Palm print is a relatively new biometric technology. Palm print textures like principle lines, wrinkles and ridges contain discriminative information which can be extracted for verification purpose. In this paper, we aim to reexamine the competitive palm code scheme proposed by [6], and introduce some improvement over the proposed method. In order to reduce the computation complexity, we adopted wavelet transform to decompose the palm print image into lower resolution. We are able to show that wavelet transform when integrated with 2D Gabor filter is able to yield verification rate of 96.70%, with an improved verification speed of less than 1 second.
Index Terms:
Biometrics, Gabor filter, Wavelet Transform, Palm Print, Competitive Rule
Citation:
Michael KO Goh, Tee Connie, Andrew BJ Teoh, David CL Ngo, "A Fast Palm Print Verification System," cgiv, pp.168-172, International Conference on Computer Graphics, Imaging and Visualisation (CGIV'06), 2006