Identity authentication method and system based on fingerprint sound waves

Abstract

The invention relates to a fingerprint identification technology, and aims to provide an identity authentication method and system based on fingerprint sound waves. Comprising the steps that a user puts the finger pulp of a finger on a friction material, the finger pulp slides on the surface of the friction material for multiple times, and a microphone records friction sound in the finger moving process; carrying out noise reduction processing on the collected sound signal, and intercepting a sliding friction audio signal after noise reduction; performing feature extraction on the intercepted audio signal to obtain a third-level audio feature based on a fingerprint feature; fusing, classifying and predicting the extracted audio features by using a plurality of classifiers by using an integrated classification learning method; and voting the prediction result according to the weight to obtain a user identity authentication result, and storing the user identity authentication result with the authentication result. According to the invention, it can be ensured that fingerprint authentication can only be based on field living body identification, and the security is greatly enhanced while the privacy of the user is protected; the fingerprint information of the user can be obtained only through the microphone, the hardware cost and the space can be saved to the maximum extent, and the method has good industrial application prospects.

Description

Technical field [0001] The present invention relates to fingerprint recognition technology, in particular to an identity authentication method and system based on fingerprint sound waves. Background [0002] From wearable devices to smart homes, electronic devices have gradually penetrated into people's lives. While the Internet of Things is in full swing, access control of electronic devices has also become an important issue. [0003] In the traditional identity authentication scheme, fingerprint recognition is one of the most commonly used technologies, but a large number of electronic devices due to the size or cost of the limit, can not configure fingerprint scanner and other identity authentication systems, so it is impossible to authenticate users to achieve access control and user privacy protection, leaving a lot of security risks. [0004] There are many schemes for voiceprint recognition, such as the long short-term memory network and clustering method proposed in the ...

AI Technical Summary

Problems solved by technology

[0003] In the traditional identity authentication scheme, fingerprint recognition is one of the most commonly used technologies. However, due to the limitation of volume or cost, a large number of electronic devices cannot be equipped with identity authentication systems such as fingerprint scanners, so it is impossible to authenticate users to achieve Access control and user privacy protection leave many security risks

[0004] There are currently many voiceprint recognition schemes, such as the scheme based on long-term short-term memory network and clustering method proposed in the document "Speaker Diarization with LSTM", and the volume-based These schemes analyze the characteristics of human speech and can realize the function of identity authentication, but they are vulnerable to replay attacks. Attackers can record and use the voice of legitimate users to deceive the voiceprint authentication system.

Images