Payment system verification method and system based on multi-factor authentication, medium and equipment

Abstract

The invention discloses a payment system verification method and system based on multi-factor authentication, a medium and equipment, and the method comprises the following steps: S1, face recognition: collecting a face image of a user through a camera, comparing the face image with a pre-stored face feature, completing face recognition if the comparison succeeds, and entering S2; if the comparison is not successful, face recognition is carried out again; s2, voiceprint recognition: collecting the sound of a user through a microphone, extracting voiceprint features, comparing the voiceprint features with pre-stored voiceprint features, and completing voiceprint recognition if the comparison is successful; if the comparison is not successful, voiceprint recognition is carried out again; and S3, password recognition: prompting the user to speak a preset voice password within a specified time, verifying whether the voice content is correct or not, completing voice password recognition, allowing payment if the password recognition is successful, and returning to S1 if the password recognition is unsuccessful. The method has the advantages that the payment system is verified by combining face recognition, voiceprint recognition and password recognition, so that the verification complexity of the payment system is improved, and the payment safety is ensured.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and particularly to a verification method, system, medium and device for a payment system based on multi-factor authentication. Background Art

[0002] With the popularization of mobile payment and online payment, the security of the payment system is crucial to the user's fund safety. The traditional single authentication method has the risk of being breached and is difficult to meet the high-security requirements. Existing payment authentication is often a single verification mode, lacking optimization, resulting in poor user experience or insufficient security. In view of this, it is necessary to provide a coil electroadhesive device. Summary of the Invention

[0003] The verification method, system, medium and device for a payment system based on multi-factor authentication provided by the present invention effectively solve the problem of the single existing payment security verification mode.

[0004] The technical solution adopted by the present invention is as follows:

[0005] The verification method for a payment system based on multi-factor authentication includes the following steps.

[0006] S1. Face recognition: Collect the user's facial image through a camera, compare it with the pre-stored facial features. If the comparison is successful, the face recognition is completed and enter S2; if the comparison is unsuccessful, re-perform face recognition.

[0007] S2. Voiceprint recognition: Collect the user's voice through a microphone, extract the voiceprint features, and compare them with the pre-stored voiceprint features. If the comparison is successful, the voiceprint recognition is completed; if the comparison is unsuccessful, re-perform voiceprint recognition.

[0008] S3. Password recognition: Prompt the user to say a preset voice password within a specified time, verify whether the voice content is correct, and complete the voice password recognition. If the password recognition is successful, payment is allowed; if the password recognition is unsuccessful, return to S1.

[0009] Furthermore: The specific method of face recognition in S1 is:

[0010] S11. Use a convolutional neural network model to extract the feature vector of the user's facial image.

[0011] S12. Compare the extracted feature vector with the pre-stored facial feature vector and calculate the cosine similarity.

[0012] S13. If the cosine similarity is greater than the preset threshold, it is determined that the face recognition is successful; otherwise, it is determined to be a failure.

[0013] Furthermore: The specific method of voiceprint recognition in S2 is:

[0014] S21. Preprocess the user's voice signal, where the preprocessing includes noise reduction and framing;

[0015] S22. Extract the MFCC features of each frame of the voice signal;

[0016] S23. Use a pre-trained GMM model to model the MFCC features and generate a voiceprint feature vector;

[0017] S24. Compare the generated voiceprint feature vector with the pre-stored voiceprint feature vector and calculate the log-likelihood ratio;

[0018] S25. If the log-likelihood ratio is greater than a preset threshold, it is determined that the voiceprint recognition is successful; otherwise, it is determined to be a failure.

[0019] Furthermore: The specific method for voice password recognition is as follows:

[0020] S31. Use a pre-trained ASR model to convert the user's voice password into text;

[0021] S32. Compare the converted text with the preset voice password text;

[0022] S33. If the texts are the same and it is completed within the specified time, it is determined that the voice password recognition is successful; otherwise, it is determined to be a failure.

