Fast Approach for Iris Detection on GPU by Applying Search Localization for Circular Hough Transform
Pinkey Kumari, Srinath R. Naidu · 2018
Security applications are adopting iris-based biometric system for identification and authorization of individuals. However, the computational cost for these algorithms is very high. In this paper, we present a faster approach to iris detection on GPU (Graphical Processing Unit) applying search localization for Hough Transform (HT). We have utilized the parallel computing capability of GPU along with Search Localization (SL) to accelerate the iris detection using Circular Hough Transform and compared the performance on CPU (Central Process Unit) and GPU. SL reduces the region over which Circular Hough Transform is performed and hence gives better performance.