An algorithm for exact retinal vein extraction

Abstract

Recently more interest in retina-based human recognition is observable. It is connected with the fact that this biometrics trait can guarantee absolute certainty in the case of human identity. In the literature, one can easily observe that most of the algorithms are based on veins system and its pattern. A method to extract recently mentioned structure is presented in this paper. The proposed approach consists of three main parts: preprocessing, segmentation and unnecessary artifacts removal. During the research, the authors used different image processing methods for veins system extraction, especially diversified binarization algorithms and edge detection approaches were tested. In the time of the experiments the authors took into consideration not only accuracy but also proposed solution efficiency. Performed tests have shown that it is clearly possible to extract veins system with satisfactory precision and efficiency.