Payment verification method, device, terminal, server and storage medium

Abstract

The embodiment of the present application discloses a payment verification method, device, terminal, server and storage medium, which belongs to the field of contactless payment technology. The method includes: obtaining a palm print image and a face image; determining at least two first candidate accounts based on the palm print image, and determining at least two second candidate accounts based on the face image; determining a first false recognition rate corresponding to the first candidate account based on the palm print similarity of the first candidate account, and determining a second false recognition rate corresponding to the second candidate account based on the face similarity corresponding to the second candidate account; determining a target account from the first candidate account and the second candidate account based on the first false recognition rate and the second false recognition rate; and making payment based on the target account. It can avoid the situation where one biometric feature is always emphasized, and there is no need to introduce additional parameters, thereby improving the accuracy of palm-face payment.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of contactless payment technology, and in particular to a payment verification method, device, terminal, server and storage medium. Background Art

[0002] Nowadays, contactless payment systems based on face recognition have been widely used. Since the current face recognition model still has difficulty distinguishing similar faces (such as identical twins), in order to solve this problem and maintain the convenience of biometrics, the current stage usually adopts the method of integrating palm features for identity recognition.

[0003] In related technologies, the system collects multiple feature recognition results of the user, including face recognition results, palm print recognition results, and palm vein recognition results. When the similarities corresponding to the three features exceed a threshold, the recognition is determined to be successful and payment is made. Alternatively, the three recognition results are fused in a weighted manner, or the three feature vectors are fused and identity recognition is performed based on the fused features.

[0004] However, identity confirmation based on three thresholds will result in a low pass rate, easily causing problems with repeated operations for payment, and low payment efficiency; the weighted calculation method of recognition results leads to over-reliance on a certain biometric feature and does not increase the accuracy of the recognition results; and the method that relies on feature fusion will introduce more hyperparameters, resulting in the fused feature not necessarily being stronger than a single biometric feature. Summary of the invention

[0005] The embodiments of the present application provide a payment verification method, device, terminal, server and storage medium, which can improve the accuracy of palm and face payment results. The technical solution is as follows:

[0006] On the one hand, an embodiment of the present application provides a payment verification method, the method comprising:

[0007] Obtain palmprint images and face images;

[0008] Determine at least two first candidate accounts based on the palm print image, and determine at least two second candidate accounts based on the face image, wherein the palm print similarity corresponding to the first candidate accounts is higher than the palm print similarity corresponding to the other accounts, and the face similarity corresponding to the second candidate accounts is higher than the face similarity corresponding to the other accounts;

[0009] Determine a first misrecognition rate corresponding to the first candidate account based on the palm print similarity of the first candidate account, and determine a second misrecognition rate corresponding to the second candidate account based on the face similarity corresponding to the second candidate account;

[0010] Determining a target account from the first candidate account and the second candidate account based on the first false recognition rate and the second false recognition rate;

[0011] Payment is made based on the target account.

[0012] On the other hand, an embodiment of the present application provides a payment verification method, the method comprising:

[0013] Collect palm print images and face images through cameras;

[0014] Sending the palm print image and the face image to a server, the server being configured to determine at least two first candidate accounts based on the palm print image and at least two second candidate accounts based on the face image, determining a target account based on a first false recognition rate corresponding to the first candidate account and a second false recognition rate corresponding to the second candidate account, and making payment based on the target account;

[0015] The payment result sent by the server is received, and a payment result interface is displayed based on the payment result, wherein the payment result is used to indicate whether the payment of the account to be paid is successful or failed.

[0016] On the other hand, an embodiment of the present application provides a payment verification device, the device comprising:

[0017] An acquisition module, used for acquiring palm print images and face images;

[0018] A first determination module is used to determine at least two first candidate accounts based on the palm print image, and to determine at least two second candidate accounts based on the face image, wherein the palm print similarity corresponding to the first candidate accounts is higher than the palm print similarity corresponding to the other accounts, and the face similarity corresponding to the second candidate accounts is higher than the face similarity corresponding to the other accounts;

[0019] a second determination module, configured to determine a first misrecognition rate corresponding to the first candidate account based on the palm print similarity of the first candidate account, and to determine a second misrecognition rate corresponding to the second candidate account based on the face similarity corresponding to the second candidate account;

[0020] a third determination module, configured to determine a target account from the first candidate account and the second candidate account based on the first false recognition rate and the second false recognition rate;

[0021] A payment module is used to make payment based on the target account.

[0022] On the other hand, an embodiment of the present application provides a payment verification device, the device comprising:

[0023] An image acquisition module, used for acquiring palm print images and face images through a camera;

[0024] a sending module, configured to send the palm print image and the face image to a server, wherein the server is configured to determine at least two first candidate accounts based on the palm print image and at least two second candidate accounts based on the face image, determine a target account based on a first false recognition rate corresponding to the first candidate account and a second false recognition rate corresponding to the second candidate account, and make payment based on the target account;

[0025] The display module is used to receive the payment result sent by the server and display a payment result interface based on the payment result, wherein the payment result is used to indicate whether the payment of the account to be paid is successful or failed.

[0026] On the other hand, an embodiment of the present application provides a terminal, which includes a processor and a memory; the memory stores at least one instruction, at least one program, a code set or an instruction set, and the at least one instruction, the at least one program, the code set or the instruction set are loaded and executed by the processor to implement the payment verification method as described in the above aspects.

[0027] On the other hand, an embodiment of the present application provides a server, comprising a processor and a memory; the memory stores at least one instruction, at least one program, a code set or an instruction set, and the at least one instruction, the at least one program, the code set or the instruction set are loaded and executed by the processor to implement the payment verification method as described in the above aspects.

[0028] On the other hand, an embodiment of the present application provides a computer-readable storage medium, in which at least one computer program is stored. The computer program is loaded and executed by a processor to implement the payment verification method as described in the above aspects.

[0029] According to one aspect of the present application, a computer program product or computer program is provided, the computer program product or computer program includes computer instructions, the computer instructions are stored in a computer-readable storage medium. A processor of a computer device or terminal reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device or terminal performs the payment verification method provided in various optional implementations of the above aspects.

[0030] The technical solution provided by the embodiment of the present application includes at least the following beneficial effects:

[0031] In an embodiment of the present application, based on the preliminary tests, the false recognition rates of palmprint recognition results and face recognition results are respectively obtained. During payment verification, based on the similarity between each candidate account and the acquired palmprint image and face image, multiple first candidate accounts with the highest palmprint similarity and multiple second candidate accounts with the highest face similarity are selected. Based on the false recognition rates corresponding to each candidate account, the target account to be paid is determined. Compared with feature fusion or weighted calculation methods, this can avoid always focusing on one feature, and the target account is determined based on the false recognition rate without introducing additional parameters. This can also make the result more realistic, so as to improve the accuracy of palm and face payment. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 is a schematic diagram of an implementation environment provided by an exemplary embodiment of the present application;

[0033] Figure 2 is a flow chart of a payment verification method provided by an exemplary embodiment of the present application;

[0034] Figure 3 is a flow chart of a payment verification method provided by another exemplary embodiment of the present application;

[0035] Figure 4 is a schematic diagram of a first mapping table provided by an exemplary embodiment of the present application;

[0036] Figure 5 is a schematic diagram of a second mapping table provided by an exemplary embodiment of the present application;

[0037] Figure 6 is a schematic diagram of a score mapping table provided by an exemplary embodiment of the present application;

[0038] Figure 7 is a flow chart of a payment verification method provided by another exemplary embodiment of the present application;

[0039] Figure 8 is a flow chart of a payment verification method provided by another exemplary embodiment of the present application;

[0040] Fig. 9 is a flow chart of a payment verification method provided by another exemplary embodiment of the present application;

[0041] Fig.10 is a flow chart of a payment verification method provided by another exemplary embodiment of the present application;

[0042] Fig.11 is a schematic diagram of image preprocessing of a palm image provided by an exemplary embodiment of the present application;

[0043] Fig.12 is a schematic diagram of image preprocessing of a facial image provided by an exemplary embodiment of the present application;

[0044] Fig.13 is a schematic diagram of palm print images of different qualities provided by an exemplary embodiment of the present application;

[0045] Fig.14 is a structural block diagram of a payment verification device provided by an exemplary embodiment of the present application;

[0046] Fig.15 is a structural block diagram of a payment verification device provided by an exemplary embodiment of the present application;

[0047] Fig.16 is a structural block diagram of a server provided by an exemplary embodiment of the present application;

[0048] Fig.17 It is a structural block diagram of a terminal provided by an exemplary embodiment of the present application. DETAILED DESCRIPTION

[0049] In order to make the objectives, technical solutions and advantages of the present application clearer, the implementation methods of the present application will be further described in detail below with reference to the accompanying drawings.

[0050] The term "multiple" as used herein refers to two or more than two. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the related objects are in an "or" relationship.

[0051] In the related art, the contactless payment method based on palm print features and facial features mainly includes three implementation methods. One is to set a palm print similarity threshold, a facial similarity threshold and a palm vein threshold. When the recognition results of the three types of biometric information meet the corresponding thresholds, the recognition is deemed to be successful. However, this method will result in an extremely low pass rate and easy verification failure, resulting in the user needing secondary verification or switching to contact payment; the second is to obtain multiple biometric recognition results, namely face recognition results, palm print recognition results, and palm vein recognition results. When the three recognition results conflict (for example, the account numbers recognized by the palm print recognition results and the face recognition results are inconsistent), one of the biometric recognition results shall prevail (such as the face recognition result). This setting method is too dependent on one of the biometric recognition results and will not bring gains to the recognition accuracy; the third is to generate a biometric feature vector and use a specific method to weightedly fuse the three vectors. However, this method relies on feature fusion, which will introduce more hyperparameters, and it is still unknown whether the fused feature is stronger than a single biometric feature.

[0052] In order to solve the problems existing in the above-mentioned payment verification methods and improve the accuracy and efficiency of payment verification, the present application provides a payment verification method. Figure 1 A schematic diagram of an implementation environment provided by an embodiment of the present application is shown. The implementation environment may include: a payment terminal 110, a server 120 and an account registration terminal 130.

