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

[JD Smith]

On Sat, 2003-12-27 at 20:48, nowozin at cs.tu-berlin.de wrote:
> Hi Pablo, hi list :)
> 
> 
> 
> On Sat, Dec 27, 2003 at 05:41:22PM +0100, Pablo d'Angelo wrote:
> 
> > [...]
> > when you create your rotationally invariant descriptor, by suming the
> > polar image, one looses all rotation information. Probably it would be
> > interesting to create a kind of "local" reference frame, from the center
> > to the strongest feature. This could then be used for corrospondence
> > analysis as well.
> 
> I am not sure if I understand completely what you mean with reference frame.
> 
> My idea is roughly this: As comparing all feature pixels with each other has
> quadratic complexity, I first build a short rough fingerprint (the rotation
> invariant one) of each feature pixel. Then, a number of closest matches by
> this fingerprint are searched for each feature pixel (default is 48 ones, so
> complexity is still quadratic in this step). The expensive comparison
> follows on just this first matches, so the complexity is now
> O(featurepixel_count). This second comparison finds the rotation required to
> create the minimum difference.
> 
> 
> > Have you read the "Recognizing Panoramas" Paper from Brown?
> > http://www.cs.ubc.ca/~mbrown/panorama/panorama.html
> 
> Ahhhhhh.... I wish I would have had that link earlier. I will read it today
> and let you know :) His results look very impressive, also his speed.
> 
> My guess is that the high speed comes from the second step (where I have
> still O(n*m) complexity), but I have to read his paper.

I corresponded with Mathew on the issue of speed.  It's not mentioned
very prominently, but the times he reports are for panoramas made with
very small images (like 640x480 or so), and not typical 3-6 megapixel
ones -- hence the speed.  He mentioned that his feature recognition is
linear in the number of pixels, but that, ideally, you'd perform feature
matching on small versions of the images to get approximate solutions,
and then employ Lucas-Kanade optimization on the full images for the
final super-accurate registration.  I can't seem to contact your site:
is it down?

JD