Algorithm 1. The algorithm of Scan-and-Cluster for saliency points estimation of iris outer boundary. |
(1) Define the scanning area from 0.25× to 0.75× height of the ROI of estimated iris mask. |
(2) Search all the points within the range for the right and left side of ROI. |
(3) Repeatedly do following step for all points: |
(a) Compute all possible paired distance of two points using Equation (1). |
(b) Find the maximum value of distance. |
(c) Form a set Ω consisting of all distance values computed from (b). |
(4) Apply K-means algorithm with the value of k = 2 on set Ω. Suppose the subset Θ denotes the cluster that has higher value. |
(5) Saliency points are recovered as all the end points corresponding to all lines that belonged to Θ. |