Personal Identification by Fusing Hand Shape Geometry and Palmprint Features

This paper presents a new personal identification approach with fusion of hand shape geometry and palmprint features based on Gabor wavelet transformation. Two kinds personal features can be extracted form the low-resolution hand images. Hand shape geometry features include length of four fingers and width of palm. The palmprint features are composed of principal lines, wrinkles, minutiae, delta points. In pattern matching, the normalization method of matching degree is proposed. At the same time square difference distance is used to calculate feature matching degree of the hand shape geometry and palmprint. The result of experiment show that this approach is the most suitable, acceptable and the higher recognition rate, respectively, using different feature extraction methods.