hugin crash and ideas

[Pablo d'Angelo]

On Mon, 06 Oct 2003, Peter Suetterlin wrote:

> > Hmm, you could also just lower the threshold. I suspect that one would
> > try shift every time no match was found...
> 
> I don't really want to lower the threshold permanently.  But when a
> too-low value is met, I cannot see where the algorithm would have put
> it.  Most probably it is quite often still correct, especially towards
> the corners of the image where the correcaltion goes down due to
> geometrical distortions.

So a message box with "low correlation, set point anyway?" question might
be another solution?

ciao
  Pablo
--
http://wurm.wohnheim.uni-ulm.de/~redman/
Please use PGP