| Biological recognition technology has been widely used in recent world. Iris recognition is an emerging biometric technology. For its exclusive, stability, collectible and not invadable, iris recognition is being more and more regarded by people. In recent years, iris recognition has made progress in technology research and application, and has a wide prospect and market. Compared with face, voice fingerprint and other biometric technology, iris recognition has higher precision. The system based on iris recognition is made up of several parts: iris-capturing device, iris preprocessing, iris feature extracting and coding, iris code matching. This paper finish the following works:1. First, estimate the iris quality and smooth filter the image.2. In iris localization, we improve the iris localization algorithm based-on boundary pick-up and Hough transform. We work out the iris coarse localization algorithm based-on block-searching and iris fine localization algorithm based-on local-area-searching Hough transform, so it can precisely and quickly localize the inner and outer boundaries.3. Then we improve the traditional iris normalization algorithm and work out the iris normalization algorithm based-on line picked-up to normalize the annular region of iris into a rectangle. In order to improve the following algorithms, we lighting correcting and histogram equalize the iris image, then we can get a better iris image.4. We excogitate the Gabor filter based-on sub-block texture analyse to extract the feature of iris normalized image and encode them. Then, we use Hamming distance to match two iris codes. In order to reduce the rotation of iris, we work out a new method.We test these algorithms by the CASIA 2.0 iris database. This method is proven by the experiment results. |