[0023] Furthermore: In the S1 face recognition, a live detection algorithm based on blink detection or head pose estimation is used to determine whether the user is a real live body. If the user is a real live body, the face recognition is valid; otherwise, the face recognition is invalid.

[0024] Furthermore: In the S2 voiceprint recognition, a live detection algorithm based on voice activity detection and voiceprint dynamic features is used to determine whether the user is a real live body. If the user is a real live body, the voiceprint recognition is valid; otherwise, the voiceprint recognition is invalid.

[0025] Furthermore: When performing S1 face recognition, the user is prompted "Please align with the camera". When performing S2 voiceprint recognition, the user is prompted "Please say the following text: XXX". When performing S3 password recognition, the user is prompted "Please say your pre-stored password within 5 seconds".

[0026] A payment system verification system based on multi-factor authentication includes

[0027] A user registration module for initial registration and storage of user biometric data and voice passwords;

[0028] A face recognition module for face verification during user payment, including image acquisition, feature extraction, feature comparison, and live detection;

[0029] A voiceprint recognition module, used for voiceprint verification during user payment, including voice collection, voiceprint feature extraction, voiceprint comparison, and liveness detection;

[0030] A voice password recognition module, used for voice password verification during user payment, including voice collection, speech recognition, text comparison, and time detection;

[0031] A result feedback module, used to allow or reject a payment request according to the verification result and feedback the verification result to the user;

[0032] A data storage and management module, responsible for the secure storage and management of user biometric data and voice passwords.

[0033] A computer-readable storage medium stores a computer program, and when the computer program is processed and executed, the steps of the payment system verification method based on multi-factor authentication are implemented.

[0034] A computer device includes a processor, a communication interface, a memory, and a communication bus. The processor, the communication interface, and the memory complete communication with each other through the communication bus. Among them:

[0035] The memory is used to store a computer program;

[0036] The processor is used to execute the steps of the payment system verification method based on multi-factor authentication by running the program stored on the memory.

[0037] Advantages of the invention:

[0038] 1. Combining face recognition, voiceprint recognition, and password recognition for payment system verification improves the complexity of payment system verification, thus ensuring payment security.

[0039] 2. Performing liveness detection in face recognition. The liveness detection can be actions such as blinking and opening the mouth, effectively preventing face recognition using forged photos and forged videos.

[0040] 3. Performing liveness detection in voiceprint recognition effectively improves the security and effectiveness of voiceprint recognition.

[0041] 4. Using a convolutional neural network model extraction algorithm improves the accuracy and verification efficiency of face recognition and avoids interference from the external environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 It is a flowchart of the payment system verification method based on multi-factor authentication provided by the embodiments of the present application. DETAILED DESCRIPTION

[0043] To make the above objects, features, and advantages of the present invention more obvious and understandable, the following provides a detailed description of the specific embodiments of the present invention in conjunction with the accompanying drawings.

[0044] As Figure 1 shown, the first embodiment of the present application is a payment system verification method based on multi-factor authentication, including the following steps:

[0045] S1. Face recognition: Collect the facial image of the user through a camera, compare it with the pre-stored facial features. If the comparison is successful, the face recognition is completed and proceed to S2; if the comparison is unsuccessful, re-perform face recognition.

[0046] S2. Voiceprint recognition: Collect the user's voice through a microphone, extract the voiceprint features, and compare them with the pre-stored voiceprint features. If the comparison is successful, the voiceprint recognition is completed; if the comparison is unsuccessful, re-perform voiceprint recognition.

[0047] S3. Password recognition: Prompt the user to say the preset voice password within a specified time, verify whether the voice content is correct, and complete the voice password recognition. If the password recognition is successful, payment is allowed; if the password recognition is unsuccessful, return to S1.

[0048] In the above design, the verification of the payment system is combined with face recognition, voiceprint recognition, and password recognition, which improves the complexity of the payment system verification and thus ensures payment security.

[0049] Specifically: The specific method of face recognition in S1 is:

[0050] S11. Use a convolutional neural network model to extract the feature vector of the user's facial image.

