Retinal image registration using control points
Conor Heneghan, P.J. Maguire, N. Ryan, Philip de Chazal · 2003
A method for registering pairs of digital ophthalmic images of the retina is presented using anatomical features as control points present in both images. In general, the matching between control points is unknown, however. An automated algorithm is used to determine the matching pairs of control points in the two images as follows. Using two control points from each image, rigid global transform (RGT) coefficients were calculated for all possible combinations of control point pairs. The set of RGT coefficients was then exhaustively searched and the cluster of coefficients associated with the matched control point pairs identified. This cluster was identified by calculating the Euclidean distance between each set of RGT coefficients and its Rth nearest neighbor and then using the EM algorithm to identify matched pairs of control points. Once control point pairs were established, registration of the two images was achieved by using linear regression to optimize the bilinear or 2nd order polynomial. Two examples of registration are shown. The first is an example of cross-modal image registration using an optical image and a fluorescein angiogram of an eye. The second examples shows an example of the registration of two images of an infant eye captured thirteen days apart.