This folder contains a number of matlab files.

dlt.m contains the main script that calls all of the other functions. It will
generate a smile and transform that smile according to a given projective
transformation H. Then, it will attempt to solve for the transformation H with the
DLT algorithm using a given number of point correspondences. Feel free to walk
through the code to get an idea of what its doing.

To get this to work, you have to implement two matlab files to implement the
algorithm we talked about in discussion. In get_Ai.m, you are given the point
correspondence, and you need to compute the 2x9 matrix Ai that represents the
constraint that the points yield. In compute_H.m, you are given the matrix A that is
formed from the given correspondences, and you need to compute the solution h. (You
may find the matlab function svd(A) useful.)

Once you get these parts working, you can see what happens if we add additional
correspondences (yielding an overdetermined system) or if the correspondences are
noisy and not perfect by changing the parameters in dlt.m.
