An efficient 2D image registration algorithm

Abstract

In this paper an efficient algorithm is presented for registering the images based on pixel similarity. The proposed algorithm is based on concept of calculating number of points in the reference image and same sized target image which satisfy the defined similarity criterion. The algorithm uses a counter to count the number of the similar intensity valued points (as per similarity criterion) in both images. The window corresponding to maximum value of counter indicates the best match. The performance of proposed algorithm is studied by registering a set of occluded and noisy images. It is observed through experimental results on different type of real world images that the proposed algorithm efficiently registers the images with very low computation time.