Evaluation of error metric for image registration of multiple images

Abstract

When one image is to be registered with multiple images of the same scene, one would like to know the accuracy of the registered image. In this paper we define the comparison of different error metrics based on the No-reference method (no true value) and reference method (true value) which evaluates the performances of image registration algorithms for sequences of images. This analysis will be a useful tool for quantifying the image registration algorithm and for tuning the parameters of the registration algorithm.