[ptx] autopano: first attempt at semi-automatic panorama stitching

[Pablo d'Angelo]

On Fri, 02 Jan 2004, Bruno Postle wrote:

> On Thu 01-Jan-2004 at 11:50:53PM +0100, Pablo d'Angelo wrote:
> > 
> > has anyone tried the current autopano program?
> 
> Anonymous CVS is normally a few hours behind, so that might not be
> buildable yet [*].
> 
> I tried it on a couple of single-row hand-held panoramas, it seemed
> relatively fast.
> 
> The control points it created were nearly correct, but they were all
> slightly wrong.  They were near enough though - If there was a
> feature in hugin to select control points and redo the
> auto-alignment, they would have been fixed.

I'm just adding the "fine tune" all function.

> However, if you have a photo that contains 'features' and you know
> roughly the type of lens and fov, there is an attribute that is
> independent of roll, pitch and yaw that can be calculated precisely:
> The angular distance between any two features in a single image.
>
> Say you have an image with three features; there will be three of
> these angles that can be calculated - If another photo has three
> features with the same three angles, then the two photos are quite
> likely to actually correspond at those points.

Very interesting idea... You guys make it really hard for me to work on my
diploma thesis ;)

Using just 3 features will be very unreliable, because a single feature
might not be detected the same way in the other picture, or it's not on the
image at all. 
One would have to find a way to combine not all features, because that would
lead to an insane amount of angles for all features. Maybe only features
inside a local neighbourhood.

> Don't know if this is the theory behind any of the other approaches,
> but calculating angles and comparing numbers is simple and very
> fast.

Not that I know. Hmm implicitly it is done in the RANSAC step of browns
algorithm, where he checks if a set of features leads to a consistent transform
between the images.

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