[0053] The payment terminal 110 is installed and runs an application 111 that supports contactless payment. When the payment terminal 110 runs the application 111, the payment terminal 110 turns on the camera to collect the user's palm print image and face image in real time, and sends the palm print image and face image to the server 120 for payment verification. When receiving the payment result sent by the server 120, it is displayed to inform the user whether the payment is successful. The application 111 can be an application that supports contactless payment, such as a shopping application, a payment application, and a financial transaction application.

[0054] The account registration terminal 130 has installed and runs an application 131 that supports contactless payment. Optionally, the application installed on the payment terminal 110 and the account registration terminal 130 is the same, or the application installed on the two terminals is the same type of application on different operating system platforms (Android or IOS), or the application installed on the two terminals is a different version of the same type of application (for example, an application that supports payment and an application that supports collection). When the account registration terminal 130 receives an account registration operation, it turns on the camera to capture the face image and palm print image of the user 132, and sends the captured image and the registered account to the server 120, so that the server 120 updates the user account database 123, which stores the user account and the corresponding face features and palm print features.

[0055] Figure 1 Only one payment terminal and one account registration terminal are shown, but in different embodiments, there are multiple other terminals that can access the server 120. Optionally, there are one or more terminals corresponding to the developer, on which a development and editing platform for an application supporting contactless payment is installed. The developer can edit and update the application on the terminal, and transmit the updated application installation package to the server 120 via a wired or wireless network. The first terminal 110 and the second terminal 130 can download the application installation package from the server 120 to update the application.

[0056] The payment terminal 110 , the account registration terminal 130 and other terminals are connected to the server 120 via a wireless network or a wired network.

[0057] The server 120 includes at least one of a single server, a server cluster consisting of multiple servers, a cloud computing platform, and a virtualization center. In an illustrative example, the server 120 includes a memory 121, a processor 122, a user account database 123, a payment verification module 124, and an input / output interface (I / O interface) 125 for users. Among them, the processor 122 is used to load the instructions stored in the server 120, and process the data in the user account database 123 and the payment verification module 124; the user account database 123 is used to store the data of the user account used by the account registration terminal 130 and other terminals, and the data includes the user account and the corresponding facial features and palm print features; the payment verification module 124 is used to determine at least two first candidate accounts based on the palm print image and at least two second candidate accounts based on the facial image when the facial features and palm print features are obtained and it is determined that payment verification is required, determine the target account based on the first false recognition rate corresponding to the first candidate account and the second false recognition rate corresponding to the second candidate account, and make payment with the target account; the user-oriented I / O interface 125 is used to establish communication and exchange data with the payment terminal 110 and / or the account registration terminal 130 through a wireless network or a wired network.

[0058] Figure 2 A flow chart of a payment verification method provided by an exemplary embodiment of the present application is shown. This embodiment is described by taking the method applied to a backend server of an application supporting a contactless payment function as an example, and the method includes the following steps.

[0059] Step 201, obtaining a palm print image and a face image.

[0060] In a possible implementation, a payment terminal is provided in the merchant. When a customer needs to make a payment, the merchant controls the payment terminal to start collecting information. The customer needs to show his palm and face so that the payment terminal can collect the palm print image and the face image and send them to the server for payment verification. The server receives the data sent by the payment terminal and obtains the palm print image and the face image.

[0061] Optionally, the payment terminal continuously collects images at a preset frequency, and sends the palm print images and face images obtained after optimization to the server by screening the collected images.

[0062] Step 202: determine at least two first candidate accounts based on the palm print image, and determine at least two second candidate accounts based on the face image.

[0063] Among them, the palm print similarity corresponding to the first candidate account is higher than the palm print similarities corresponding to other accounts, and the face similarity corresponding to the second candidate account is higher than the face similarities corresponding to other accounts.

[0064] In a possible implementation, a feature library is stored in the server, and the feature library contains feature data of all accounts of the payment application, that is, the correspondence between the account and the facial features and palm print features. For example, the first feature library contains the palm print features corresponding to the account, and the second feature library contains the facial features corresponding to the account. When the server obtains the palm print image and the facial image and performs payment verification, it extracts features from the palm print image and the facial image respectively, and then performs feature comparison between the palm print features corresponding to the palm print image and the palm print features in the first feature library to obtain the palm print similarity corresponding to each account, and performs feature comparison between the facial features corresponding to the facial image and the facial features in the second feature library to obtain the facial similarity corresponding to each account.

[0065] Since palm prints and faces are two independent biometric features, the server compares the features of the two separately, and thus also determines the corresponding candidate accounts separately. The result of the palm print feature comparison may be different from the result of the face feature comparison, that is, the account with the highest similarity may not be the same, so the server needs to determine at least two first candidate accounts and at least two second candidate accounts respectively, and determine a target account for payment.

[0066] For example, the server extracts features from the palmprint image, performs feature comparison based on the extracted features and the palmprint features in the first feature library, calculates the palmprint similarity (e.g., cosine similarity) corresponding to each account, and determines the n accounts with the highest palmprint similarity as the first candidate accounts, where n is an integer greater than 1, and determines the second candidate account in the same way.

[0067] Step 203, based on the palm print similarity of the first candidate account, determine the first misrecognition rate corresponding to the first candidate account, and based on the face similarity corresponding to the second candidate account, determine the second misrecognition rate corresponding to the second candidate account.

[0068] In one possible implementation, the correspondence between palmprint similarity and the first false recognition rate, and the correspondence between face similarity and the second false recognition rate are stored during the service period, so that the first false recognition rate corresponding to the first candidate account is determined based on the palmprint similarity of the first candidate account, and the second false recognition rate corresponding to the second candidate account is determined based on the face similarity corresponding to the second candidate account.

[0069] The first false recognition rate is the probability that the palm print similarity meets the account screening conditions and does not belong to the account corresponding to the palm print image; the second false recognition rate is the probability that the face similarity meets the account screening conditions and does not belong to the account corresponding to the face image. For example, the first false recognition rate of the first candidate account A is 0.002, which means that the probability of error in determining account A as the account corresponding to the palm print image based on the palm print features is 0.002.

[0070] Step 204 : determining a target account from the first candidate account and the second candidate account based on the first false recognition rate and the second false recognition rate.

[0071] In order to improve the accuracy of payment verification, the server does not directly determine the target account based on the first false recognition rate or the second false recognition rate, but determines the comprehensive score of each candidate account based on the first false recognition rate and the second false recognition rate of each candidate account, and then obtains the target account. Among them, there may be some candidate accounts that are only determined as the first candidate account or the second candidate account. The server cannot directly obtain the two false recognition rates of such accounts, but scores them based on the preset false recognition rates.

[0072] For example, the first candidate account includes account A, account B and account C, and the second candidate account includes account B, account C and account D. For account B and account C, the server can directly obtain the first misrecognition rate and the second misrecognition rate. For account A, it can only directly obtain its first misrecognition rate based on the corresponding relationship. Since account A has a low facial similarity and is not determined as the second candidate account, the server determines the preset misrecognition rate as the second misrecognition rate of account A. The preset misrecognition rate is high (for example, 1), which means that if account A is determined as the target account based on facial features, the probability of the recognition result being wrong is high. Similarly, the server determines the preset misrecognition rate as the first misrecognition rate of account C.

[0073] Step 205: Make payment based on the target account.

[0074] After the server determines the target account, it makes payment through the target account based on the payment amount, payment method and other data contained in the payment request data sent by the payment terminal, and returns the payment result to the payment terminal.

[0075] To summarize, in the embodiments of the present application, the false recognition rates of palmprint recognition results and face recognition results are obtained respectively through preliminary tests. During payment verification, based on the similarity between each candidate account and the acquired palmprint image and face image, multiple first candidate accounts with the highest palmprint similarity and multiple second candidate accounts with the highest face similarity are selected. Based on the false recognition rates corresponding to each candidate account, the target account to be paid is determined. Compared with feature fusion or weighted calculation, this method can avoid always focusing on one feature, and the target account is determined based on the false recognition rate without introducing additional parameters. This can also make the result more realistic, so as to improve the accuracy of palm and face payment.

[0076] Figure 3 A flow chart of a payment verification method provided by another exemplary embodiment of the present application is shown. This embodiment is described by taking the method applied to a backend server of an application supporting a contactless payment function as an example, and the method includes the following steps.

[0077] Step 301, obtaining a palm print image and a face image.

[0078] Step 302, determining at least two first candidate accounts based on the palm print image, and determining at least two second candidate accounts based on the face image.

[0079] The specific implementation of step 301 to step 302 can refer to the above-mentioned step 201 to step 202, and the embodiment of the present application will not be repeated here.

[0080] Step 303, querying the first mapping table to determine the first similarity interval corresponding to the palm print similarity of the first candidate account, wherein the first mapping table contains a correspondence between the similarity interval and the first false recognition rate.

[0081] The server stores a first mapping table obtained based on a false recognition rate test. The server determines a first similarity interval to which the palm print similarity of the candidate account belongs, thereby determining its first false recognition rate based on the first mapping table.

[0082] In another possible implementation, the first mapping table stores the corresponding relationship between the palm print similarity threshold and the first false recognition rate, which means that when a certain palm print similarity threshold is used for payment verification, an account with a palm print similarity higher than the palm print similarity threshold is determined as a target account, but the probability that the account does not belong to the account corresponding to the acquired palm print image, that is, the probability of an incorrect recognition result. Therefore, there is a negative correlation between the palm print similarity threshold and the first false recognition rate, that is, the lower the palm print similarity threshold, the more accounts that meet the screening conditions, and the greater the probability that the server determines the wrong account as the target account.

[0083] Indicative, Figure 4 A first mapping table is shown, which contains first misrecognition rates corresponding to three palm print similarity thresholds, and the first misrecognition rates are expressed by order of magnitude. For example, when 0.85 is used as the palm print similarity threshold, the probability that the recognition result is wrong, that is, the account that meets the palm print similarity threshold but does not actually belong to the acquired palm print image is 10^(-3).

