Non-rigid Image Registration

Image registration, in particular medical image registration, has been subject to extensive studies in the past years. Itsversatile and important applications have attracted researchers from various branches, including numerical analysts. Thegoal of image registration is to find a transformation that aligns one image to another. To be successful, the registrationscheme has to be adapted to the problem under consideration. Depending on the application at hand, it is often desirableto constrain the wanted deformation. The idea is to incorporate higher level information about the expected deformation.In this note, we examine some intuitive ways of adding user knowledge. Moreover we show, that all of these may bephrased in a variational setting allowing for high performance numerics. In addition, we present various instructiveexamples.