RETINAL ANGIOGRAM REGISTRATION BY ESTIMATION OF DISTRIBUTION ALGORITHM

Retinal fundus photographs are employed as standard diagnostic tools in ophthalmology. We employ optimization techniques for registration of retinal angiograms, using non-linear pre-processing (Wiener filtering and morphological gradient) and computation of a similarity criterion. The present work makes a comparison between different optimization techniques, namely the optical flow minimization method, the Nelder-Mead local search, the CEDA and CHEDA metaheuristics. The impact of the resolution and median filtering of gradient image is studied and the robustness of the approaches is tested through experimental studies, performed on ICG angiographies. Our proposed method has shown interesting results, especially for high resolution registration problems.