[0084] The acquisition process of the first mapping table is as follows: select x (for example, 10,000) accounts as test accounts; establish a first test set and a second test set, wherein the first test set consists of a registered palm print image corresponding to each test account, a total of x images, and the second test set consists of y test palm print images corresponding to each test account, a total of xy images (for example, 10,000*3 images, y is a positive integer); calculate the palm print similarity between each test palm print image in the second test set and each registered palm print image in the first test set, that is, for each test palm print image An image, which corresponds to a pair of positive sample pairs (accounts are consistent) and (x-1) pairs of negative sample pairs (accounts are inconsistent), and each sample pair corresponds to a palmprint similarity, with a total of xy pairs of positive sample pairs and x*(xy-1) pairs of negative sample pairs; a palmprint similarity threshold is set for a negative sample set composed of negative sample pairs, a first false recognition rate corresponding to the palmprint similarity threshold is determined based on the number of negative sample pairs that meet the palmprint similarity threshold and the total number of negative sample pairs, different palmprint similarity thresholds are selected respectively, the corresponding first false recognition rates are obtained, and a first mapping table is established.

[0085] Step 304: determine a first misrecognition rate based on the first similarity interval and the first mapping table.

[0086] The server determines a first misrecognition rate corresponding to each first candidate account based on the first mapping table.

[0087] Step 305, query the second mapping table to determine the second similarity interval corresponding to the face similarity of the second candidate account, and the second mapping table contains the corresponding relationship between the similarity interval and the second false recognition rate.

[0088] Step 306: Determine a second misrecognition rate based on the second similarity interval and the second mapping table.

[0089] Correspondingly, similar to the determination process of the first false recognition rate, the server determines the second false recognition rate of the second candidate account in the same way. The second mapping table stores the correspondence between the face similarity threshold and the second false recognition rate, which means that when a certain face similarity threshold is used for payment verification, an account with a face similarity higher than the face similarity threshold is determined as the target account, but the probability that the account does not belong to the account corresponding to the acquired face image, that is, the probability that the recognition result is wrong. Therefore, there is a negative correlation between the face similarity threshold and the second false recognition rate, that is, the lower the face similarity threshold, the more accounts that meet the screening conditions, and the greater the probability that the server determines the wrong account as the target account.

[0090] Indicative, Figure 5 A second mapping table is shown, which contains second false recognition rates corresponding to four face similarity thresholds, and the second false recognition rates are expressed by order of magnitude. For example, when 0.8 is used as the face similarity threshold, the probability that the recognition result is wrong, that is, the account that meets the face similarity threshold but does not actually belong to the acquired face image is 10^(-4).

[0091] The process of formulating the second mapping table is similar to that of the first mapping table, and will not be repeated herein in the embodiment of the present application.

[0092] Step 307 , determining the account misrecognition rate of the candidate account based on the first misrecognition rate, the second misrecognition rate, and the account type of the candidate account.

[0093] Among them, the account types include intersection accounts and non-intersection accounts, and the intersection accounts belong to both the first candidate account and the second candidate account. The account misrecognition rate is used to indicate the probability that the palm print similarity and face similarity of the candidate account meet the account screening conditions and do not belong to the account corresponding to the palm print image and the face image. For example, the first candidate account determined by the server includes account A, account B and account C, and the second candidate account includes account A, account B and account D, then account A and account B belong to the intersection account, and account C and account D belong to the non-intersection account.

[0094] For the intersection account, the server can directly obtain its first false recognition rate and second false recognition rate, while for the non-intersection account, since the similarity of a certain feature is low, it is not determined as a candidate account corresponding to the feature. In a possible implementation, step 307 includes the following steps:

[0095] Step 307a, in response to the candidate account being an intersection account, determining the account misrecognition rate of the candidate account based on the first misrecognition rate and the second misrecognition rate corresponding to the candidate account.

[0096] In one possible implementation, the server determines the account misrecognition rate of the candidate account based on the first misrecognition rate and the second misrecognition rate. For example, the server determines the account misrecognition rate as the product of the first misrecognition rate and the second misrecognition rate. The account misrecognition rate is used to indicate the probability that when the palm print similarity corresponding to the candidate account satisfies the adopted palm print similarity threshold and the face similarity satisfies the adopted face similarity threshold, the candidate account does not belong to the account corresponding to the acquired palm print image and face image.

[0097] Step 307b, in response to the candidate account being a non-intersection account and belonging to the first candidate account, determining the second misrecognition rate of the candidate account to be 1; and determining the account misrecognition rate of the candidate account based on the first misrecognition rate and the second misrecognition rate corresponding to the candidate account.

[0098] Step 307c, in response to the candidate account being a non-intersection account and belonging to the second candidate account, determining the first misrecognition rate of the candidate account to be 1; and determining the account misrecognition rate of the candidate account based on the first misrecognition rate and the second misrecognition rate corresponding to the candidate account.

[0099] Since the similarity of non-intersection accounts corresponding to a certain feature is low and they are not determined as candidate accounts corresponding to the feature, in order to make the screening of target accounts more stringent and avoid the emergence of multiple candidate accounts with low account misrecognition rates and high verification scores, the server directly determines the first misrecognition rate or the second misrecognition rate of non-intersection accounts as 1, that is, 10^0.

[0100] In a possible implementation, the first misrecognition rate and the second misrecognition rate are expressed by orders of magnitude. Due to defects in the feature extraction model, errors in similarity calculation, unclear palm print images or facial images obtained, etc., the actual account misrecognition rate of the server is usually higher than the account misrecognition rate obtained by theoretical calculation, that is, there is a certain loss. The above step of "determining the account misrecognition rate of the candidate account based on the first misrecognition rate and the second misrecognition rate corresponding to the candidate account" can also specifically include the following steps:

[0101] Step 1: Determine a target misrecognition rate for the candidate account based on the product of the first misrecognition rate and the second misrecognition rate.

[0102] Step 2: Determine the account misrecognition rate based on the target misrecognition rate and the order of magnitude loss. The order of magnitude loss is the difference in order of magnitude between the misrecognition rate obtained by theoretical calculation and the misrecognition rate obtained by actual testing.

[0103] Ideally, the distribution of facial features and palm print features in space is independent of each other, so the first misrecognition rate and the second misrecognition rate can be directly multiplied to calculate the comprehensive account misrecognition rate corresponding to the candidate account. However, in actual application scenarios, there are intersecting parts in the misrecognition space of faces and palm prints. In order to ensure the authenticity and reliability of the misrecognition rate calculation results, the developers calculated the gap between the actual misrecognition rate and the theoretical misrecognition rate, determined the order of magnitude loss, and enabled the server to determine the account misrecognition rate of the candidate account based on the target misrecognition rate and the order of magnitude loss.

[0104] Based on the example in step 303, the calculation process of the order of magnitude loss is as follows: for x samples in the sample space, when the error recognition rate corresponding to the palm print similarity threshold T1 is 10^(-a), and the error recognition rate corresponding to the face similarity threshold T2 is 10^(-b), it is calculated that the theoretical error recognition rate should be 10^[-(a+b)]. Then, among the x samples, the number of samples num that are theoretically misidentified is theoretical It should be x*10^[-(a+b)]; then, based on the palmprint recognition technology and the corresponding palmprint similarity threshold T1, filter out the palmprint misidentification samples, the number of which is num face Based on this batch of palmprint misidentification samples, we use face recognition technology and the corresponding face similarity threshold T2 to filter out the samples of face misidentification, the number of which is num face+palm , then the order of magnitude loss waste can be calculated as 10^waste=num face+palm / num theoretical , the calculated loss value may not be an integer. In this case, the rounding function ceil() is set, and the order of magnitude loss corresponding to the threshold combination T1 and T2 is finally determined to be ceil(waste). By calculating the order of magnitude loss corresponding to each threshold combination, the loss with the largest value is determined as the order of magnitude loss of the final system, so that the selection of target accounts is more stringent.

[0105] Illustratively, the server determines the first candidate account, including account A, with a palm print similarity of 0.9, account B, with a palm print similarity of 0.7, and account C, with a palm print similarity of 0.7; the second candidate account includes account B, with a face similarity of 0.8, account A, with a face similarity of 0.8, and account D, with a face similarity of 0.7. The server determines that the first misrecognition rate of account A is 10^(-3), the first misrecognition rate of account B is 10^(-2), the first misrecognition rate of account C is 10^(-1), the second misrecognition rate of account A is 10^(-4), the second misrecognition rate of account B is 10^(-4), and the second misrecognition rate of account D is 10^(-1), then the target misrecognition rate of account A is 10^(-7), the target misrecognition rate of account B is 10^(-6), the target misrecognition rate of account C is 10^(-2), and the target misrecognition rate of account D is 10^(-2). If the order of magnitude loss is 10^2, the account misrecognition rate of account A is 10, the target misrecognition rate of account A is 10^(-5), the target misrecognition rate of account B is 10^(-4), the target misrecognition rate of account C is 10^0, and the target misrecognition rate of account D is 10^0.

[0106] Step 308, determining a target account from candidate accounts based on the account misrecognition rate.

[0107] The server may directly determine the candidate account with the lowest account misidentification rate and below the misidentification rate threshold as the target account. In a possible implementation, step 308 specifically includes the following steps:

[0108] Step 308a, determining the verification score of the candidate account based on the account misrecognition rate and the score mapping table.

[0109] The score mapping table contains the corresponding relationship between the account misrecognition rate and the verification score, and the account misrecognition rate and the verification score are negatively correlated. The server queries the score mapping table based on the account misrecognition rate to determine the verification score of each candidate account.

[0110] Indicative, Figure 6 A score mapping table is shown, in which the account misrecognition rate is expressed based on orders of magnitude, which contains verification scores corresponding to each order of magnitude of the account misrecognition rate from 10^0 to 10^(-7).

[0111] Step 308b: In response to the verification score corresponding to the candidate account satisfying the score threshold, the candidate account with the highest verification score is determined as the target account.

[0112] Step 308c: In response to the verification scores corresponding to the candidate accounts being less than the score threshold, it is determined that the target account does not exist.

[0113] In a possible implementation, a score threshold (e.g., 90 points) is set in the server, and the server determines the candidate account with the highest verification score among the candidate accounts that are higher than the score threshold as the target account. If there is no candidate account in the verification result whose score is higher than the score threshold, it is determined that the payment verification has failed.

[0114] Step 309: Make payment based on the target account.

[0115] The specific implementation of step 309 can refer to the above step 205, and the embodiment of the present application will not be repeated here.

