Research on Identification Verification Based on Iris
Zhu Ai-qing · Jisuanji fangzhen · 2011
Identification verification problem was studied.Traditional identification verification methods have different defects,while iris has the strongpoints of uniqueness and stability,non-invasive and etc,therefore,it is very suitable for identification verification.In order to improve the accuracy of the identification verification,this paper presented a new iris verification method.Firstly,integral-differential and polar methods located and normalized iris;then wavelet torque and principal component analysis were used to extract and select the iris features,reducing the characteristics dimensions and decreasing the computational complexity.Then Hai Ming distance was used to match the iris and realize the identity verification.Using CASIA iris database to carry out the simulation experiment,the experimental results show that,compared with other identity verification algorithms,the proposed method improves the accuracy of identity verification,shortens verification time and improves the verification efficiency.It is very suitable for real-time demand high identity verification occasion.