Hypergeometric Filters For Optical Flow and A ne Matching

This paper proposes new \hypergeometric" lters for the problem of image matching under the translational and aane model. This new set of lters has the following advantages: (1) High-precision registration of two images under the translational and aane model. Because the window eeects are eliminated, we are able to achieve superb performance in both translational and aane matching. (2) AAne matching without exhaustive search or image warping. Due to the recursiveness of the lters in the spatial domain, we are able to analytically express the relation between lter outputs and the six aane parameters. This analytical relation enables us to directly compute these aane parameters. (3) Generality. The approach we demonstrate here can be applied to a broad class of matching problems, as long as the transformation between the two image patches can be mathematically represented in the frequency domain.