky (very stable).

Markus:

You could try to smooth it with r.series, doing vertical averaging?
Or is it spatial jitter?

There is both slight spatial & temporal jitter between each image. I
wish to average out the temporal jitter without it getting blurred by
the slow spatial drift.

The drift seems to be both small & linear, so I might be able to correct
for that by manually shifting cells by the required amount, but I was
hoping for something that would automatically "lock on" to the features
in the image & save me the gererational losses introduced by an extra
step + clumsy human fingers.

Hamish