This document presents a pupil detection technique using the center of gravity method. It first applies Gaussian filtering, double thresholding, and morphological closing to an eye image to isolate the pupil region. It then uses the center of gravity method to calculate the x and y coordinates of the pupil center by dividing the total pixel values in each dimension by the total number of black pixels. Experimental results on the CASIA iris database demonstrate the accuracy of the proposed computationally efficient pupil detection method. A hardware implementation of the technique is also presented, which could be used for real-time iris localization in biometric recognition applications.