Image registration based on physical forces

Abstract

In this paper, we proposed a new method for image registration which the registration parameters are translation and rotation. This method is based on physical forces. A unique feature of the proposed method is that it assumes images like charged materials that attract each other. We can also assume one of the images like a hole and another one like a material that fill the hole. In both cases one of the images moves in the same direction as the applied force and the other one is still. It moves consequently until the resultant force is equal to zero. This approach does not estimate these parameters separately. They are estimated simultaneously leading to a better optimized set of registration parameters. We compute the order of this algorithm on 2D and 3D Images. We show that the approach is successful and yields to better results than MI and correlation-like methods.