Fusing multi-biometrics authorization with PCA

The recognition based on Fusing Multi-Biometrics is a hot point in authentications. A new method fusing multi-biometric features base on PCA is proposed in this paper. At first, the features are extracted from the voice, face images and online handwritings of one person. Then, three features are fused by the means of PCA method. Finally, the authorization is implemented through classification by minimizing the Euclidean distance of different people. The dataset we built includes the voice, face image and online handwritings of 20 volunteers. Our method performs much better on this dataset when the EER (Equal Error Rate) is 1%.