Abstract
To reduce the influence of the eyelid for the iris recognition rate, an eyelid detection algorithm for the iris recognition is proposed. The grayscale morphological operations are employed to remove the interference of the eyelash and the light spot to the eyelid region. The points of the minimum grayscale value of each column in the eyelid region are extracted as edge points. The least squares parabolic fitting eyelids is carried out for the edge points of the eyelid. The eyelid parabolic moving up and down in neighborhood region is used to localizing the eyelid precisely, when the grayscale mutation happens. The experimental results show that the eyelid localization algorithm can detect the eyelid effectively and quickly. Compared with the hough transform algorithm, the accurate rate is the same, but the speed is improved about 4 seconds.
Sign-in/Register to access full text options 
Free) 
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
