Localization method of imperfect iris images
Li Ping Su · Applied science and technology · 2012
The iris localization is an important part of iris recognition.However,the iris image collected by a high unforced iris recognition system is usually not perfect,for example,eyelid noise,eyelashes noise and reflection in pupil may be existing in iris images.It brings lots of adverse effect to iris localization,and thereby the accuracy of iris recognition is depressed.Based on the structure characteristic of an iris image,a simple and practicable method is proposed to locate the inner and outer edges of the iris.In order to effectively improve the iris localization precision,on the basis of gray threshold,improvement measures are designed to eliminate the adverse effects brought by eyelid noise,eyelashes noise and reflection in the pupil.