[0116] In an embodiment of the present application, the account misrecognition rate of the candidate account is obtained based on the first misrecognition rate and the second misrecognition rate of the candidate account, and then a score is given based on the account misrecognition rate. The target account is determined based on the misrecognition rates of the two biometric features, so that the palm print features and the facial features complement each other, avoiding the misrecognition caused by the emphasis on one of the biometric features; and when calculating the account misrecognition rate, on the basis of calculating the theoretical misrecognition rate, the error between the theoretical calculation and the actual application is taken into account, and an order of magnitude loss is set, so that the account misrecognition rate is closer to the real data, thereby improving the authenticity and reliability of the account misrecognition rate and making the screening of the target account more stringent.

[0117] The above embodiment shows the payment verification process after the server receives the payment request. In a possible implementation, the server also receives an account registration request sent by the user's registration terminal and a facial image and palm print image used for registration to update the feature library. Figure 7 A flow chart of a payment verification method provided by another exemplary embodiment of the present application is shown. This embodiment is described by taking the method applied to a backend server of an application supporting a contactless payment function as an example, and the method includes the following steps.

[0118] Step 701: Receive an account registration request and payment verification data sent by an account registration terminal.

[0119] The payment verification data includes the registered account number, registered face image, and registered palm print image.

[0120] When a user registers an account, the registration terminal turns on the camera and prompts the user to collect biometric information to obtain the user's registered face image and registered palm print image. When it is determined that the registered face image and registered palm print image meet the registration conditions (clear and complete), the registration terminal sends the payment verification data and account registration request containing the registered account, registered face image and registered palm print image to the server, so that the server creates the corresponding account and updates the feature library based on the payment verification data and account registration request.

[0121] Step 702, feature extraction is performed on the registered face image and the registered palm print image to obtain the depth feature corresponding to the registered account, where the depth feature includes palm print feature and face feature.

[0122] After receiving the account registration request and payment verification data, the server performs deep feature extraction on the registered face image and the registered palm print image to obtain the palm print features and face features corresponding to the registered account.

[0123] Step 703: Update the feature library based on the registered account and the deep features.

[0124] The server adds the registered account to the feature library, and adds the palm print features and facial features corresponding to the registered account. For example, the feature library includes a first feature library for storing palm print features and a second feature library for storing facial features. The server creates a registered account, and adds the palm print features corresponding to the registered account to the first feature library, and adds the facial features corresponding to the registered account to the second feature library.

[0125] Figure 8 The interactive process between the registration terminal and the backend server during account registration is shown, and the process includes the following steps: Step 801, the terminal enables the payment application; Step 802, the terminal obtains the registration account identifier; Step 803, the terminal collects the palm print image; Step 804, the terminal collects the face image; Step 805, the terminal sends the image to the server; Step 806, the server performs palm print recognition; Step 807, the server performs face recognition; Step 808, the server updates the feature library and returns the registration result to the terminal.

[0126] In the embodiment of the present application, the server receives the account registration request and payment verification data sent by the registration terminal, and updates the feature library so that the user who registers the account can subsequently perform contactless payment.

[0127] The above embodiments are steps performed by the backend server of the payment application during the payment verification process. Before the server performs feature comparison, the payment terminal also needs to perform image preprocessing on the collected images to ensure that high-quality palm print images and face images are sent to the server, thereby ensuring the accuracy of feature recognition and feature comparison in the backend server and improving the accuracy of payment verification. Fig. 9 A flow chart of a payment verification method provided by an exemplary embodiment of the present application is shown. This embodiment is described by taking the method applied to a payment terminal supporting a contactless payment function as an example, and the method includes the following steps.

[0128] Step 901, collecting palm print images and face images through a camera.

[0129] In one possible implementation, merchants, banks, and other places where payment takes place are equipped with payment terminals, which are used to collect biometric images of users, namely palm print images and facial images, and send them to a server for feature recognition, so that the server makes payments based on the recognition results, thereby replacing traditional contact payment methods (such as password payment, fingerprint recognition payment, etc.).

[0130] Step 902: Send the palm print image and the face image to the server.

[0131] Among them, the server is used to determine at least two first candidate accounts based on palm print images, and at least two second candidate accounts based on face images, determine the target account based on the first false recognition rate corresponding to the first candidate accounts and the second false recognition rate corresponding to the second candidate accounts, and make payments based on the target account.

[0132] In one possible implementation, the terminal collects palm print images and facial images at a certain frequency, and determines whether the acquired images meet the image recognition conditions (for example, the palm and face are complete, the angle is correct, the image is clear, etc.). When palm print images and facial images that meet the image recognition conditions are detected, the collection is stopped, and the qualified images are sent to the server for payment verification to avoid payment verification errors or failures due to unclear collected images.

[0133] Step 903, receiving the payment result sent by the server, and displaying the payment result interface based on the payment result, wherein the payment result is used to indicate whether the payment of the account to be paid is successful or failed.

[0134] After the server makes payment based on the target account, it determines whether the payment is successful or failed, and returns the result to the payment terminal, where the types of payment failure include insufficient balance, identity verification failure, etc. The terminal displays the payment result interface to feedback the payment result to the user.

[0135] To sum up, during payment verification, based on the similarity between each candidate account and the acquired palm print image and face image, multiple first candidate accounts with the highest palm print similarity and multiple second candidate accounts with the highest face similarity are selected, and based on the misrecognition rate corresponding to each candidate account, the target account to be paid is determined. Compared with feature fusion or weighted calculation methods, it can avoid the situation where one feature is always focused, and the target account is determined based on the misrecognition rate without introducing additional parameters. It can also make the result more in line with reality, so as to improve the accuracy of palm and face payment.

[0136] Since payment terminals are usually located in public places, the images they collect may contain more information, such as multiple faces. In order to avoid interfering with the feature recognition of the background server and improve the efficiency and accuracy of feature recognition, the payment terminal needs to pre-process the collected images to ensure that high-quality face images and palm print images are sent to the server. Fig.10 A flow chart of a payment verification method provided by an exemplary embodiment of the present application is shown. This embodiment is described by taking the method applied to a payment terminal supporting a contactless payment function as an example, and the method includes the following steps.

[0137] Step 1001, in response to detecting a palm through a first camera, determining that a payment instruction is received.

[0138] Since the payment terminal is usually located in a public place, there may be a large number of people and dense distribution near the payment terminal, and the user to pay may not be prepared for feature collection or has not decided to pay. Therefore, if the image is directly collected when the payment initiation operation is detected, it may result in an invalid image, or even the features of other people may be mistakenly collected or the payment process may be triggered by mistake. In one possible implementation, the payment terminal adopts a mechanism of raising the palm to start the payment process, that is, when the user confirms to pay, he shows his palm to the camera set on the terminal. After the terminal detects the palm raising operation, it confirms that the payment instruction has been received and starts the payment process.

[0139] Step 1002: Capture a palm image through a first camera, and capture a face image through a second camera.

[0140] In a possible implementation, in application scenarios such as subway gates, since palm and face scanning require a high amount of time, the terminal is provided with at least a first camera and a second camera having different shooting angles, which are used to collect palm print images and face images, respectively, to improve payment efficiency.

[0141] Step 1003, performing image preprocessing on the palm image to obtain a palm print image, and performing image preprocessing on the face image to obtain a face image.

[0142] In a possible implementation manner, the process of performing image preprocessing of the palm image by the terminal includes the following steps:

[0143] Step 1003a: input the palm image into a palm detector to obtain palm frame selection data, where the palm frame selection data is used to indicate the area corresponding to the palm in the palm image.

[0144] In a possible implementation, a palm detector is deployed in the terminal, where the palm detector is a cascaded convolutional neural network (CNN) and is used to determine an approximate area where a palm is located in an image.

[0145] Step 1003b, cropping the palm image based on the palm frame selection data to obtain the middle palm image.

[0146] The palm frame selection data output by the palm detector is used to indicate the position of the palm area in the palm image. In order to avoid the area selected by the palm detector not including the complete palm area, resulting in incomplete subsequent palm prints and incorrect palm print recognition, the terminal further determines the area corresponding to the palm frame selection data as the first palm area, determines the second palm area based on the first palm area and the first preset ratio, and crops according to the second palm area to obtain the middle palm image. The first preset ratio is less than 1 (for example, 1:1.5, 1:2, etc.), that is, the second palm area covers the first palm area and has a larger range to ensure that the complete palm prints are included in the middle palm image.

[0147] like Fig.11 As shown, the palm area indicated by the palm frame selection data output by the palm detector is the area shown by the box in the palm middle image 1101. The terminal crops the palm image based on the palm frame selection data and the first preset ratio to obtain the palm middle image 1101, that is, the palm middle image 1101 includes the image area corresponding to the palm frame selection data and has a larger range.

[0148] Step 1003c: input the middle palm image into the palm key point detection model to obtain the palm key points in the middle palm image.

[0149] Based on the palm detection frame determined in the previous step, the palm key points are obtained using the palm key point detection model, where the palm key point detection model is a regression model that outputs at least three palm key points for constructing a coordinate system and determining the ROI position.

[0150] Indicatively, Fig.11 As shown, the terminal inputs the palm middle image 1101 into the palm key point detection model to obtain the positions of the palm key points A, B and C.

[0151] Step 1003d, based on the palm coordinate system corresponding to the palm key points, determine the palm region of interest (ROI), where the palm ROI is the area in the palm region that contains the palm prints.

[0152] Step 1003e: determine the image corresponding to the palm ROI as the palm print image.

[0153] like Fig.11 As shown, the terminal establishes a coordinate system with the straight line where the key point A corresponding to the index finger and the key point C corresponding to the ring finger are located as the horizontal axis, and the straight line passing through the key point B corresponding to the middle finger and perpendicular to the horizontal axis as the vertical axis. The center point D of the palm print is found along the negative direction of the vertical axis, and the length of BD is equal to the length of AC at the first preset multiple (for example, two-thirds times). The second preset multiple of the length of AC (for example, six-seven times) is used as the side length of the palm ROI area, and the area with point D as the center and the side parallel to the horizontal axis or the vertical axis is used as the palm ROI area. The palm intermediate image 1101 is cropped according to the palm ROI to obtain the palm print image 1102.

[0154] In a possible implementation manner, the process of performing image preprocessing on the facial image by the terminal includes the following steps:

[0155] Step 1003f, input the facial image into a face detector to obtain face frame selection data, where the face frame selection data is used to indicate the area corresponding to the face in the facial image.

