A feature extraction method of palmprint recognition based on Two-Dimensional Principal Component Analysis (2DPCA) is proposed in this work. A series of experiments were performed on the PolyU- Online- Palmprint -Database with a nearest neighbor classifier and cosine distance. The recognition rate is 99.14%. The 2DPCA method has more recognition accuracy and more computationally efficient than PCA, especially in the small training samples. At the same time the selection of threshold has been researched in different application systems.
展开▼
