Feature selection for biometrie iris recognition
Kateryna Olehivna Ivanko, N. Budik, Н. Г. Иванушкина · 2017
The approaches to biometrie identification of a person by iris recognition are considered in the paper. The coefficients obtained by two-dimensional discrete Fourier transform, two-dimensional discrete cosine transform, and two-dimensional wavelet transform are compared as features for iris recognition. Also, a comparison of the accuracy of different classifiers for the examined features is carried out. On the basis of this, a combination of the features and a classifier that provides the greatest accuracy of iris recognition is chosen.