Iris is taken in consideration as very appropriate for the identification and the verification of humans because the error average
is the lowest in ultimate biometric identity technique and provides an excellent recognition execution. In this research Canny Edge technique
has been adopted and processed to extract the edges of pupil / iris. Circular Hough Transform has been applied for finding the radius and
the pupil center.. Daugman Rubber Sheet has been applied to acquire the normalization iris by transform from Cartesian to polar coordinate.
All experiments were performed on the CASIA V3 dataset of 300 images (60 templates and 240 test images) provided by the Chinese academy
of sciences. Independent Component Analysis (ICA) has best performed on iris images with a goal to discovery a linear representation of non Gaussian data, for this purpose, ICA is implemented to find the iris
features.
In this research, the way is adopted in our calculation using whitens ICA and Euclidean distance. The outcomes show that
the technique gave best recognition accuracy. The recognition accuracy reached 97.3%
تاريخ النشر
15/09/2019
الناشر
Aus Revista
رقم المجلد
26
رقم العدد
2
رابط DOI
DOI:10.4206/aus.2019.n26.2.15
الصفحات
113-115
الكلمات المفتاحية
Iris recognition, Whitens Independent component Analysis, Circular Hough Transform, and Daugman rubber sheet model