[0156] In a possible implementation, a face detector is deployed in the terminal, and the face detector is a cascaded CNN model for determining an approximate area where a face is located in an image.

[0157] Step 1003g, cropping the facial image based on the facial frame selection data to obtain an intermediate facial image.

[0158] The face frame selection data output by the face detector is used to indicate the position of the face area in the face image. In order to avoid the situation where the area selected by the face detector does not contain a complete face, resulting in incomplete subsequent face and incorrect face recognition, the terminal further determines the area corresponding to the face frame selection data as the first face area, determines the second face area based on the first face area and the second preset ratio, and crops according to the second face area to obtain a face intermediate image. The second preset ratio is less than 1 (for example, 1:1.5, 1:2, etc.), that is, the second face area covers the first face area and has a larger range to ensure that the complete face is included in the face intermediate image.

[0159] like Fig.12 As shown, the face area indicated by the face frame selection data output by the face detector is the area shown by the box in the face intermediate image 1201. The terminal crops the face image based on the face frame selection data and the second preset ratio to obtain the face intermediate image 1201, that is, the face intermediate image 1201 includes the image area corresponding to the face frame selection data and has a larger range.

[0160] Step 1003h, input the middle face image into the face key point detection model to obtain the face key points in the middle face image.

[0161] Based on the face detection frame obtained in the previous step, the key points of the facial features are regressed based on the face key point detection model. Optionally, the face key point detection model and the palm key point detection model are two independent models, or they are the same model. Fig.12 As shown, the terminal inputs the intermediate face image 1201 into the key point detection model to obtain the position data of 5 facial key points.

[0162] Step 1003i, determining a face ROI region based on face key points, determining an image corresponding to the face ROI region as a face image, wherein the face ROI region is a region in the face region that contains face features.

[0163] The terminal determines the face ROI area based on the location of the face key points, such as Fig.12 As shown, the terminal determines the face ROI area based on the face key points, and crops the face intermediate image 1201 based on the face ROI area to obtain the face image 1202.

[0164] Step 1004: input the palm print image and the face image into a quality assessment model to obtain a palm print quality score and a face quality score.

[0165] The quality score is used to indicate the feature recognizability of the image. Specifically, the quality assessment model in the embodiment of the present application assesses the quality of the palm print image and the face image from multiple dimensions such as image clarity, feature integrity, image brightness, and shooting angle to obtain a quality score. Fig.13 As shown in the figure, image quality can be roughly divided into five categories. Except for the normal type of image, the quality scores of the other four categories are low. Through the quality score, the terminal can select the best image from a series of video frames captured and then transmit it to the server, thereby improving the accuracy and availability of payment verification.

[0166] In a possible implementation, the payment terminal inputs the palmprint image and face image obtained after image preprocessing into the quality assessment model respectively, obtains the palmprint quality score of the palmprint image and the face quality score of the face image, and screens out high-quality palmprint images and face images based on the quality scores.

[0167] Optionally, the quality assessment of the palm print image and the face image adopts the same quality assessment model, or the palm print image and the face image are respectively input into two independent quality assessment models. The embodiment of the present application is not limited to this. The embodiment of the present application is illustrated by taking the palm print image and the face image as input into two independent quality assessment models respectively as an example. When the terminal is provided with a first camera and a second camera, the face image and the palm image can be collected at the same time, and the face image and the palm image can be synchronously preprocessed and the quality assessment is performed to reduce the time consumed by the payment process.

[0168] Step 1005: In response to the fact that the palm print quality score and the face quality score are both higher than the quality score threshold, the palm vein image is input into a liveness recognition model to obtain a liveness recognition result of the liveness recognition model.

[0169] The terminal is also provided with an infrared camera. When receiving a payment instruction, the terminal collects the palm image through the first camera and the face image through the second camera, and at the same time collects the vein image corresponding to the palm through the infrared camera. Accordingly, the terminal performs image preprocessing on the vein image to obtain the vein ROI of the vein image, and extracts the image corresponding to the vein ROI as the palm vein image, and inputs the palm vein image into the liveness recognition model to obtain the liveness recognition result of the liveness recognition model.

[0170] The terminal first inputs the palm vein image into the liveness model, and based on the output of the model, determines whether the user currently performing the payment operation is a live user. Through the liveness recognition mechanism, the terminal will be able to avoid two-dimensional attack methods such as photos, and improve the security of contactless payment.

[0171] Optionally, in response to the palm print quality score being lower than the quality score threshold, the next palm print image is collected through the first camera; in response to the face quality score being lower than the quality score threshold, the next face image is collected through the second camera. That is, the terminal continuously collects palm images and face images, and if a palm print image whose quality score reaches the quality score threshold is collected, the first camera stops shooting, and if a palm print image whose quality score reaches the quality score threshold is collected, the second camera stops shooting.

[0172] Step 1006: In response to the live body recognition result that the image acquisition object is a live body, the palm print image and the face image are sent to the server.

[0173] When the liveness recognition result indicates that the image acquisition object is a live object, the terminal will send the palm print image and face image to the server for identity authentication and payment, otherwise the payment process will be terminated. In a possible implementation, in order to avoid the user's payment failure caused by the wrong model detection result and improve the security of the system, when the liveness recognition result indicates that the image acquisition object is a non-live object, the palm image and face image are re-collected. If the liveness recognition result indicates that the same image acquisition object is a non-live object, the number of times reaches the threshold, the payment process is terminated and an alarm message is generated.

[0174] Step 1007, receiving the payment result sent by the server, and displaying the payment result interface based on the payment result, wherein the payment result is used to indicate whether the payment of the account to be paid is successful or failed.

[0175] The specific implementation of step 1007 can refer to the above step 903, and the embodiment of the present application will not be repeated here.

[0176] In the embodiment of the present application, a payment mechanism of raising the palm to start the payment process is adopted, combined with liveness recognition, to ensure that the user makes payment when he is sure of his willingness to pay, thereby improving the security of contactless payment; the collected palm images and facial images are image preprocessed, and the image corresponding to the ROI is sent to the server for feature recognition and comparison to improve the recognition efficiency of the server; by performing quality inspection, it is ensured that high-quality images are sent to the server to improve the accuracy and availability of payment verification.

[0177] In another possible implementation, in order to reduce hardware design and cost, in a scenario where the payment time requirement is not high, the terminal can collect the palm image and the face image serially by multiplexing the same camera, and the above step 901 further includes the following steps:

[0178] Step 901a, in response to detecting a palm through the camera, determining that a payment instruction is received, and capturing a palm image through the camera.

[0179] Similarly, the terminal triggers the payment process based on the mechanism of raising the palm. When the palm is detected, the palm image and palm vein image are collected through the camera. When the liveness recognition is performed based on the palm vein image and it is confirmed that it is a live palm, the palm image is further pre-processed.

[0180] Step 901b, performing image preprocessing on the palm image to obtain a palm print image.

[0181] The terminal determines the palm ROI through the palm detector and the palm key point detection model, and then performs quality assessment. When a palm print image with a quality score that meets the quality score threshold is collected, the palm print image is sent to the server, and the face image is continued to be collected through the camera. In a possible implementation, the terminal is also provided with an infrared camera. While collecting palm images through an ordinary camera, the infrared camera collects palm vein images for liveness recognition.

[0182] Step 901c, multiplexing the camera to capture the facial image.

[0183] Step 901d, performing image preprocessing on the facial image to obtain a facial image.

[0184] Correspondingly, after the terminal captures the facial image through the camera, it determines the face ROI based on the face detector and the face key point detection model, and then performs a quality assessment. When a face image with a quality score that meets the quality score threshold is captured, the face image is sent to the server, ending the image detection link.

[0185] Fig.14: is a structural block diagram of a payment verification device provided by an exemplary embodiment of the present application, the device comprising:

[0186] An acquisition module 1401 is used to acquire a palm print image and a face image;

[0187] A first determination module 1402 is used to determine at least two first candidate accounts based on the palm print image, and to determine at least two second candidate accounts based on the face image, wherein the palm print similarity corresponding to the first candidate accounts is higher than the palm print similarity corresponding to the other accounts, and the face similarity corresponding to the second candidate accounts is higher than the face similarity corresponding to the other accounts;

[0188] A second determination module 1403 is used to determine a first misrecognition rate corresponding to the first candidate account based on the palm print similarity of the first candidate account, and to determine a second misrecognition rate corresponding to the second candidate account based on the face similarity corresponding to the second candidate account;

[0189] A third determination module 1404, configured to determine a target account from the first candidate account and the second candidate account based on the first false recognition rate and the second false recognition rate;

[0190] The payment module 1405 is used to make payment based on the target account.

[0191] Optionally, the second determining module 1403 includes:

[0192] A first query unit, configured to query a first mapping table to determine a first similarity interval corresponding to the palm print similarity of the first candidate account, wherein the first mapping table includes a correspondence between the similarity interval and the first false recognition rate;

[0193] A first determining unit, configured to determine the first misrecognition rate based on the first similarity interval and the first mapping table;

[0194] A second query unit, configured to query a second mapping table to determine a second similarity interval corresponding to the face similarity of the second candidate account, wherein the second mapping table includes a correspondence between the similarity interval and the second false recognition rate;

[0195] A second determining unit is configured to determine the second misrecognition rate based on the second similarity interval and the second mapping table.

[0196] Optionally, the third determining module 1404 includes:

[0197] a third determining unit, configured to determine an account misrecognition rate of the candidate account based on the first misrecognition rate, the second misrecognition rate, and an account type of the candidate account, the account type comprising an intersection account and a non-intersection account, the intersection account belonging to both the first candidate account and the second candidate account;

[0198] A fourth determining unit is used to determine the target account from the candidate accounts based on the account misrecognition rate.

[0199] Optionally, the third determining unit is further configured to:

[0200] In response to the candidate account being the intersection account, determining the account misrecognition rate of the candidate account based on the first misrecognition rate and the second misrecognition rate corresponding to the candidate account;

[0201] In response to the candidate account being the non-intersection account and belonging to the first candidate account, determining the second misrecognition rate of the candidate account to be 1; determining the account misrecognition rate of the candidate account based on the first misrecognition rate and the second misrecognition rate corresponding to the candidate account;

[0202] In response to the candidate account being the non-intersection account and belonging to the second candidate account, determining the first misrecognition rate of the candidate account to be 1; and determining the account misrecognition rate of the candidate account based on the first misrecognition rate and the second misrecognition rate corresponding to the candidate account.

