Abstract
Cross-spectral iris recognition represents the ability of the system to identify the iris images acquired in different electromagnetic spectrums. An iris captured in the near-infrared spectrum (NIR) is matched with an iris obtained in the visual light spectrum (VIS) to boost the recognition performance. In cross-spectral iris recognition, the illumination factor between NIR and VIS images significantly degrades the recognition performance. Therefore, the existing method only achieved recognition performance with an equal error rate (EER) larger than 5%, and it is a challenging issue for cross-spectral performance to have EER below 5%. In this paper, we improve iris recognition performance by concatenating the Gradientfaces-based normalization technique (GRF) to a standard (conventional) iris recognition method to alleviate the illumination effect. In addition, we integrate the GRF with a Gabor filter, a difference of Gaussian (DoG) filter, and texture descriptors, namely a binary statistical image feature (BSIF) and a local binary pattern (LBP). The experimental results show that the GRF can boost the cross-spectral iris recognition performance with an EER equals to 1.69%. In addition, the best cross-spectral iris recognition performance is achieved when the GRF is integrated with the Gabor filter and the BSIF.
Highlights
In recent decades, iris recognition has received considerable interest in human identification [1], [2]
The Gradientfacesbased normalization technique (GRF) is a gradient-based orientation which considers the relationship of neighbouring pixel points and is suitable to be combined with texture features such as binary statistical image feature (BSIF), which extract the statistical features from near-infrared spectrum (NIR) and visual light spectrum (VIS) images
The best framework is achieved using the combination of a Gabor GRF fused with BSIF descriptors and Difference of Gaussian (DoG) filtering with an equal error rate (EER) of 1.02% at genuine acceptance rate (GAR) 98.97%
Summary
Iris recognition has received considerable interest in human identification [1], [2]. The iris recognition in NIR imaging is more robust compared to VIS imaging because of fewer reflections; the NIR imaging dismisses almost all of the prominent pattern or patterns of the pigment melanin in the iris In this case, cross-spectral iris matching schemes are considered as some of the most accurate recognition frameworks [15]. We propose a novel framework of cross-spectral iris matching using an integration of Gradientface-based normalization (GRF). We concatenated the GRF with a Gabor filter, a DoG filter, BSIF, and local texture descriptors, namely an LBP, to obtain the best integration descriptors This method is the first in the literature that implements the GRF to reduce illumination variation in cross-spectral iris recognition. A novel feature extraction method for cross-spectral iris matching that is illumination invariant and representative in both NIR and VIS imaging, based on integrated Gradientface-based normalization and the texture descriptors.
Sign-in/Register to view highlights & summary 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
