Fluorescein angiography is a common procedure in ophthalmic practice,
mainly to evaluate vascular retinopathies and choroidopathies from seq
uences of ocular fundus images. In order to compare the images, a reli
able overlying is essential. This paper proposes some methods for the
recovery of irregular motion in fluorescein angiograms (FA). The overl
ying is done by a three step procedure: detection of relevant points,
matching points from different images and estimation of the assumed li
near geometric transformation. A stochastic model (closely related to
the general linear model) allows to fuse the second and third steps, T
wo different estimators of the geometric transformation are proposed a
nd tested with real FAs. Images from choroido-retinal diseases have be
en analysed: diabetic retinopathy, vein occlusions and choroidal neova
scular membrane. Results have been evaluated using a different number
of relevant points with different spatial arrangements. Registration a
ccuracy is evaluated as the mean squared error between real and transf
ormed relevant point locations for those points not used to estimate t
he transformation. (C) 1997 Elsevier Science B.V.