Multi-algorithmic IRIS Recognition
Goski Sathish, S. Saravanan, S. Narmadha, S. Uma Maheswari · International Journal of Computer Applications · 2012
ABSTRACT Modern societies give higher relevance to personal recognition system that contribute to the increase of security and reliability, essentially due to terrorism and other extremism or illegal activities. The objective of this work is to present a multi-algorithmic biometric authentication system for physical access control based on iris pattern for high security access. The CASIA database of IRIS images provided by C hinese A cademy of S ciences I nstitute of A utomation is used and the system is implemented in MATLAB. large database due to high pattern variability among different The iris recognition is based on Daugman's approach and multiple classifiers using Hamming distance and Neural networks. In Daugman's approach, the iris features are extracted using 2D Gabor Wavelets. The proposed work provides match for iris pattern if situations. Though many security forces have launched hamming distance is below 0.15 whereas for the existing works it is 0.20. The Neural Classifier uses a feed forward network with three hidden layers legible fingerprint biometrics system. Thus iris has been taken and used after normalization and feature extraction phase. Features given to neural network are Energy, Entropy, Standard deviation, Covariance. The error rate has been reduced from e