Face payment security method based on secure unit and trusted execution environment

Abstract

The present invention provides a face payment security method based on a security unit and a trusted execution environment. Through the face recognition trusted application and the face recognition camera to collect face liveness detection data and face image data, the trusted execution environment calculates the liveness percentage and then calls the security unit to obtain the percentage signature result and ciphertext data. The payment application determines whether the liveness percentage is greater than the preset liveness percentage threshold. If it is greater, the rich execution environment combines the password keyboard to perform user identity authentication and the remaining transactions. In this solution, the security unit provides secure cryptographic algorithm services and face recognition data protection keys, the trusted execution environment ensures the secure execution of the face liveness detection algorithm, and ensures the integrity, authentication and confidentiality of each data in the face payment process through direct connection with the face recognition camera.

Description

[0001] This application is a divisional application with the application date of August 30, 2019, application number: 2019108129583, and the invention name: Facial payment security method and platform based on security unit and trusted execution environment. Technical Field

[0002] The present invention relates to the field of face payment, and in particular to a face payment security method based on a security unit and a trusted execution environment. Background Art

[0003] With the rapid development and evolution of artificial intelligence technology, face recognition technology has gradually been applied to the field of financial payment, and the emergence of face recognition payment applications has introduced potential security risks to the original payment system. Without targeted security protection, face recognition-related algorithms and face image data are vulnerable to external malicious attacks in the common operating system of existing payment applications, causing users' payment transactions to face severe security threats. Designing a face recognition secure payment application platform can effectively improve the defense capability of payment applications and ensure the security and stability of payment processes using face recognition as a way to identify and authenticate users. The secure payment application platform is based on a security unit and relies on a trusted execution environment to prevent the face liveness detection results and user face images from being maliciously tampered with, stolen and forged, thereby ensuring terminal payment security. The security unit solves the problems of core key storage and cryptographic algorithm services. The trusted execution environment solves the problems of face liveness detection algorithm execution, algorithm results and face image integrity, authentication and confidentiality protection.

[0004] The secure element (SE) used in the payment application platform is usually an independent security chip that can prevent external malicious parsing attacks and protect the security of core sensitive data. It has cryptographic algorithm logic circuits in the chip and can provide secure cryptographic algorithm services to the outside. The concept of SE in the financial field originated from financial IC card chips. Later, SE chips with similar functions were gradually promoted on financial transaction terminals. In recent years, mobile terminals have also begun to be equipped with dedicated embedded SE chips. SE can not only prevent logical attacks from the software layer, but also resist physical attacks. Even if it is physically destroyed and disassembled, it can protect the security of the data stored in it. SE has extremely high security protection capabilities, but its computing and storage resources are limited. It is usually only used to protect core sensitive security key data and provide underlying cryptographic algorithm services. It is not suitable for the protection of large-scale data and more complex program logic execution.

[0005] Trusted Execution Environment (TEE), borrowed from the idea of ​​trusted computing technology, aims to protect security-sensitive code execution and related data information from attacks and damage by malicious adversaries, and is the basis for building a trusted mobile terminal platform. TEE is mainly composed of a microkernel operating system, isolated from a rich execution environment (REE) composed of ordinary terminal operating systems (such as Android and iOS). TEE can resist attacks from the software layer, and its security protection capability is lower than SE, but it actually runs on the terminal's main control chip (CPU), has strong computing power, and can execute complex logic programs. The International Standards Organization (GlobalPlatform, GP) developed a standard white paper for TEE in 2011 and provided a system architecture design guide. ARM proposed the TrustZone isolation technology that provides hardware support for TEE, and a highly secure isolation environment can be built through a bottom-up approach. In recent years, building TEE on mobile terminal devices has become a standard configuration for mobile phone manufacturers.

[0006] Face recognition technology has been widely used in security, finance, smart home and other fields, with the purpose of achieving face comparison and identification, and using the face as a biometric feature to complete the identification and authentication of individual users. Face recognition is applied to the payment field to identify and confirm the identity of the payment subject through facial feature recognition. It can not only identify the user (and its corresponding account) (analogous to determining the bank card number), but also can be used as an identity authentication factor to confirm the identity (analogous to entering a password). During use, in order to improve security, face recognition is often combined with authentication methods such as user password input to complete the identity authentication of the payment user. The algorithms related to the face recognition process and payment security mainly include liveness detection algorithms and face comparison algorithms. The former is mainly used to determine that the face image comes from a real live body to prevent fake attacks such as photos and videos, and the latter is mainly used to determine the identity of the current user in a large amount of user image data. In payment applications, only the liveness detection algorithm is generally integrated, while the face comparison algorithm is mainly deployed in the backend server of the payment institution or authentication institution, and is used in conjunction with a large number of face image databases.

