A Novel Unified-Noise Algorithm for Iris Recognition

The iris identification as one of the most effective ways in biometrics identification, which obtain more and more attention in all walks of life to a wide range of applications, but the recognition of the quality and speed still exist noise interference such as eyelash, eyelid, and sampling flare. In the paper we put forward a method that unified noise used to between login iris and sample iris images aiming at the widespread adoption of matching method. Basing on this way, we can effectively reduce the noise interference and increase for identification of utilization as soon as possible. It is found that the method can be applied to reduce regional differences resulted from noise regional disunity errors, and make us to get more effective iris image area, especially in inferiorly non-aggressive iris image area. Our study Indicates the result of calculation iris matching has better effect in direct gray-level surface.