First of all sorry if the title is somewhat opaque, the problem I am trying to solve is already hard to explain properly in my first language.
So, let's consider we have a plane, rectangle target in a three dimensional space. From two different points of views (the images of two cameras for example), we have the coordinates of each corner of that rectangle.
Knowing these coordinates, how could we translate the coordinates of any point on one of the images to the ones it would have on the other image ?
I made a (very) simple drawing that might help understanding what I'm looking for: http://i.imgur.com/x3QgZ.jpg
In this situation the first observer is right in front of the target, the other one is slightly shifted to the right. We know all coordinates but the ones of the red dot on the second point of view, which we are looking for.