Robust rigid image registration with arbitrary extrinsic photometric noise

A recently proposed multi parametric quadratic programming (QP) based approach to image matching under complex photometric variations has shown quite an improvement over other state-of-the-art algorithms. However, it is not an entirely satisfactory solution for many practical applications as it does not account for geometric dissimilarities between the images. We extend the QP-based method and propose a new method for robust rigid image registration under both photometric and geometric deviations. The new algorithm retains the distinctive photometric modelling power of its predecessor, while it can also handle affine geometric transformations.