[0203] Optionally, the first misrecognition rate and the second misrecognition rate are expressed by order of magnitude;

[0204] The third determining unit is further configured to:

[0205] Determining a target misrecognition rate for the candidate account based on a product of the first misrecognition rate and the second misrecognition rate;

[0206] The account misrecognition rate is determined based on the target misrecognition rate and the order of magnitude loss, where the order of magnitude loss is the difference in order of magnitude between the misrecognition rate obtained by theoretical calculation and the misrecognition rate obtained by actual testing.

[0207] Optionally, the fourth determining unit is further configured to:

[0208] Determine the verification score of the candidate account based on the account misrecognition rate and a score mapping table, wherein the score mapping table includes a correspondence between the account misrecognition rate and the verification score, and the account misrecognition rate and the verification score are negatively correlated;

[0209] In response to the verification score corresponding to the candidate account satisfying the score threshold, determining the candidate account with the highest verification score as the target account;

[0210] In response to the verification scores corresponding to the candidate accounts being less than the score threshold, it is determined that the target account does not exist.

[0211] Optionally, the device further comprises:

[0212] A receiving module, used to receive an account registration request and payment verification data sent by an account registration terminal, wherein the payment verification data includes a registered account, a registered face image, and a registered palm print image;

[0213] A feature extraction module, used to extract features from the registered face image and the registered palm print image, and obtain a depth feature corresponding to the registered account, wherein the depth feature includes a palm print feature and a face feature;

[0214] An updating module is used to update a feature library based on the registered account and the deep features.

[0215] Fig.15 : is a structural block diagram of a payment verification device provided by an exemplary embodiment of the present application, the device comprising:

[0216] The image acquisition module 1501 is used to acquire palm print images and face images through a camera;

[0217] a sending module 1502, configured to send the palm print image and the face image to a server, wherein the server is configured to determine at least two first candidate accounts based on the palm print image and at least two second candidate accounts based on the face image, determine a target account based on a first false recognition rate corresponding to the first candidate account and a second false recognition rate corresponding to the second candidate account, and make payment based on the target account;

[0218] Display module 1503 is used to receive the payment result sent by the server and display a payment result interface based on the payment result, wherein the payment result is used to indicate whether the payment of the account to be paid is successful or failed.

[0219] Optionally, the terminal is provided with a first camera and a second camera;

[0220] The image acquisition module 1501 includes:

[0221] a fifth determining unit, configured to determine that a payment instruction is received in response to detecting a palm through the first camera;

[0222] A first acquisition unit, configured to acquire a palm image through the first camera, and acquire a facial image through the second camera;

[0223] The first processing unit is used to perform image preprocessing on the palm image to obtain the palmprint image, and perform image preprocessing on the face image to obtain the face image.

[0224] Optionally, the image processing unit is further used for:

[0225] Inputting the palm image into a palm detector to obtain palm frame selection data, where the palm frame selection data is used to indicate an area corresponding to the palm in the palm image;

[0226] Cropping the palm image based on the palm frame selection data to obtain a palm middle image;

[0227] Inputting the palm middle image into a palm key point detection model to obtain palm key points in the palm middle image;

[0228] Determine a palm region of interest ROI based on a palm coordinate system corresponding to the palm key points, wherein the palm ROI is a region of the palm region that includes palm lines;

[0229] The image corresponding to the palm ROI is determined as the palmprint image.

[0230] Optionally, the image processing unit is further used for:

[0231] Inputting the facial image into a face detector to obtain face frame selection data, where the face frame selection data is used to indicate an area corresponding to a face in the facial image;

[0232] Cropping the facial image based on the facial frame selection data to obtain an intermediate facial image;

[0233] Inputting the face intermediate image into a face key point detection model to obtain face key points in the face intermediate image;

[0234] A face ROI region is determined based on the face key points, and an image corresponding to the face ROI region is determined as the face image. The face ROI region is a region in the face region that contains face features.

[0235] Optionally, the terminal is further provided with an infrared camera, and the infrared camera is used to collect palm vein images; the sending module 1502 includes:

[0236] A quality detection unit, used for inputting the palm print image and the face image into a quality assessment model to obtain a palm print quality score and a face quality score, wherein the quality score is used to indicate the clarity and completeness of the image;

[0237] a liveness detection unit, configured to input the palm vein image into a liveness recognition model in response to the palm print quality score and the face quality score being both higher than a quality score threshold, to obtain a liveness recognition result of the liveness recognition model, wherein the liveness recognition result is used to indicate whether the image acquisition object is a live object or a non-live object;

[0238] A sending unit is used for sending the palm print image and the face image to the server in response to the living body recognition result indicating that the image acquisition object is a living body.

[0239] Optionally, the image acquisition module 1501 includes:

[0240] a second acquisition unit, configured to, in response to detecting a palm through the camera, determine that a payment instruction has been received, and acquire a palm image through the camera;

[0241] A second processing unit, configured to perform image preprocessing on the palm image to obtain the palm print image;

[0242] A third acquisition unit, used for multiplexing the camera to acquire facial images;

[0243] The third processing unit is used to perform image preprocessing on the facial image to obtain the facial image.

[0244] To summarize, in the embodiments of the present application, the false recognition rates of palmprint recognition results and face recognition results are obtained respectively through preliminary tests. During payment verification, based on the similarity between each candidate account and the acquired palmprint image and face image, multiple first candidate accounts with the highest palmprint similarity and multiple second candidate accounts with the highest face similarity are selected. Based on the false recognition rates corresponding to each candidate account, the target account to be paid is determined. Compared with feature fusion or weighted calculation, this method can avoid always focusing on one feature, and the target account is determined based on the false recognition rate without introducing additional parameters. This can also make the result more realistic, so as to improve the accuracy of palm and face payment.

[0245] Please refer to Fig.16 , which shows a schematic diagram of the structure of a server provided by an embodiment of the present application. Specifically:

[0246] The server 1600 includes a central processing unit (CPU) 1601, a system memory 1604 including a random access memory (RAM) 1602 and a read only memory (ROM) 1603, and a system bus 1605 connecting the system memory 1604 and the central processing unit 1601. The server 1600 also includes a basic input / output (I / O) controller 1606 for facilitating information transmission between various components in the computer, and a large-capacity storage device 1607 for storing an operating system 1613, application programs 1614, and other program modules 1615.

[0247] The basic input / output system 1606 includes a display 1608 for displaying information and an input device 1609 such as a mouse and a keyboard for user inputting information. The display 1608 and the input device 1609 are connected to the central processing unit 1601 through an input / output controller 1610 connected to the system bus 1605. The basic input / output system 1606 may also include an input / output controller 1610 for receiving and processing inputs from a plurality of other devices such as a keyboard, a mouse, or an electronic stylus. Similarly, the input / output controller 1610 also provides output to a display screen, a printer, or other types of output devices.

[0248] The mass storage device 1607 is connected to the central processing unit 1601 via a mass storage controller (not shown) connected to the system bus 1605. The mass storage device 1607 and its associated computer-readable medium provide non-volatile storage for the server 1600. That is, the mass storage device 1607 may include a computer-readable medium (not shown) such as a hard disk or a compact disc read-only memory (CD-ROM) drive.

[0249] Without loss of generality, the computer readable medium may include computer storage media and communication media. Computer storage media include volatile and non-volatile, removable and non-removable media implemented by any method or technology for storing information such as computer readable instructions, data structures, program modules or other data. Computer storage media include RAM, ROM, Erasable Programmable Read Only Memory (EPROM), flash memory or other solid-state storage technology, CD-ROM, Digital Video Disc (DVD) or other optical storage, cassettes, magnetic tapes, disk storage or other magnetic storage devices. Of course, those skilled in the art will appreciate that the computer storage medium is not limited to the above. The above-mentioned system memory 1604 and mass storage device 1607 can be collectively referred to as memory.

[0250] According to various embodiments of the present application, the server 1600 may also be connected to a remote computer on a network through a network such as the Internet. That is, the server 1600 may be connected to a network 1612 through a network interface unit 1611 connected to the system bus 1605, or the network interface unit 1611 may be used to connect to other types of networks or remote computer systems (not shown).

[0251] The memory also includes at least one instruction, at least one program, code set or instruction set, which is stored in the memory and configured to be executed by one or more processors to implement the above-mentioned payment verification method.

[0252] Please refer to Fig.17 , which shows a block diagram of a terminal 1700 provided by an exemplary embodiment of the present application. The terminal 1700 may be a portable mobile terminal, such as a smart phone, a tablet computer, a Moving Picture Experts Group Audio Layer III (MP3) player, or a Moving Picture Experts Group Audio Layer IV (MP4) player. The terminal 1700 may also be referred to as a user device, a portable terminal, or other names.

[0253] Typically, the terminal 1700 includes a processor 1701 and a memory 1702 .

[0254] The processor 1701 may include one or more processing cores, such as a 4-core processor, an 8-core processor, etc. The processor 1701 may be implemented in at least one hardware form of digital signal processing (DSP), field-programmable gate array (FPGA), and programmable logic array (PLA). The processor 1701 may also include a main processor and a coprocessor. The main processor is a processor for processing data in the awake state, also known as a central processing unit (CPU); the coprocessor is a low-power processor for processing data in the standby state. In some embodiments, the processor 1701 may be integrated with a graphics processing unit (GPU), which is responsible for rendering and drawing the content to be displayed on the display screen. In some embodiments, the processor 1701 may also include an artificial intelligence (AI) processor, which is used to process computing operations related to machine learning.

[0255] The memory 1702 may include one or more computer-readable storage media, which may be tangible and non-transitory. The memory 1702 may also include a high-speed random access memory, and a non-volatile memory, such as one or more disk storage devices, flash memory storage devices. In some embodiments, the non-transitory computer-readable storage medium in the memory 1702 is used to store at least one instruction, which is used to be executed by the processor 1701 to implement the method provided in the embodiment of the present application.

[0256] In some embodiments, the terminal 1700 may further include: a peripheral device interface 1703 and at least one peripheral device. Specifically, the peripheral device includes: at least one of a radio frequency circuit 1704, a touch display screen 1705, a camera 1706, an audio circuit 1707, a positioning component 1708 and a power supply 1709.

