Iris images are increasingly employed in national ID programs and largescale iris databases have been developed. Conventional iris images for the biometrics databases are acquired from close distances under near infrared wavelengths. However the surveillance data is often acquired under visible wavelengths. Therefore, applications like watch-list identification and surveillance at-a-distance require accurate irismatching capability for images acquired under different wavelengths. In this context, simultaneously acquired visible iris images should be matched with the iris images acquired under near infrared illumination to ascertain cross-spectral iris recognition accuracy. This paper describes the need for such cross-spectral iris recognition capability and proposes the development of bi-spectral iris image database to advance the much needed research in this area.
|Name||Advances in Intelligent Systems and Computing|
- Control and Systems Engineering
- Computer Science(all)