Iris recognition based on support vector machine
Yang Shu-fan · Journal of Changchun University of Technology · 2006
First the sampled iris images are preprocessed in the iris recognition system to get the stripe images.With the method of PCA,the features are extracted for reducing the dimension and a 40-dimention corresponded with a training sample.Then the iris recognition is carried out,with the Sequential Minimal Optimization(SMO),by using the Support Vector Machine(SVM).The mean recognition rate is about 94.3%,which shows that the recognition method decrease the training time and improve the training efficiency.