Biometrics is a technology that uses human character to do person identification by computer . Compared with the traditional methods,it is more convenient,safer,not easy to be forgotten and not easy to be replaced . However,identification that uses only one character is not fitful for application . According to this,in this thesis,new methods about multi-biometrics are introduced . Then,a multi-biometrics system based on face and voice is realized.The following is done in this thesis:Firstly,a multi-biometrics model is introduced based on information fusion. Also,a face recognition module based on PCA and a speaker recognition module based on MFCC and VQ are realized. Secondly,in the different level of pattern recognition,we realize feature-level fusion based on neural network and score-level fusion based on multi-classifier. Thirdly,Some opinions about application of multi-biometrics are given and corresponding prototype systems are realized.
|