VeriEye iris recognition technology detects partial obstructions, irregularities and deformations in the iris shape

(16/06/2008)

No two irises are exactly alike, and although they may appear to be round, no iris is perfectly circular. Neurotechnology has announced the addition of VeriEye iris recognition technology to the company's suite of biometric identification algorithms and software development products.

VeriEye SDK features an advanced iris recognition algorithm that rapidly detects and compensates for partial obstructions, irregularities and deformations in the iris shape - even when the eyes are gazing away. This provides a high degree of accuracy that makes VeriEye suitable for use in a wide range of applications for security, access control, time and attendance and other systems that require fast and reliable identification.

Unlike some iris identification systems that approximate the iris shape as a circle, Neurotechnology's new iris segmentation algorithm can detect the boundaries of irises that are not perfectly circular, are partly occluded by eyelids or eyelashes or deformed in shape because the eye is looking away from the image capture device. This enables VeriEye to more precisely extract all visible iris texture and significantly enhance iris recognition accuracy and reliability.

Iris texture is modeled through unique features that remain true even with varying contrast and illumination levels. VeriEye takes less than a half-second to create a record of unique iris features from an unsegmented human eye image.

The new iris matching algorithm in VeriEye rapidly compares two sets of iris features and is similarly capable of detecting and compensating for obstructions on iris texture and in deformed "looking away" iris images. Matching speeds against existing iris records can range from 50,000 to 150,000 irises per second, depending on configuration.

"We are very pleased to add VeriEye SDK for iris recognition to our line of biometric fingerprint, face and object recognition technologies," said Justas Kranauskas, Neurotechnology's iris identification project lead. "Having a full suite of biometric recognition modalities allows us to provide a best- fit solution for each customer."

VeriEye provides a high degree of accuracy and reliability. When tested on the publicly available dataset ICE2005 Exp 1 (NIST ICE-1), Neurotechnology's VeriEye iris recognition algorithm achieved the following results at a matching speed of 50,000 iris templates per second:
- 0.11% False Rejection Rate (FRR) at a False Acceptance Rate (FAR) of 0.1%
- 0.16% FRR at FAR=0.01%, and
- 0.33% FRR at FAR=0%.

These are the best known results ever to have been achieved on this database. Similar results are achieved on other datasets.

VeriEye supports the following standards:
- ANSI INCITS 379-2004 (American National Standard for Information Technology - Iris Image Interchange Format)
- ISO/IEC 19794-6 (Information technology - Biometric data interchange formats - Iris image data)

VeriEye Standard SDK and VeriEye Extended SDK are available now with licensing options through Neurotechnology or from distributors worldwide. A 30-day trial version with full functionality is also available for download.

Related topics:  Building access control system   Eye biometrics 

Print version | Email to a friend | Related articles

Data breaches: Trends, costs and best practices gives you all the latest information on securing personal and corporate data, key recommendations for immediate action to improve data security, and how to respond to data breaches.

Other Security news and resources