Abstract: The accuracy and speed of iris boundary localization affect recognition system performance in the iris recognilion system. Based on analyzing some prevailing iris recognition algorithms, the edge information of the iris image was extracted by the edge detection operator which is based on Canny's thought, iris inner circle and outer circle were loca lined by Hough transform within the small block images by incorporating prior knowledge, and experimental result shows that this localization method improves the boundary localization speed and it also ensures the localization accuracy. The yawp from iris region includes the eyelid, eyelash, eyelid shadow and specular reflections. The segmental-secondary linear localization method adopting edge detection and Radon transform was proposed to remove the interference from the eyelid on the eyelid localization, the eyelash yawp and eyelid shadow were removed by threshold method, and experimental result shows that the algorithm is efficient and accurate.

Key words: Iris localization, Edge extraction, Hough transform, Eyelid localization