A fingerprint and voiceprint fusion identity authentication method

Abstract

The invention belongs to the technical field of biometric identity authentication, and is a fingerprint and voiceprint fusion identity authentication method. Firstly, fingerprint images and voice signals are collected, and after corresponding processing, fingerprint and voiceprint fusion authentication templates are generated; then user fingerprint images are collected and processed to obtain The user fingerprint feature sequence, calculate the cosine similarity with the fingerprint template to obtain the fingerprint similarity score, collect the user voice signal and perform framing, MFCC sequence calculation and other processing to obtain the user voiceprint feature sequence, and calculate the voiceprint similarity score , and then use the posterior probability to judge the fingerprint and voiceprint fusion according to the fingerprint and voiceprint similarity scores. In each authentication process, the present invention dynamically adjusts the proportion of fingerprints and voiceprints in fusion authentication according to the level of posterior probability, and optimizes the final fusion authentication effect.

Description

technical field [0001] The invention belongs to the technical field of biometric identity authentication, and in particular relates to an identity authentication method based on the dynamic fusion of fingerprints and voiceprints. Background technique [0002] With the vigorous development of the Internet + industry, the demand for user ubiquitous identity authentication is becoming stronger and stronger. At present, password authentication is still the main means of user identity authentication in the Internet-based service industry, but at the same time, identity authentication means based on human biometric features are rapidly becoming popular. Although password authentication is easy to deploy, it has the problem of insufficient password strength due to limited user memory ability. Biometric authentication can relieve the burden of human memory key, and the convenience of deployment can be improved with the progress of software framework and biometric acquisition equipme...

AI Technical Summary

Problems solved by technology

[0004] In order to solve the problems existing in the existing multi-modal biometric authentication technology, the present invention proposes a fingerprint and voiceprint fusion identity authentication method, which uses dynamic Bayesian analysis to dynamically perform fingerprint and voiceprint biometrics in the judgment stage. Fusion computing can improve the authentication accuracy by more than 5% compared with existing methods

Images