TECHNIQUES FOR TEMPORAL REGISTRATION OF RETINAL IMAGES (MP-P7)

Author(s) : Bin Fang (Department of Computer Science, HKBU, Hong Kong) Wynne Hsu (Department of Computer Science, NUS, Singapore) Mongli Lee (Department of Computer Science, NUS, Singapore)

Abstract : Temporal registration of retinal images is fundamental in tracking the evolution of eye-related diseases and providing important information for physicians to decide on further treatments. The vascular structure of the retina is the most appropriate feature representation for retinal fundus image registration. This paper describes a fast chamfer matching applied to the vascular structure to align pairs of fundus images. While the fast chamfer matching is able to achieve good alignment in most cases, it may be trapped in some local minimal in other cases. To alleviate this problem, we propose a non-parametric elastic matching method. The two matching algorithms are tested on 98 pairs of temporal fundus images. We found that elastic matching gives 100% success in registration while there are 3 failure cases for the fast chamfer matching method.

Menu