A tunable incremental factor algorithm in fundus indocyanine green angiogram registration and mosaicing

A tunable incremental factor is introduced to the inverse compositional image alignment method in registration of fundus blood vessel angiograms under confocal scanning laser ophthalmoscope, to build angiogram mosaics which have larger field of view without loss of resolution to assist diagnosis. The algorithm costs less iterations to converge usually and hardly causes degradation to the registration precision, mainly due to the favorable convexity of the normalized Sum of Squared Differences (SSD) function between the angiograms. The effectiveness and efficiency of the tunable incremental factor algorithm, as well as the convexity of the normalized SSD function, is exemplified in the experimental results on clinical indocyanine green angiograms.