Noise Detection of Iris Image Based on Tezture Analysis
Noise detection is very important in an iris recognition system. A novel noise detection method for iris images is presented in this paper. According to the texture feature of different noises, a 2-D circular Gabor Filter is designed to detect specular reflection and estimate pupils location, then, a 1-D peak Gabor Filter is proposed to detect eyelid boundary and eyelashes. Furthermore, eyelid is localized on eyelid boundary image by parabolic Integrodifferential operator. In terms of the experiment on the CASIA-IrisV3-Lamp iris database, which contains 16214 iris images, the correct rate of pupil estimation is 100％ and eyelid localization is 99.4％ respectively. The results show that the proposed method is quite effective.
Iris Recognition Noise Detection Tezture Feature Gabor Filter Integrodifferential Operator
ZHANG Xiangde WANG Qi ZHU Hegui YAO Cuili GAO Longcheng LIU Xianyan
College of Sciences, Northeastern University, Shenyang 110004, China
国际会议
2009年中国控制与决策会议(2009 Chinese Control and Decision Conference)
广西桂林
英文
2366-2370
2009-06-17（万方平台首次上网日期，不代表论文的发表时间）