[0007] In order to reduce the security risks of payment transactions, existing payment application devices have made certain security designs in the operating system, physical hardware and password input protection. However, for face recognition algorithms and related processes, the above security designs are not enough to protect the safe execution and results of the algorithms, and face image data is at risk of leakage, tampering and forgery. The face recognition secure payment application platform based on TEE and SE can effectively solve related security issues, but in specific implementation and application, there are still the following problems:

[0008] 1. Existing payment applications are not designed specifically to protect face liveness detection algorithms, and are unable to protect the integrity, authentication, and confidentiality of algorithm execution and algorithm results.

[0009] 2. Existing payment applications do not provide protection against facial recognition processes.

[0010] 3. Facial image data is the user's private data. Once leaked, it will bring long-term security risks to users. Existing payment applications process facial images in the common operating system REE. Since REE is easily hacked, facial image data also faces a high risk of leakage. Therefore, how to design security protection solutions for facial image data at various stages such as terminal generation, processing and transmission to ensure the security of image data in the facial recognition payment process has become an urgent problem to be solved. Summary of the invention

[0011] The present invention provides a face payment security method based on a security unit and a trusted execution environment, which is used to solve the problem in the prior art that the integrity, authentication and confidentiality of face payment related algorithm execution, algorithm results, face recognition process and face image data cannot be protected during face payment.

[0012] In order to achieve the above-mentioned purpose, the technical solution of the present invention provides a face payment security method based on a security unit and a trusted execution environment, including: a face recognition trusted application in a trusted execution environment generates a liveness percentage based on the collected face liveness detection data and face image data. The liveness percentage and face image data are signed and encrypted in the security unit, and the generated ciphertext data is returned to the face recognition trusted application. The payment application determines whether the liveness percentage is greater than the preset liveness percentage threshold. If it is greater, identity authentication is performed, and the remaining transactions are performed after passing.

[0013] As a preferred embodiment of the above technical solution, preferably, the face recognition trusted application collects face liveness detection data and face image data, including: the payment application determines whether the collection of the face liveness detection data and face image data is completed. If not, the payment application calls the face recognition trusted application through the trusted execution environment. The face recognition trusted application activates the face recognition camera to capture face data.

[0014] As a preferred embodiment of the above technical solution, preferably, obtaining the liveness percentage includes: after the face recognition camera is activated by the face recognition trusted application in the trusted execution environment, it captures the face liveness detection data and face image data. The face recognition trusted application calls the liveness detection algorithm to calculate the liveness percentage based on the face liveness detection data and face image data. Among them, the face image and the face liveness detection data collected by the face recognition camera are only transmitted to the corresponding face recognition tool in the trusted execution environment through the trusted execution environment channel.

[0015] As a preferred embodiment of the above technical solution, preferably, the security unit signs and encrypts the liveness percentage and face image data, and then transmits the generated signature result and ciphertext data back to the face recognition trusted application, including: the face recognition trusted application calls the security unit interface, and the security unit performs a signature operation on the liveness percentage and the face image data to generate a signature result. The security unit calls a symmetric encryption algorithm and a corresponding key to perform an encryption operation on the liveness percentage, the signature result, and the face image data, generates ciphertext data, and transmits the ciphertext data back to the face recognition trusted application.

[0016] As a preferred embodiment of the above technical solution, preferably, the payment application determines whether the liveness percentage is greater than the preset liveness percentage threshold, and if so, performs identity authentication, including: if the collection of the face liveness detection data and face image data is completed, the liveness percentage and the generated ciphertext data are directly used to perform the payment authentication process; the payment application compares the liveness percentage with the preset liveness percentage threshold, if it is greater, the password keyboard is unlocked, and the payment application obtains the payment password. The payment application transmits the generated ciphertext data to the authentication server, and the authentication server performs identity authentication on the user. If it passes, the remaining transaction is performed, otherwise the transaction ends.