[0051] S12. Compare the extracted feature vector with the pre-stored facial feature vector and calculate the cosine similarity.

[0052] S13. If the cosine similarity is greater than the preset threshold, it is determined that the face recognition is successful; otherwise, it is determined to be a failure.

[0053] In the above design, the convolutional neural network model extraction algorithm is adopted to improve the accuracy and verification efficiency of face recognition and avoid interference from the external environment.

[0054] Specifically: The specific method of voiceprint recognition in S2 is:

[0055] S21. Preprocess the user's voice signal, and the preprocessing includes noise reduction and framing.

[0056] S22. Extract the MFCC features of each frame of the voice signal.

[0057] S23. Use the pre-trained GMM model to model the MFCC features and generate the voiceprint feature vector;

[0058] S24. Compare the generated voiceprint feature vector with the pre-stored voiceprint feature vector and calculate the log-likelihood ratio;

[0059] S25. If the log-likelihood ratio is greater than the preset threshold, it is determined that the voiceprint recognition is successful; otherwise, it is determined to be a failure.

[0060] In the above design, the voiceprint recognition adopts the voiceprint feature extraction and comparison algorithm based on MFCC and GMM, uses the MFCC features to effectively capture the unique features of the sound, and combines the GMM model to improve the accuracy of voiceprint recognition.

[0061] Specifically: The specific method of the voice password recognition is as follows:

[0062] S31. Use the pre-trained ASR model to convert the user's voice password into text;

[0063] S32. Compare the converted text with the preset voice password text;

[0064] S33. If the texts are the same and are completed within the specified time, it is determined that the voice password recognition is successful; otherwise, it is determined to be a failure.

[0065] In the above design, the deep learning-based ASR model can accurately convert speech into text, improve the reliability of voice password recognition, and increase the difficulty of attacking the payment verification system.

[0066] Specifically: In the S1 face recognition, use the live detection algorithm based on blink detection or head pose estimation to determine whether the user is a real live body. If the user is a real live body, the face recognition is valid; otherwise, the face recognition is invalid.

[0067] In the above design, the live detection can be actions such as blinking and opening the mouth, which can effectively prevent face recognition using forged photos and forged videos.

[0068] Specifically: In the S2 voiceprint recognition, use the live detection algorithm based on voice activity detection and voiceprint dynamic features to determine whether the user is a real live body. If the user is a real live body, the voiceprint recognition is valid; otherwise, the voiceprint recognition is invalid.

[0069] In the above design, the voiceprint recognition can use the VAD algorithm based on energy threshold or spectral features to analyze the spectral features of the sound signal and detect whether the speech signal contains real speech activity. It can effectively improve the security of voiceprint recognition.

[0070] Specifically: when performing S1 face recognition, the user is prompted "Please align with the camera"; when performing S2 voiceprint recognition, the user is prompted "Please say the following text: XXX"; when performing S3 password recognition, the user is prompted "Please say your pre-stored password within 5 seconds."

[0071] The second embodiment provided by this application is a verification system for a payment system based on multi-factor authentication, including

[0072] A user registration module for initial registration and storage of user biometric data and voice passwords;

[0073] A face recognition module for face verification during user payment, including image acquisition, feature extraction, feature comparison, and liveness detection;

[0074] A voiceprint recognition module for voiceprint verification during user payment, including voice acquisition, voiceprint feature extraction, voiceprint comparison, and liveness detection;

[0075] A voice password recognition module for voice password verification during user payment, including voice acquisition, speech recognition, text comparison, and time detection;

[0076] A result feedback module for allowing or rejecting a payment request according to the verification result and feeding back the verification result to the user;

[0077] A data storage and management module responsible for secure storage and management of user biometric data and voice passwords.

[0078] The third embodiment provided by this application is a computer-readable storage medium storing a computer program, and when the computer program is processed and executed, the steps of the verification method for the payment system based on multi-factor authentication are implemented. In addition, the computer-readable storage medium of this embodiment can adopt any combination of one or more readable storage media, where the readable storage media include electrical, optical, electromagnetic, infrared, or semiconductor systems, devices, or components, or any combination of the above.

