Periocular-Assisted Multi-Feature Collaboration for Dynamic Iris Recognition
Iris recognition has emerged as one of the mostaccurate and convenient biometric for person identification andhas been increasingly employed in a wide range of e-securityapplications. The quality of iris images acquired at-a-distanceor under less constrained imaging environments is known todegrade the iris recognition accuracy. The periocular informationis inherently embedded in such iris images and can be exploitedto assist in the iris recognition under such non-ideal scenarios.Our analysis of such iris templates also indicates significantdegradation and reduction in the region of interest, where theiris recognition can benefit from a similarity distance that canconsider importance of different binary bits, instead of thedirect use of Hamming distance in the literature. Periocularinformation can be dynamically reinforced, by incorporating thedifferences in the effective area of available iris regions, for moreaccurate iris recognition. This paper presents such a periocular-assisted dynamic framework for more accurate less-constrainediris recognition. The effectiveness of this framework is evaluatedon three publicly available iris databases using within-datasetand cross-dataset performance evaluation, e.g., improvement inthe recognition accuracy of 22.9%, 10.4% and 14.6% on threedatabases under both the verification and recognition scenarios.
. 
Downloading Code
The code can be downloaded from this weblink.