[0017] The technical solution of the present invention provides a secure method for face payment. Through a face recognition trusted application, face liveness detection data and face image data collected by a face recognition camera are obtained. After obtaining the liveness percentage, the trusted execution environment obtains ciphertext data through a security unit. The payment application determines whether the liveness percentage is greater than a preset liveness percentage threshold. If it is greater, identity authentication is performed in combination with an encryption key. After passing the authentication, the remaining transactions are performed in a rich execution environment in combination with a password keyboard.

[0018] The advantages of the present invention are:

[0019] 1. The SE-based terminal platform of the present invention can securely manage data protection keys related to the face recognition process. SE can provide secure cryptographic algorithm services. TEE ensures the secure execution of the face liveness detection algorithm. TEE is directly connected to the face recognition camera. The overall terminal platform realizes the protection of the data accuracy, integrity, authentication and confidentiality of the face liveness detection results and face recognition images.

[0020] 2. The present invention can be closely integrated with the original payment process of the payment application. The face recognition security payment method based on the terminal platform can effectively reduce the security risks introduced by the face recognition process to traditional payment transactions, and can resist malicious attacks on the terminal platform from the software layer and some hardware layers, thereby improving the security of face recognition payment as a whole.

[0021] 3. The terminal platform and payment method designed by the present invention comply with the general SE and TEE management mechanism, and cooperate with the mature TSM (Trusted Service Management Platform) and TAM (Trusted Authentication Management Platform) mechanisms, which can effectively realize the key lifecycle management and authentication of the secure payment application platform, and the lifecycle management and authentication of TEE and TA, so that the platform can be quickly integrated into the existing trusted management and authentication system environment, thereby further enhancing the underlying security management capabilities of face recognition payment transactions. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, a brief introduction is given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0023] Figure 1 A flowchart of a secure facial payment method provided by an embodiment of the present invention.

[0024] Figure 2 The specific payment process of the face payment security method provided by the embodiment of the present invention Figure 1 .

[0025] Figure 3 The specific payment process of the face payment security method provided by the embodiment of the present invention Figure 2 .

[0026] Figure 4 A structural diagram of a secure face payment method based on hardware and software implementation provided by an embodiment of the present invention Figure 1 .

[0027] Figure 5A structural diagram of a secure face payment method based on hardware and software implementation provided by an embodiment of the present invention Figure 2 . DETAILED DESCRIPTION

[0028] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0029] The technical solution of the present invention is now specifically described. Figure 1 A schematic diagram of a flow chart provided for an embodiment of the present invention, such as Figure 1 As shown, including:

[0030] Step 101: The payment application determines whether the collection of the face liveness detection data and the face image data is completed.

[0031] If not completed, the payment application calls the face recognition trusted application through the trusted execution environment, and the face recognition trusted application activates the face recognition camera to capture face data and then executes step 102. If completed, execute step 104.

[0032] Step 102: The face recognition trusted application obtains the liveness percentage through the collected face liveness detection data and face image data.

[0033] The face recognition trusted application calls the liveness detection algorithm to obtain the liveness percentage by calculation according to the face liveness detection data and the face image data;

[0034] Among them, the face image and the face liveness detection data collected by the face recognition camera are transmitted to the corresponding face recognition tool in the trusted execution environment only through the trusted execution environment channel.

[0035] Step 103: The face recognition trusted application obtains ciphertext data according to the liveness percentage and the face image data.

[0036] The face recognition trusted application in the trusted execution environment calls the security unit interface, and the security unit performs a signature operation on the liveness percentage and the face image data to generate a signature result. The security unit calls the symmetric encryption algorithm and the corresponding key to perform an encryption operation on the liveness percentage, the signature result, and the face image data to obtain ciphertext data. The encrypted content of the ciphertext data includes the face image data, the liveness percentage, and the signature result.

[0037] Step 104: The payment application determines whether the liveness percentage is greater than the preset liveness percentage threshold. If so, the ciphertext data is transmitted to the authentication server for identity authentication. If passed, the identity authentication and the remaining transaction are completed in conjunction with the password keyboard. If less, the transaction ends.

