Palmprint verification based on textural features by using Gabor filters based GLCM and wavelet

The palmprint is one of the most reliable physiological characteristics Among different approaches that exist in biometric. palmprint due to having high acceptability, stability and low cost of implementation has drawn attention from researchers. In this paper, we considered the palmprint as a texture and applied two types of feature extraction methods, namely Gabor filters based Gray-Level Co-occurrence Matrix and Discrete Wavelet Transform. In total 350 features that are extracted by these approaches, fifty superior features selected by the forward feature selection algorithm. Features are classified with new method of using reference features in order to achieve higher resolution and by using K-Nearest Neighbor and Fuzzy K-Nearest Neighbor classifiers. In CASIA testing database of 5,502 palmprint samples from 312 palms, we achieved Equal Error Rate of 1.25% ± 0.56 and Accuracy of 98.75% ± 0.56 with 60% train by K-Nearest Neighbor classifier.