Image enhancement and iris localization based on 2D complex matched filter for noisy images

Iris image preprocessing is one of the most important steps in iris recognition system. It includes both iris localization and some solutions for in-band noise removal like lighting, eyelashes, eyelids, reflections, etc. In this paper, it proposed a new iris-preprocessing algorithm, in which adopted on one of the famous phenomena on communication system (Matched filter). After this, it made some enhancement for noisy databases that is necessary in the automatic iris recognition system. The algorithm is test to be in effect through our experiments on 2400 iris images from The Institute of Automation, Chinese Academy of Sciences and 2435 iris images University of Bath database. Our goal is to remove the unwanted information and emphasis the wanted one in less time for noisy images. This improvement is significant for the real time application systems that are sensitive to noise, demand precise segmentation results and has time constraints.