[0038] Specifically, the payment application compares the liveness percentage with the preset liveness percentage threshold. If it is greater than the liveness percentage threshold, the password keyboard is unlocked and the payment application obtains the payment password. At the same time, the payment application transmits the ciphertext data to the authentication server, and the authentication server authenticates the user. If the authentication passes, the remaining transaction is carried out, otherwise the transaction ends. When the payment application compares the liveness percentage with the preset liveness percentage threshold, the liveness percentage appears in the rich execution environment of the payment application in plain text.

[0039] The technical solution of the present invention is now further described in detail using a specific embodiment: the ID and key identification involved in the embodiment of the technical solution of the present invention are shown in Table 1 below.

[0040] Table 1: ID and key identification

[0041]

[0042]

[0043] The specific payment process of the face payment security method provided by the present invention is as follows: Figure 2 As shown, the specific description is as follows: Among them, the security unit SE, the trusted execution environment TEE, the rich execution environment REE, the trusted application TA (Trusted Application), and the payment APP are the above-mentioned payment applications, and the payment applications are installed in the operating terminals of merchants or consumers.

[0044] Step 201: A merchant or consumer operates a payment APP in the terminal platform REE to trigger a face recognition payment service.

[0045] Step 202 , the payment APP determines whether the consumer's facial image capture and liveness detection are completed. If completed, execute step 210 ; otherwise, execute step 203 .

[0046] Step 203: The payment APP activates the face recognition TA.

[0047] The face recognition function module of the payment app calls the face recognition TA in the trusted execution environment TEE through the TEE manager, sends the desired call instruction to TA, and the face recognition TA enters the active state.

[0048] Step 204: The face recognition camera captures a face image.

[0049] Among them, the face recognition TA activates the face recognition camera through the internal driver of TEE, and the face camera is ready to capture the face image according to the instruction financial activity status. The face recognition camera is usually set on the operation terminal of the merchant or consumer.

[0050] Step 205: Send the face liveness detection data l and the face image data p to the face recognition TA.

[0051] Specifically, the face recognition camera captures the consumer's face liveness detection data l and face image data p respectively, and directly transmits the two data to the face recognition TA in the TEE.

[0052] Step 206: The face recognition TA calculates the liveness percentage r.

[0053] Specifically, the face recognition TA calls the corresponding algorithm of the face liveness detection SDK to process and identify the face liveness detection data l, and obtains the liveness percentage r corresponding to the data.

[0054] Step 207: The face recognition TA generates a signature result.

[0055] Among them, the face recognition TA calls the SE interface through TEE, sends the liveness percentage r and face image data p to SE, and SE calls the signature algorithm SIG using the private key SK Face Perform signature operation on the liveness percentage r and the face image data p to obtain the signature result s, as follows:

[0056] s=SIG(SK Face ,r||p).

[0057] Step 208: SE calls the encryption algorithm to process the data and obtain the ciphertext data m.

[0058] SE calls the symmetric encryption algorithm ENC and uses the key FaceKey to perform encryption operations on the signature result s, the liveness percentage r, and the face image data p, and obtains the ciphertext data m sent to the background. Specifically, m = ENC (FaceKey, s||r||p). After SE completes the operation, it sends the ciphertext data m and the liveness percentage r to the face recognition TA of TEE. The key FaceKey is the key in the security unit SE.

[0059] Step 209: The face recognition TA receives the ciphertext data m and the liveness percentage r and sends them to the payment APP.

[0060] At this point, the payment app has completed facial image capture and liveness detection, and obtained relevant data for verifying consumer identity information, waiting to be sent to the background authentication server.

[0061] Specifically, the ciphertext data m is the encryption result of the signature result s, the liveness percentage r and the face image data p.

[0062] Step 210: The payment APP determines whether the liveness detection is passed according to the liveness percentage r. If it is passed, step 211 is executed, otherwise the transaction ends.

[0063] The payment app makes a preliminary judgment based on the value of r. If the set liveness percentage threshold is reached, the subsequent face payment transaction steps will be carried out. If the threshold requirement is not reached, the transaction will end and the payment will fail.

[0064] At this time, the liveness percentage r appears in REE, so it is an unreliable data. The judgment that it has reached the liveness threshold is only a preliminary judgment here, and the relevant data will be finally judged by the background authentication server.

[0065] Step 211, the payment APP calls the payment function module and enters the payment process.

[0066] Step 212: The consumer enters the transaction password through the password keyboard.

