Analogizing the Thinning Algorithm and Elicitation of Vascular Landmark in Retinal Images

1. Introduction

Retinal pattern is a unique biometric identification which forms the inner layer of the eye where even identical twins do not share the same pattern. The pattern remains unchanged from birth to death. The retinal blood vessel pattern plays a vital role as a secured personal identification mark. The fundus image obtained from the retina scanners can be accessed only from living human beings because the retina decays soon after the death. So it is mainly used in military bases and governments where very high secured identifications are required. In the article by Kataryzna and Adam (2004) the authors say that the retinal blood vessels are the land mark of fundus images. A typical fundus image and the important parts of retina are shown in Figure 1.

Figure 1.

Sample retinal image

The retina is found in the posterior region of an eye which forms a tissue like substance mainly composed of Optic nerve, blood vessels and Macula. In the article by Fabiola and Edgardo (2009) the authors says that the blood vessel pattern forms a unique structure for each individual and are analyzed which results in tiny characteristics points.

3. Proposed Methodology

This paper focuses on the thinning methods for the retinal image. The retinal characteristics are obtained and compared by applying Zhang suen and morphological thinning algorithm. The different stages used in Retinal minutiae extraction are shown in Figure 2

Figure 2.

Stages in retinal feature extraction