Abstract
In this paper, we focus on common approaches in the area of eye and iris localisation. We describe in detail a new method for detecting iris in digital images. Our method is simple yet effective and estimation of the two circular boundaries is based on curve fitting and cost function maximisation. When we are searching for limbic boundary, we use the inverse gradient and star-like sampling scheme. Rather, an analytical approach is used during detection of pupillary boundary. We define the cost function reflecting the properties of brightness distribution in the pupil region. Our algorithm can be described in three simple steps; at first, we detect the bright point inside the pupil; at second, outer limbic boundary is found via statistical measurements of outer boundary points; and at third, inner boundary points are searched by means of defined cost function maximisation.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