[0067] The payment device SDK activates the PINPAD password keyboard, and the consumer enters the password dedicated to face recognition payment transactions through the PINPAD. The securely protected consumer password is transmitted to the payment APP through the payment device SDK.

[0068] Step 213: The authentication server authenticates the consumer's identity using the received ciphertext data m and password. If the authentication is successful, step 214 is executed.

[0069] The payment APP packages the ciphertext data m and the securely protected consumer facial recognition payment transaction password, and transmits them to the backend server of the relevant organization with a biometric recognition platform to identify and authenticate the consumer's identity and map the consumer's transaction account for subsequent actual amount transactions.

[0070] Step 214, the payment APP determines whether the transaction is completed, if so, executes step 215, otherwise the transaction ends.

[0071] Specifically, in step 214, the payment APP independently determines whether the payment password is correct and whether the face matches the consumer himself. If both are correct, step 215 is executed, otherwise the transaction ends.

[0072] After completing the consumer identity authentication through facial data, the relevant server and terminal platform jointly complete the original payment transaction process, complete the actual operation of the corresponding account and the corresponding transaction amount through the traditional payment channel, and return the result to the terminal platform.

[0073] Step 215: Display the transaction results.

[0074] The payment APP determines whether the final payment transaction has been successfully completed, and displays the final transaction result on the terminal platform screen for confirmation by merchants and consumers.

[0075] The hardware implementation based on the method provided by the present invention includes: PINPAD password keyboard, security unit SE and face camera; the software implementation based on the method provided by the present invention includes: REE and TEE environment. Among them, the payment APP, payment device SDK and TEE manager are running in REE, and the payment APP includes a payment function module and a face recognition function module; the face recognition trusted application TA is running in TEE.

[0076] The payment method provided by the present invention is now described by combining hardware and software:

[0077] The payment application (payment APP) 31 is used to call up the face recognition trusted application if the collection of face liveness detection data and face image data is not completed; it is also used to transmit the ciphertext data generated by the security unit to the authentication server, authenticate the user through the authentication server, and perform the payment process.

[0078] The face recognition trusted application TA32 is used to activate the face recognition camera and obtain the liveness percentage. It contains the face liveness detection algorithm SDK.

[0079] The face recognition camera 33 is used to capture face liveness detection data and face image data and send them to the trusted execution environment. This camera is dedicated to capturing face image data during face recognition and has hardware support for liveness detection. It can be a 3D structured light, TOF camera or infrared binocular camera, driven by the corresponding algorithm in the TEE. The face recognition camera is only directly connected to the TEE, not to the REE. The generated face image raw data can only be directly transmitted to the corresponding face recognition trusted application TA in the TEE for processing.

[0080] The trusted execution environment TEE34 is used to call the symmetric encryption algorithm in the security unit SE to perform a signature operation on the liveness percentage obtained by the face recognition trusted application and the face image data collected by the face recognition camera, generate a signature result, and send an encryption instruction to the security unit to instruct the security unit to generate ciphertext data m.

[0081] The trusted execution environment 34 is implemented based on the ARM TrustZone hardware architecture and is a proprietary environment that is isolated from the REE. Here we mainly describe the TEE OS layer. The face recognition trusted application TA is securely executed in the TEE. The face recognition trusted application TA32 runs the face liveness detection algorithm SDK321, which can perform liveness detection and other related algorithms on the acquired face data. The TEE stores the TEE identity serial number ID TEE 、TEE authentication private key SK TEE and TA authentication public key PK TA_Certify ; TEE and REE communicate data through the TEE manager in REE, and TEE can also execute other security-sensitive TAs.

[0082] The security unit SE35 is used to perform encryption operation on the liveness percentage, the signature result generated by the trusted execution environment 34 and the face image data collected by the face recognition camera to obtain ciphertext data.

[0083] Specifically, the security unit SE is directly located on the platform motherboard and is used to sign the face liveness detection results, encrypt the face image, and provide secure cryptographic algorithms for the upper layer, including signatures, symmetric encryption and digest algorithms, etc. SE has a unique identity identification serial number ID SE , Face liveness detection result signature private key SK Face And the face image symmetric encryption key FaceKey. SE is directly connected to the platform's TEE, only accepts valid instructions sent by TEE, and returns the output results to the TEE for application.

