Automatic Rigid Registration and Analysis of Colour Fundus Image in Patients with Diabetic Retinopathy

Abstract

This paper describes a new method for detection biffurcation points of retinal vascular tree. This approach uses Hessian matrix for blood vessel segmentation, followed by morphological thinning and binary matched filtering. These bifurcation points are consequently used as landmarks for rigid image registration based on the iterative closest point method. Registered images were used for simple analysis with the aim to enhance the new pathologies in analyzed image sequence.