[photo-3d] Algorithm Wanted

[Bruce Springsteen <bsspringsteen@xxxxxxxxx>]

A sidebar to the tutorial.
In his quiz explanation, George wrote:

>You can think of the stereoscopic deviations 
>as vectors which, in a properly recorded and mounted stereo
>pair, lie in the horizontal direction.  Stereo projection
>or anaglyph pairs give you a nice opportunity to study the
>orientation of these vetors.

I'm looking for the most efficient algorithm - a set of steps - for
locating and aligning those vectors in an un-registered pair of images. 
In other words, solve the following problem:

You have two large prints, unmounted left and right images of a stereo
pair, marked L and R.  There is no obvious indication in the pictures
(like a horizon) as to where horizontal should be in the stereopair.  The
edges of the prints are not reliable as a guide - say the prints are cut
round or ragged.  What then is the most efficient way to discover the
direction of deviations in the pair?  You may mark or damage the prints if
necessary to discover the answer - we assume that new clean prints may be
made, once the vectors are located.

I'm interested in purely geometric answers, as well as real-world
practical ones.  I have no "right" solution to offer, just my own ideas.

Bruce

__________________________________________________
Do You Yahoo!?
Yahoo! Mail - Free email you can access from anywhere!
http://mail.yahoo.com/