[0084] The authentication server 36 is used to authenticate the current user based on the encrypted data and the biometric features stored therein, and transmit the authentication data back to the payment APP 31 .

[0085] The rich execution environment REE37 is used to run the payment application used by the payment institution for face recognition transactions. The payment application runs a payment function module and a face recognition function module, which respectively handle the conventional payment process and the face recognition process. The rich execution environment runs a payment tool to assist in completing the payment transaction, and the payment tool interacts with the password keyboard for data. It is an ordinary Android operating system that executes the application, component services and drivers of the original terminal. It runs a payment APP31 to implement the upper interface and UI of the payment transaction function and the face recognition function. The APP is mainly implemented by the payment institution. The payment function module 311 of the payment APP31 is mainly responsible for processing the request, response and data package of the payment related functions; the face recognition function module 312 is mainly responsible for processing the request, response and data package of the face recognition related functions. The payment device SDK371 is also running in REE37 to process the original payment transaction algorithm and protocol. This part is related to the background rules of the payment institution. TEE manager 372 runs in REE37 to implement data communication and command calling functions between TEE and REE. When the payment APP calls the face recognition trusted application TA, it is responsible for sending a call request and receiving the data results returned by it.

[0086] PINPAD password keyboard 38 is used by users to input payment passwords into payment applications. Specifically, it is used to safely input facial transaction passwords during facial payment. The PINPAD is a physical keyboard, not a virtual keyboard. The PINPAD needs to pass the inspection and certification of the relevant departments of the financial industry. It contains an independent security chip. When in use, it connects with the backend of the payment institution to directly obtain the encryption protection key and integrity protection password assigned to it. When the user enters the facial transaction password, the encryption and integrity protection calculations will be completed directly in the PINPAD, and then sent to the backend of the payment institution. The user password will not appear in plain text in other environments including the payment APP. The PINPAD is connected to the rich execution environment REE and mainly interacts with the payment device SDK.

[0087] The facial payment security method based on a security unit and a trusted execution environment provided by the present invention can be deployed in ordinary merchants to complete payment transactions for goods or services through facial recognition for consumers alone. It can also be used in conjunction with existing payment MIS (management information system) cash registers, self-service vending machines and POS terminals to quickly transform the original equipment to have facial recognition payment functions.

[0088] Furthermore, the above-mentioned facial recognition payment method interacts with the authentication server, TSM (Trusted Service Management Platform) and TAM (Trusted Authentication Management Platform) for relevant information and data to support and complete the facial recognition secure payment process.

[0089] The trusted service management platform TSM is used to interact with facial recognition payment methods through a secure channel, to issue, register and update the ciphertext data generated by the security unit and the remaining applications, and to protect facial image data in transactions. Specifically, TSM can be deployed in the background of the payment institution, or it can be independently operated and maintained by an institution. It is mainly used to manage the SE of the terminal platform and requires data interaction with the payment APP background service. Before the terminal platform leaves the factory, the manufacturer presets the key for SE, and a secure channel can be established between TSM and the terminal platform. During the use of the terminal platform, the Applet applications and function keys run by SE can be sent to SE by TSM through a secure channel, and the update and upgrade of these data are also sent through TSM. The SE keys managed by TSM are FaceKey and SK Face When a terminal SE completes registration at TSM, TSM will have the ID corresponding to SE SE , FaceKey and PK Face , TSM transmits this data in some form and through some means to the authentication server. The authentication server will use this data as the basis for verification when verifying the face payment transaction on the terminal platform.

[0090] A trusted authentication management platform is used to interact with the face recognition payment method through a secure channel, to issue, register and update the keys of the trusted execution environment and the face recognition trusted applications in the trusted execution environment, and to authenticate the identity of the trusted execution environment.

[0091] TAM can be deployed in the background of the payment institution or independently operated and maintained by an institution. It is mainly used to manage the TEE of the terminal platform and needs to interact with the payment APP background service. Before the terminal platform leaves the factory, the manufacturer presets the key for TEE, and a secure channel can be established between TAM and the terminal platform. During the use of the terminal platform, the TEE OS image, the TA running in TEE, and the functional keys can all be sent to the terminal platform by TAM through a secure channel. At the same time, the update and upgrade of these data are also sent through TAM. The TEE keys managed by TAM include SK TEE and PK TA_Certify When a terminal TEE completes registration at TAM, TAM will have the ID corresponding to the TEE TEE ,PK TEE and SK TA_Certify , TAM will ID TEE and PK TEE The two data are transmitted to the authentication server in some form and by some means. The authentication server will use these two data as the basis for verification when verifying the TEE identity of the terminal platform.

