Personal identification based on multi-modal hand-print features

A novel personal identification approach using hand-based features is proposed in this paper. In contrast with the existing approaches, this paper extracts fusion biometric features, including hand shape, finger-print and palm-print to implement coarse-to-fine dynamic identification. Hand shape feature is used to guide the fast selection of a small set of similar candidates from database in coarse level matching stage. And the palm-print and finger-print features are used for fine-level identification. To facilitate the identification, all the extracted features are presented as one dimension (1D) signal. Having located 1D signal of finger and virtual circle on palm, wavelet zero-crossing is used to extract feature. In specific, the fusion matching mechanism is applied in decision stage. And the experimental result of 97.2% correct ratio demonstrates the effectiveness of the proposed method done on captured palm database.