[0257] The peripheral device interface 1703 may be used to connect at least one peripheral device related to input / output (I / O) to the processor 1701 and the memory 1702. In some embodiments, the processor 1701, the memory 1702, and the peripheral device interface 1703 are integrated on the same chip or circuit board; in some other embodiments, any one or two of the processor 1701, the memory 1702, and the peripheral device interface 1703 may be implemented on a separate chip or circuit board, which is not limited in this embodiment.

[0258] The radio frequency circuit 1704 is used to receive and transmit radio frequency (RF) signals, also known as electromagnetic signals. The radio frequency circuit 1704 communicates with the communication network and other communication devices through electromagnetic signals. The radio frequency circuit 1704 converts the electrical signal into an electromagnetic signal for transmission, or converts the received electromagnetic signal into an electrical signal. Optionally, the radio frequency circuit 1704 includes: an antenna system, an RF transceiver, one or more amplifiers, a tuner, an oscillator, a digital signal processor, a codec chipset, a user identity module card, and the like. The radio frequency circuit 1704 can communicate with other terminals through at least one wireless communication protocol. The wireless communication protocol includes, but is not limited to: the World Wide Web, a metropolitan area network, an intranet, various generations of mobile communication networks (2G, 3G, 4G and 5G), a wireless local area network and / or a wireless fidelity (Wireless Fidelity, WiFi) network. In some embodiments, the radio frequency circuit 1704 may also include circuits related to near field communication (NFC), which is not limited in this application.

[0259] The touch display screen 1705 is used to display the UI. The UI may include graphics, text, icons, videos, and any combination thereof. The touch display screen 1705 also has the ability to collect touch signals on the surface or above the surface of the touch display screen 1705. The touch signal can be input to the processor 1701 as a control signal for processing. The touch display screen 1705 is used to provide virtual buttons and / or virtual keyboards, also known as soft buttons and / or soft keyboards. In some embodiments, the touch display screen 1705 can be one, and the front panel of the terminal 1700 is set; in other embodiments, the touch display screen 1705 can be at least two, which are respectively set on different surfaces of the terminal 1700 or are folded; in some other embodiments, the touch display screen 1705 can be a flexible display screen, which is set on the curved surface or folded surface of the terminal 1700. Even, the touch display screen 1705 can also be set to a non-rectangular irregular shape, that is, a special-shaped screen. The touch display screen 1705 can be made of materials such as a liquid crystal display (LCD) and an organic light-emitting diode (OLED).

[0260] The camera assembly 1706 is used to capture images or videos. Optionally, the camera assembly 1706 includes a front camera and a rear camera. Typically, the front camera is used to realize video calls or selfies, and the rear camera is used to realize the shooting of photos or videos. In some embodiments, there are at least two rear cameras, which are any one of a main camera, a depth of field camera, and a wide-angle camera, so as to realize the background blur function by fusion of the main camera and the depth of field camera, and the fusion of the main camera and the wide-angle camera to realize panoramic shooting and virtual reality (VR) shooting functions. In some embodiments, the camera assembly 1706 may also include a flash. The flash can be a monochrome temperature flash or a dual-color temperature flash. A dual-color temperature flash refers to a combination of a warm light flash and a cold light flash, which can be used for light compensation at different color temperatures.

[0261] The audio circuit 1707 is used to provide an audio interface between the user and the terminal 1700. The audio circuit 1707 may include a microphone and a speaker. The microphone is used to collect sound waves from the user and the environment, and convert the sound waves into electrical signals and input them into the processor 1701 for processing, or input them into the radio frequency circuit 1704 to achieve voice communication. For the purpose of stereo acquisition or noise reduction, there may be multiple microphones, which are respectively arranged at different parts of the terminal 1700. The microphone may also be an array microphone or an omnidirectional acquisition microphone. The speaker is used to convert the electrical signal from the processor 1701 or the radio frequency circuit 1704 into sound waves. The speaker may be a traditional film speaker or a piezoelectric ceramic speaker. When the speaker is a piezoelectric ceramic speaker, it can not only convert the electrical signal into sound waves audible to humans, but also convert the electrical signal into sound waves inaudible to humans for purposes such as ranging. In some embodiments, the audio circuit 1707 may also include a headphone jack.

[0262] The positioning component 1708 is used to locate the current geographical location of the terminal 1700 to implement navigation or location-based services (LBS). The positioning component 1708 can be a positioning component based on the Global Positioning System (GPS) of the United States, the Beidou system of China, or the Galileo system of Russia.

[0263] The power supply 1709 is used to power various components in the terminal 1700. The power supply 1709 can be an alternating current, a direct current, a disposable battery, or a rechargeable battery. When the power supply 1709 includes a rechargeable battery, the rechargeable battery can be a wired rechargeable battery or a wireless rechargeable battery. A wired rechargeable battery is a battery charged through a wired line, and a wireless rechargeable battery is a battery charged through a wireless coil. The rechargeable battery can also be used to support fast charging technology.

[0264] In some embodiments, the terminal 1700 further includes one or more sensors 1710 , including but not limited to: an acceleration sensor 1711 , a gyroscope sensor 1712 , a pressure sensor 1713 , a fingerprint sensor 1714 , an optical sensor 1715 , and a proximity sensor 1716 .

[0265] The acceleration sensor 1711 can detect the magnitude of acceleration on the three coordinate axes of the coordinate system established by the terminal 1700. For example, the acceleration sensor 1711 can be used to detect the components of gravity acceleration on the three coordinate axes. The processor 1701 can control the touch display screen 1705 to display the user interface in a horizontal view or a vertical view according to the gravity acceleration signal collected by the acceleration sensor 1711. The acceleration sensor 1711 can also be used for collecting game or user motion data.

[0266] The gyro sensor 1712 can detect the body direction and rotation angle of the terminal 1700, and the gyro sensor 1712 can cooperate with the acceleration sensor 1711 to collect the user's 3D actions on the terminal 1700. The processor 1701 can implement the following functions based on the data collected by the gyro sensor 1712: motion sensing (such as changing the UI according to the user's tilt operation), image stabilization during shooting, game control, and inertial navigation.

[0267] The pressure sensor 1713 can be set on the side frame of the terminal 1700 and / or the lower layer of the touch display screen 1705. When the pressure sensor 1713 is set on the side frame of the terminal 1700, it can detect the user's holding signal of the terminal 1700, and perform left and right hand recognition or shortcut operation according to the holding signal. When the pressure sensor 1713 is set on the lower layer of the touch display screen 1705, the operability control on the UI interface can be controlled according to the user's pressure operation on the touch display screen 1705. The operability control includes at least one of a button control, a scroll bar control, an icon control, and a menu control.

[0268] The fingerprint sensor 1714 is used to collect the user's fingerprint to identify the user's identity based on the collected fingerprint. When the user's identity is identified as a trusted identity, the processor 1701 authorizes the user to perform relevant sensitive operations, including unlocking the screen, viewing encrypted information, downloading software, paying, and changing settings. The fingerprint sensor 1714 can be set on the front, back, or side of the terminal 1700. When a physical button or a manufacturer logo is set on the terminal 1700, the fingerprint sensor 1714 can be integrated with the physical button or the manufacturer logo.

[0269] The optical sensor 1715 is used to collect the ambient light intensity. In one embodiment, the processor 1701 can control the display brightness of the touch display screen 1705 according to the ambient light intensity collected by the optical sensor 1715. Specifically, when the ambient light intensity is high, the display brightness of the touch display screen 1705 is increased; when the ambient light intensity is low, the display brightness of the touch display screen 1705 is reduced. In another embodiment, the processor 1701 can also dynamically adjust the shooting parameters of the camera assembly 1706 according to the ambient light intensity collected by the optical sensor 1715.

[0270] The proximity sensor 1716, also called a distance sensor, is usually arranged on the front of the terminal 1700. The proximity sensor 1716 is used to collect the distance between the user and the front of the terminal 1700. In one embodiment, when the proximity sensor 1716 detects that the distance between the user and the front of the terminal 1700 is gradually decreasing, the processor 1701 controls the touch display screen 1705 to switch from the screen-on state to the screen-off state; when the proximity sensor 1716 detects that the distance between the user and the front of the terminal 1700 is gradually increasing, the processor 1701 controls the touch display screen 1705 to switch from the screen-off state to the screen-on state.

[0271] Those skilled in the art will understand that Fig.13 The structure shown in the figure does not constitute a limitation on the terminal 1700, and the terminal 1700 may include more or less components than those shown in the figure, or combine certain components, or adopt a different component arrangement.

[0272] An embodiment of the present application further provides a computer-readable storage medium, which stores at least one instruction, and the at least one instruction is loaded and executed by a processor to implement the payment verification method described in the above embodiments.

[0273] According to one aspect of the present application, a computer program product or computer program is provided, the computer program product or computer program includes computer instructions, the computer instructions are stored in a computer-readable storage medium. A processor of a computer device or terminal reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device or terminal performs the payment verification method provided in various optional implementations of the above aspects.

[0274] Those skilled in the art should be aware that in one or more of the above examples, the functions described in the embodiments of the present application can be implemented with hardware, software, firmware, or any combination thereof. When implemented using software, these functions can be stored in a computer-readable storage medium or transmitted as one or more instructions or codes on a computer-readable storage medium. Computer-readable storage media include computer storage media and communication media, wherein the communication media include any media that facilitates the transmission of a computer program from one place to another. The storage medium can be any available medium that a general or special-purpose computer can access.

[0275] The above description is only an optional embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A payment verification method, characterized in that: The method comprises: Obtain palmprint images and face images; Determine at least two first candidate accounts based on the palm print image, and determine at least two second candidate accounts based on the face image, wherein the palm print similarity corresponding to the first candidate accounts is higher than the palm print similarity corresponding to the other accounts, and the face similarity corresponding to the second candidate accounts is higher than the face similarity corresponding to the other accounts; Querying a first mapping table to determine a first similarity interval corresponding to the palm print similarity of the first candidate account, wherein the first mapping table includes a correspondence between the similarity interval and the first false recognition rate; Determining the first misrecognition rate based on the first similarity interval and the first mapping table; Querying a second mapping table to determine a second similarity interval corresponding to the face similarity of the second candidate account, wherein the second mapping table includes a correspondence between the similarity interval and the second false recognition rate; Determining the second misrecognition rate based on the second similarity interval and the second mapping table; Determining the account misrecognition rate of the candidate account based on the first misrecognition rate, the second misrecognition rate, and an account type of the candidate account, the account type including an intersection account and a non-intersection account, the intersection account belonging to both the first candidate account and the second candidate account; Determining a target account from the candidate accounts based on the account misrecognition rate; Payment is made based on the target account.