[0092] The authentication server is mainly deployed in the background of the payment institution to verify the legal identity of the terminal platform and the consumer. It is the main body for deploying the background face recognition algorithm during the face payment transaction process, and is also the core component for authenticating the consumer account of the face payment transaction. In the face recognition security payment process, the terminal platform (payment APP) generates face recognition package data and sends it to the authentication server. The authentication server uses the corresponding key data transmitted by TSM to unpack and verify the terminal platform data. After making a qualified judgment on the live percentage r, the face image data p and the consumer's face transaction password are used to compare and retrieve in the own face image database to determine the specific identity and consumption account of the corresponding consumer, and then the payment transaction process of the specific amount is completed through the traditional path, thereby completing the core verification step of the face recognition payment transaction and realizing the face recognition security payment application method function.

[0093] The technical solution of the present invention is now described in combination with the actual operation process.

[0094] The consumer starts the payment application 31 in the terminal device to start payment. The payment application 31 activates the TEE manager 372, thereby calling the face recognition trusted application TA32 in the trusted execution environment TEE34. The face recognition trusted application TA32 activates the face recognition camera 33 to collect the consumer's face image and transmits the image back to the face recognition trusted application TA32. The face liveness detection algorithm SDK321 calculates the face image to obtain the liveness percentage r and the face pixel image. The face recognition trusted application TA32 sends the liveness percentage r and the face pixel image to the security unit SE35 for calculation to obtain the ciphertext data m. The ciphertext data m is then transmitted back to the face recognition trusted application TA32, and then transmitted back to the payment application 31 after passing through the TEE manager 372. The payment application sends the ciphertext data m to the authentication server 36 for authentication, and then transmits the authentication result back to the payment application 31. If the authentication is successful, the password is entered on the password keyboard 38, and then the payment device SDK371 calculates the result and sends it to the payment application 31. The payment function module 311 in the payment application 31 executes the payment process.

[0095] The technical solution of the present invention provides a secure method for face payment. Through a face recognition trusted application, face liveness detection data and face image data collected by a face recognition camera are obtained. After obtaining the liveness percentage, the trusted execution environment obtains ciphertext data through a security unit. The payment application determines whether the liveness percentage is greater than a preset liveness percentage threshold. If it is greater, identity authentication is performed in combination with an encryption key. After passing the authentication, the remaining transactions are performed in a rich execution environment in combination with a password keyboard.

[0096] The advantages of the present invention are:

[0097] 1. The SE-based terminal platform of the present invention can securely manage data protection keys related to the face recognition process. SE can provide secure cryptographic algorithm services. TEE ensures the secure execution of the face liveness detection algorithm. TEE is directly connected to the face recognition camera. The overall terminal platform realizes the protection of the data accuracy, integrity, authentication and confidentiality of the face liveness detection results and face recognition images.

[0098] 2. The present invention can be closely integrated with the original payment process of the payment application. The face recognition security payment method based on the terminal platform can effectively reduce the security risks introduced by the face recognition process to traditional payment transactions, and can resist malicious attacks on the terminal platform from the software layer and some hardware layers, thereby improving the security of face recognition payment as a whole.

[0099] 3. The terminal platform and payment method designed by the present invention comply with the general SE and TEE management mechanism, and cooperate with the mature TSM (Trusted Service Management Platform) and TAM (Trusted Authentication Management Platform) mechanisms, which can effectively realize the key lifecycle management and authentication of the secure payment application platform, and the lifecycle management and authentication of TEE and TA, so that the platform can be quickly integrated into the existing trusted management and authentication system environment, thereby further enhancing the underlying security management capabilities of face recognition payment transactions.

