Vessel Tracking for Retina Images Based on Fuzzy Ant Colony Algorithm
Authors: Sina Hooshyar, Rasoul Khayati and Reza Rezai.
Abstract:
In this paper we present a novel fuzzy algorithm for vessel tracking in retina images. The main tools of this system are Ant Colony Optimization algorithm (ACO) and eigenvector analysis of Hessian matrix. ACO, inspired by food-searching behaviors of ants and performs well in discrete optimization, has been used for optimizing objective function of fuzzy C-means(FCM) model and clustering pixels into vessel and background clusters and Hessian matrix has been used for determining vessel direction in tracking process. Estimating full vessel parameters, overcoming initialization and profile modeling in related works and handling junction of vessels in retina image are the most important advantages of this method. Experiments and results of proposed algorithm in ocular fundus image show its good performance in vessel tracking and parameters estimating.