2. The method according to claim 1, characterized in that The determining the account misrecognition rate of the candidate account based on the first misrecognition rate, the second misrecognition rate, and the account type of the candidate account includes: In response to the candidate account being the intersection account, determining the account misrecognition rate of the candidate account based on the first misrecognition rate and the second misrecognition rate corresponding to the candidate account; In response to the candidate account being the non-intersection account and belonging to the first candidate account, determining the second misrecognition rate of the candidate account to be 1; determining the account misrecognition rate of the candidate account based on the first misrecognition rate and the second misrecognition rate corresponding to the candidate account; In response to the candidate account being the non-intersection account and belonging to the second candidate account, determining the first misrecognition rate of the candidate account to be 1; and determining the account misrecognition rate of the candidate account based on the first misrecognition rate and the second misrecognition rate corresponding to the candidate account.

3. The method according to claim 2, characterized in that The first false recognition rate and the second false recognition rate are expressed by order of magnitude; The determining the account misrecognition rate of the candidate account based on the first misrecognition rate and the second misrecognition rate corresponding to the candidate account includes: Determining a target misrecognition rate for the candidate account based on a product of the first misrecognition rate and the second misrecognition rate; The account misrecognition rate is determined based on the product of the target misrecognition rate and the order of magnitude loss. The order of magnitude loss is the difference in order of magnitude between the theoretical misrecognition rate and the actual misrecognition rate. The theoretical misrecognition rate is equal to the product of the misrecognition rate corresponding to the palmprint similarity threshold and the misrecognition rate corresponding to the face similarity threshold for multiple sample accounts in the sample space.

4. The method according to any one of claims 1 to 3, characterized in that: The determining a target account from the candidate accounts based on the account misrecognition rate includes: Determine the verification score of the candidate account based on the account misrecognition rate and a score mapping table, wherein the score mapping table includes a correspondence between the account misrecognition rate and the verification score, and the account misrecognition rate and the verification score are negatively correlated; In response to the verification score corresponding to the candidate account satisfying the score threshold, determining the candidate account with the highest verification score as the target account; In response to the verification scores corresponding to the candidate accounts being less than the score threshold, it is determined that the target account does not exist.

5. The method according to any one of claims 1 to 3, characterized in that: The method further comprises: Receiving an account registration request and payment verification data sent by an account registration terminal, wherein the payment verification data includes a registered account number, a registered face image, and a registered palm print image; Perform feature extraction on the registered face image and the registered palm print image to obtain depth features corresponding to the registered account, where the depth features include palm print features and face features; A feature library is updated based on the registered account and the deep feature.

6. A payment verification method, characterized in that: The method comprises: Collect palm print images and face images through cameras; The palm print image and the face image are sent to a server, the server is used to determine at least two first candidate accounts based on the palm print image, and at least two second candidate accounts based on the face image, query a first mapping table to determine a first similarity interval corresponding to the palm print similarity of the first candidate account, the first mapping table contains a correspondence between the similarity interval and the first false recognition rate; determine the first false recognition rate based on the first similarity interval and the first mapping table; query a second mapping table to determine a second similarity interval corresponding to the face similarity of the second candidate account, the second mapping table contains a correspondence between the similarity interval and the second false recognition rate; determine the second false recognition rate based on the second similarity interval and the second mapping table, determine the account false recognition rate of the candidate account based on the first false recognition rate, the second false recognition rate and the account type of the candidate account, the account type includes an intersection account and a non-intersection account, the intersection account belongs to both the first candidate account and the second candidate account; determine a target account from the candidate account based on the account false recognition rate, and make payment based on the target account; The payment result sent by the server is received, and a payment result interface is displayed based on the payment result, wherein the payment result is used to indicate whether the payment of the account to be paid is successful or failed.

7. The method according to claim 6, characterized in that The terminal is provided with a first camera and a second camera; The collecting of palm print images and face images by using a camera includes: In response to detecting a palm through the first camera, determining that a payment instruction is received; Capturing a palm image through the first camera, and capturing a facial image through the second camera; The palm image is subjected to image preprocessing to obtain the palmprint image, and the face image is subjected to image preprocessing to obtain the face image.

8. The method according to claim 7, characterized in that The performing image preprocessing on the palm image to obtain the palmprint image comprises: Inputting the palm image into a palm detector to obtain palm frame selection data, where the palm frame selection data is used to indicate an area corresponding to the palm in the palm image; Cropping the palm image based on the palm frame selection data to obtain a palm middle image; Inputting the palm middle image into a palm key point detection model to obtain palm key points in the palm middle image; Based on the palm coordinate system corresponding to the palm key points, determine the palm ROI area, where the palm ROI area is an area corresponding to the palm and includes palm lines; The image corresponding to the palm ROI area is determined as the palmprint image.

9. The method according to claim 7, characterized in that: The performing image preprocessing on the facial image to obtain the facial image includes: Inputting the facial image into a face detector to obtain face frame selection data, where the face frame selection data is used to indicate an area corresponding to a face in the facial image; Cropping the facial image based on the facial frame selection data to obtain an intermediate facial image; Inputting the face intermediate image into a face key point detection model to obtain face key points in the face intermediate image; A face ROI region is determined based on the face key points, and an image corresponding to the face ROI region is determined as the face image. The face ROI region is a region containing face features in the region corresponding to the face.

10. The method according to any one of claims 7 to 9, characterized in that: The terminal is also provided with an infrared camera, which is used to collect palm vein images; The sending the palm print image and the face image to the server includes: Inputting the palmprint image and the face image into a quality assessment model to obtain a palmprint quality score and a face quality score, wherein the quality score is used to indicate the feature recognizability of the image; In response to the palm print quality score and the face quality score both being higher than a quality score threshold, inputting the palm vein image into a liveness recognition model to obtain a liveness recognition result of the liveness recognition model, wherein the liveness recognition result is used to indicate whether the image acquisition object is a live object or a non-live object; In response to the living body recognition result that the image acquisition object is a living body, the palm print image and the face image are sent to the server.

11. The method according to any one of claims 6 to 9, characterized in that: The collecting of palm print images and face images by using a camera includes: In response to detecting a palm through the camera, determining that a payment instruction has been received, capturing an image of the palm through the camera; Performing image preprocessing on the palm image to obtain the palmprint image; Reusing the camera to collect facial images; Perform image preprocessing on the facial image to obtain the facial image.

12. A payment verification device, characterized in that: The device comprises: An acquisition module, used for acquiring palm print images and face images; A first determination module is used to determine at least two first candidate accounts based on the palm print image, and to determine at least two second candidate accounts based on the face image, wherein the palm print similarity corresponding to the first candidate accounts is higher than the palm print similarity corresponding to the other accounts, and the face similarity corresponding to the second candidate accounts is higher than the face similarity corresponding to the other accounts; A second determination module is configured to query a first mapping table to determine a first similarity interval corresponding to the palm print similarity of the first candidate account, wherein the first mapping table includes a correspondence between a similarity interval and a first false recognition rate; determine the first false recognition rate based on the first similarity interval and the first mapping table; query a second mapping table to determine a second similarity interval corresponding to the face similarity of the second candidate account, wherein the second mapping table includes a correspondence between a similarity interval and a second false recognition rate; determine the second false recognition rate based on the second similarity interval and the second mapping table; a third determination module, configured to determine an account misrecognition rate of the candidate account based on the first misrecognition rate, the second misrecognition rate, and an account type of the candidate account, the account type including an intersection account and a non-intersection account, the intersection account belonging to both the first candidate account and the second candidate account; and determine a target account from the candidate account based on the account misrecognition rate; A payment module is used to make payment based on the target account.

13. A payment verification device, characterized in that: The device comprises: An image acquisition module, used for acquiring palm print images and face images through a camera; a sending module, configured to send the palm print image and the face image to a server, wherein the server is configured to determine at least two first candidate accounts based on the palm print image and at least two second candidate accounts based on the face image, query a first mapping table to determine a first similarity interval corresponding to the palm print similarity of the first candidate account, wherein the first mapping table includes a correspondence between the similarity interval and a first false recognition rate; determine the first false recognition rate based on the first similarity interval and the first mapping table; query a second mapping table to determine a second similarity interval corresponding to the face similarity of the second candidate account, wherein the second mapping table includes a correspondence between the similarity interval and a second false recognition rate; determine the second false recognition rate based on the second similarity interval and the second mapping table, determine the account false recognition rate of the candidate account based on the first false recognition rate, the second false recognition rate and the account type of the candidate account, wherein the account type includes an intersection account and a non-intersection account, wherein the intersection account belongs to both the first candidate account and the second candidate account; determine a target account from the candidate account based on the account false recognition rate, and make payment based on the target account; The display module is used to receive the payment result sent by the server and display the payment result interface based on the payment result, wherein the payment result is used to indicate whether the payment of the account to be paid is successful or failed.

14. A terminal, characterized in that: The terminal includes a processor and a memory; the memory stores at least one instruction, at least one program, a code set or an instruction set, and the at least one instruction, the at least one program, the code set or the instruction set is loaded and executed by the processor to implement the payment verification method as described in any one of claims 1 to 5.

15. A server, characterized in that: The server includes a processor and a memory; the memory stores at least one instruction, at least one program, a code set or an instruction set, and the at least one instruction, the at least one program, the code set or the instruction set is loaded and executed by the processor to implement the payment verification method as described in any one of claims 6 to 11.

16. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores at least one computer program, and the computer program is loaded and executed by a processor to implement the payment verification method according to any one of claims 1 to 5, or the payment verification method according to any one of claims 6 to 11.

17. A computer program product, comprising computer instructions, which are stored in a computer-readable storage medium; a processor of a terminal reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the terminal executes the payment verification method as described in any one of claims 1 to 5, or a processor of a server reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the server executes the payment verification method as described in any one of claims 6 to 11.

Images