Geometrical Regularization of Displacement Fields with Application to Biological Image Registration

摘要

This article tackles the registration of 2-D biological images (histological sections, autoradiographs, cryosections, etc.). The large variety of registration applications - 3D volume reconstruction, cross-dye histology gene mapping, etc. - induce an equally diverse set of requirements in terms of accuracy and robustness. In turn, these directly translate into regularization constraints on the deformation model, which should ideally be specifiable in a user-friendly fashion. We propose an adaptive regularization approach where the rigidity constraints are informed by the registration application at hand and whose support is controlled by the geometry of the images to be registered. We investigate the behavior of this technique and discuss its sensitivity to the rigidity parameter.