[0100] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A secure face payment method based on a secure unit and a trusted execution environment, characterized in that: include: The payment application, face recognition trusted application, face recognition camera, security unit, trusted execution environment, authentication server, rich execution environment, password keyboard, trusted service management platform, and trusted authentication management platform work together to achieve: A payment application, which is used to activate the TEE manager if the collection of face liveness detection data and face image data is not completed. The payment application calls the face recognition trusted application in the trusted execution environment and is also used to implement the payment process; The face recognition trusted application is used to activate the face recognition camera and to collect face liveness detection data l and face image data p from the face recognition camera directly connected to the TEE, and directly transfer the two data to the face recognition trusted application in the TEE. The face recognition trusted application calculates the liveness percentage r. The interface used to call the security unit through TEE, send the liveness percentage r and face image data p to the security unit, and the security unit calls the signature algorithm SIG using the private key SK Face Perform signature operation on the liveness percentage r and the face image data p to obtain a signature result s; Used to send the received ciphertext data m and liveness percentage r to the payment application. At this time, the payment application has completed the capture of the face image and liveness detection, and obtained the ciphertext data m for verifying the consumer's identity information, waiting to be sent to the background authentication server; A face recognition camera, used to capture the face liveness detection data and face image data, and send them to the trusted execution environment; The security unit is used to perform encryption operation on the liveness percentage r, the signature result s generated by the trusted execution environment, and the face image data p collected by the face recognition camera by calling the encryption algorithm, generate ciphertext data m and return it to the trusted execution environment; The payment application is further used to transmit the ciphertext data generated by the security unit to an authentication server, and perform identity authentication on the user through the authentication server. The trusted execution environment is used to call the signature algorithm in the security unit to use the private key to perform a signature operation on the liveness percentage obtained by the face recognition trusted application and the face image data collected by the face recognition camera to generate a signature result; and used to call the symmetric encryption algorithm in the security unit to use the encryption key to perform encryption operation on the liveness percentage, the signature result and the face image data, send an encryption instruction to the security unit, and generate ciphertext data; The authentication server is used to authenticate the current user based on the encrypted data, payment password and stored biometric features; A rich execution environment is used to run a payment application used by a payment institution for face recognition transactions. The payment application runs a payment function module and a face recognition function module, which respectively handle conventional payment processes and face recognition processes. A payment tool runs in the rich execution environment to assist in completing payment transactions. The payment tool interacts with the PIN pad for data. The password keyboard is used by the user to enter the payment password into the payment application; The trusted service management platform is used to exchange data with the face recognition payment platform through a secure channel, to issue, register and update the security unit related keys and remaining applications, and to protect the face image data in the transaction; the security unit related keys managed by the trusted service management platform include the face image security unit encryption key FaceKey and the liveness recognition result security unit signature private key SK Face After the terminal's security unit completes registration on the trusted service management platform, the trusted service management platform will register the security unit with the ID serial number SE , FaceKey, the encryption key of the face image security unit, and PK, the signature verification public key of the liveness recognition result security unit Face , transmitted to the authentication server, for verification by the authentication server when verifying the face payment transaction on the terminal platform; A trusted authentication management platform, used to exchange data with the face recognition payment platform through a secure channel, to issue, register and update the keys of the trusted execution environment and the face recognition trusted applications in the trusted execution environment, and to authenticate the identity of the trusted execution environment; The payment application is also used to determine whether the liveness percentage is greater than a preset liveness percentage threshold. If so, identity authentication is performed and the remaining transactions are carried out after passing.

2. The face payment security method based on a security unit and a trusted execution environment according to claim 1 is characterized in that: The face recognition trusted application in the trusted execution environment includes: The payment application determines whether the collection of the face liveness detection data and the face image data is completed; If not completed, the payment application calls the face recognition trusted application through the trusted execution environment; The face recognition trusted application activates the face recognition camera to capture face data.

3. The face payment security method based on a security unit and a trusted execution environment according to claim 2 is characterized in that: The payment application determines whether the liveness percentage is greater than a preset liveness percentage threshold, and if so, performs identity authentication, including: If the collection of the face liveness detection data and the face image data is completed, the payment authentication process is directly executed using the liveness percentage and the generated ciphertext data; The payment application compares the liveness percentage with a preset liveness percentage threshold, and if the liveness percentage is greater than the liveness percentage threshold, the password keyboard is unlocked, and the payment application obtains the payment password; The payment application transmits the generated ciphertext data to the authentication server, and the authentication server performs identity authentication on the user. If the identity authentication is successful, the remaining transaction is performed, otherwise the transaction ends.

Images