Palmprint recognition based on two-dimensional Gabor wavelet transform and two-dimensional principal component analysis

Abstract

A method of palmprint recognition based on Two-Dimensional Gabor(2DGabor ) Wavelet Transform and Two-Dimensional Principal Component Analysis(2DPCA) is proposed. This method firstly carries out 2DGabor wavelet transform for the palmprint image, so that the amplitude of the filtered image can be used as the eigenvector for the palmprint image. And the 2DPCA is used to decrease the dimension of the eigenvector, and the nearest neighbor classifier is employed for palmprint classification.Finally, simulation experiment to compare the method of palmprint recognition based on 2DGabor wavelet and 2DPCA with the method based on Gabor wavelet and PCA is conducted with the assistance of the palmprint image database provided by Hongkong Polytechnic University. The experiment shows that the palmprint recognition method based on 2DGabor wavelet transform and 2DPCA is more effective. © 2011 Springer-Verlag.