Active shape models—their training and application
Computer Vision and Image Understanding
Pattern Classification (2nd Edition)
Pattern Classification (2nd Edition)
Multispectral Iris Analysis: A Preliminary Study51
CVPRW '06 Proceedings of the 2006 Conference on Computer Vision and Pattern Recognition Workshop
Robust modified active shape model for automatic facial landmark annotation of frontal faces
BTAS'09 Proceedings of the 3rd IEEE international conference on Biometrics: Theory, applications and systems
Iris localization in frontal eye images for less constrained iris recognition systems
Digital Signal Processing
| Hi-index | 0.00 |
Uniqueness of iris patterns among individuals has resulted in the ubiquity of iris recognition systems in virtual and physical spaces, at high security facilities around the globe. Traditional methods of acquiring iris patterns in commercial systems scan the iris when an individual is at a predetermined location in front of the scanner. Most state-of-the-art techniques for unconstrained iris acquisition in literature use expensive customequipment and are composed of amulticamera setup, which is bulky, expensive, and requires calibration. This paper investigates a method of unconstrained iris acquisition and recognition using a single commercial off-the-shelf (COTS) pan-tilt-zoom (PTZ) camera, that is compact and that reduces the cost of the final system, compared to other proposed hierarchical multicomponent systems. We employ state-of-the-art techniques for face detection and a robust eye detection scheme using active shape models for accurate landmark localization. Additionally, our system alleviates the need for any calibration stage prior to its use. We present results using a database of iris images captured using our system, while operating in an unconstrained acquisition mode at 1.5m standoff, yielding an iris diameter in the 150-200 pixels range.