[0079] The fourth embodiment provided by this application is a computer device, including a processor, a communication interface, a memory, and a communication bus, where the processor, the communication interface, and the memory complete communication with each other through the communication bus: where

[0080] The memory is used to store a computer program;

[0081] The processor is configured to execute the steps of the payment system verification method based on multi-factor authentication by running the program stored in the memory. As an implementation manner of the present invention, the communication bus mentioned in the above terminal may be a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, or the like. The communication bus may be divided into an address bus, a data bus, a control bus, and the like.

[0082] As an implementation manner of the present invention, the communication interface is used for communication between the above terminal and other devices.

[0083] As an implementation manner of the present invention, the memory may include a Random Access Memory (RAM), and may also include a non-volatile memory, such as at least one disk memory. Optionally, the memory may also be at least one storage device located far from the aforementioned processor.

[0084] As an implementation manner of the present invention, the above processor may be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it may also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0085] The fifth embodiment provided by this application is a payment system verification method based on multi-factor authentication, including the following steps: S1. Facial recognition: Collect the user's facial image through a camera and compare it with the pre-stored facial features. If the comparison is successful, facial recognition is completed and S2 is entered; if the comparison is unsuccessful, facial recognition is performed again; S2. Voiceprint recognition: Collect the user's voice through a microphone, extract the voiceprint features, and compare them with the pre-stored voiceprint features. If the comparison is successful, voiceprint recognition is completed; if the comparison is unsuccessful, voiceprint recognition is performed again; S3. Password recognition: Prompt the user to say the preset voice password within the specified time, verify whether the voice content is correct, and complete the voice password recognition. If the password recognition is successful, payment is allowed; if the password recognition is unsuccessful, return to S1. The specific method of facial recognition in S1 is: S11. Use a convolutional neural network model to extract the feature vector of the user's facial image; S12. Compare the extracted feature vector with the pre-stored facial feature vector and calculate the cosine similarity; S13. If the cosine similarity is greater than the preset threshold, it is determined that the facial recognition is successful; otherwise, it is determined to be a failure. The specific method of voiceprint recognition in S2 is: S21. Preprocess the user's voice signal, and the preprocessing includes noise reduction and frame segmentation; S22. Extract the MFCC features of each frame of the voice signal; S23. Use a pre-trained GMM model to model the MFCC features to generate a voiceprint feature vector; S24. Compare the generated voiceprint feature vector with the pre-stored voiceprint feature vector and calculate the log-likelihood ratio; S25. If the log-likelihood ratio is greater than the preset threshold, it is determined that the voiceprint recognition is successful; otherwise, it is determined to be a failure. The specific method of the voice password recognition is: S31. Use a pre-trained ASR model to convert the user's voice password into text;

[0086] S32. Compare the converted text with the preset voice password text; S33. If the texts are the same and it is completed within the specified time, it is determined that the voice password recognition is successful; otherwise, it is determined to be a failure. In the S1 facial recognition, a live detection algorithm based on blink detection or head pose estimation is used to determine whether the user is a real live body. If the user is a real live body, the facial recognition is valid; otherwise, the facial recognition is invalid. In the S2 voiceprint recognition, a live detection algorithm based on voice activity detection and voiceprint dynamic features is used to determine whether the user is a real live body. If the user is a real live body, the voiceprint recognition is valid; otherwise, the voiceprint recognition is invalid. When performing S1 facial recognition, the user is prompted "Please align with the camera", when performing S2 voiceprint recognition, the user is prompted "Please say the following text: XXX", and when performing S3 password recognition, the user is prompted "Please say your pre-stored password within 5 seconds".

