An algorithm for multistage projective transformation adjustment for image superimposition

We propose a multistage algorithm for constructing a projective transformation based on arbitrary sets of keypoint pairs, with the sets containing more than four keypoint pairs. In the course of calculation of a homography matrix, unsuccessful keypoint pairs are detected and rejected. Experimental results on image superimposition using projective transformations for homogeneous and heterogeneous images are presented. Methods for assessing the quality of image superimposition, which include the local and integral estimates of the superimposition quality, are proposed.