[0087] For further detailed description, it should be understood that the above are only specific embodiments of the present invention and are not used to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A payment system verification method based on multi-factor authentication, characterized in that: It includes the following steps: S1. Face recognition: Collect the user's facial image through a camera, compare it with the pre-stored facial features. If the comparison is successful, the face recognition is completed and proceed to S2; if the comparison is unsuccessful, re-perform face recognition. S2. Voiceprint recognition: Collect the user's voice through a microphone, extract the voiceprint features, and compare them with the pre-stored voiceprint features. If the comparison is successful, the voiceprint recognition is completed; if the comparison is unsuccessful, re-perform voiceprint recognition. S3. Password recognition: Prompt the user to say the preset voice password within a specified time, verify whether the voice content is correct, and complete the voice password recognition. If the password recognition is successful, payment is allowed; if the password recognition is unsuccessful, return to S1.

2. The verification method of the payment system based on multi-factor authentication according to claim 1, characterized in that: The specific method of face recognition in S1 is as follows: S11. Use a convolutional neural network model to extract the feature vector of the user's facial image. S12. Compare the extracted feature vector with the pre-stored facial feature vector and calculate the cosine similarity. S13. If the cosine similarity is greater than the preset threshold, it is determined that the face recognition is successful; otherwise, it is determined to be a failure.

3. The verification method of the payment system based on multi-factor authentication according to claim 1, characterized in that: The specific method of voiceprint recognition in S2 is as follows: S21. Preprocess the user's voice signal, and the preprocessing includes noise reduction and framing. S22. Extract the MFCC features of each frame of the voice signal. S23. Use a pre-trained GMM model to model the MFCC features and generate a voiceprint feature vector. S24. Compare the generated voiceprint feature vector with the pre-stored voiceprint feature vector and calculate the log-likelihood ratio. S25. If the log-likelihood ratio is greater than the preset threshold, it is determined that the voiceprint recognition is successful; otherwise, it is determined to be a failure.

4. The authentication method for a payment system based on multi-factor authentication according to claim 1, wherein: The specific method of the voice password recognition is as follows: S31. Use a pre-trained ASR model to convert the user's voice password into text. S32. Compare the converted text with the preset voice password text. S33. If the texts are the same and are completed within the specified time, it is determined that the voice password recognition is successful; otherwise, it is determined to be a failure.

5. The verification method of the payment system based on multi-factor authentication according to claim 1, characterized in that: In the face recognition of S1, a live detection algorithm based on blink detection or head pose estimation is used to determine whether the user is a real live body. If the user is a real live body, the face recognition is valid; otherwise, the face recognition is invalid.

6. The verification method of the payment system based on multi-factor authentication according to claim 1, wherein: In the voiceprint recognition of S2, a live detection algorithm based on voice activity detection and voiceprint dynamic features is used to determine whether the user is a real live body. If the user is a real live body, the voiceprint recognition is valid; otherwise, the voiceprint recognition is invalid.

7. The verification method of the payment system based on multi-factor authentication according to claim 1, wherein: When performing face recognition in S1, prompt the user "Please align with the camera". When performing voiceprint recognition in S2, prompt the user "Please say the following text: XXX". When performing password recognition in S3, prompt the user "Please say your pre-stored password within 5 seconds".

8. A payment system verification system based on multi-factor authentication, characterized in that: It includes a user registration module for the initial registration and storage of the user's biometric data and voice password. a face recognition module for face verification during user payment, including image acquisition, feature extraction, feature comparison, and live detection. a voiceprint recognition module for voiceprint verification during user payment, including voice acquisition, voiceprint feature extraction, voiceprint comparison, and live detection. The voice password recognition module is used for voice password verification during user payment, including voice collection, voice recognition, text comparison, and time detection; The result feedback module is used to allow or reject the payment request according to the verification result and feedback the verification result to the user; The data storage and management module is responsible for the secure storage and management of user biometric data and voice passwords.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is processed and executed, the steps of the payment system verification method based on multi-factor authentication described in any one of claims 1 to 7 are implemented.

10. A computer device, characterized in that: It includes a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory complete mutual communication through the communication bus: wherein: The memory is used for storing a computer program; The processor is used to execute the steps of the payment system verification method based on multi-factor authentication described in any one of claims 1 to 7 by running the program stored on the memory.

Images