Identity identification method, device and equipment and storage medium

Abstract

The present disclosure relates to an identity recognition method and device, equipment and storage medium, and relates to the technical field of Internet. The method comprises: in response to an identity recognition request carrying a certificate image, obtaining an image feature of the certificate image; calling a prediction model of each of a plurality of identity verification platforms based on the image feature, the prediction model being used to predict the difference between an identity recognition result output by the identity verification platform and a real recognition result; obtaining a prediction value corresponding to each of the plurality of identity verification platforms based on the prediction model of each of the plurality of identity verification platforms, the prediction value being used to represent the predicted difference between the identity recognition result output by the identity verification platform for the image feature and the real recognition result of the image feature; selecting a target identity verification platform with a prediction value meeting a condition from the plurality of identity verification platforms for identity recognition, thereby reducing the rate of false positives when identity recognition is performed through the identity verification platform, improving the success rate of identity recognition, and thus improving the efficiency of identity recognition.

Description

Technical Field

[0001] This disclosure relates to the field of Internet technology, and in particular to an identity recognition method, apparatus, device, and storage medium. Background Technology

[0002] Platforms requiring real-name authentication typically integrate with multiple identity verification platforms, relying on third-party authentication platforms to achieve real-name authentication. In related technologies, during identity verification, an authentication platform is usually randomly selected, and this platform automatically identifies the object based on the document image. However, during automatic identification, the authentication platform may make false positives. These false positives refer to situations where a document image that appears normal to a human might be interpreted by the authentication platform as having glare, incomplete borders, or other defects, leading to recognition failure and reduced efficiency. Summary of the Invention

[0003] This disclosure provides an identity recognition method, apparatus, device, and storage medium. The method reduces the false positive rate when performing identity recognition through an identity verification platform, thereby improving the success rate of identity recognition and ultimately increasing its efficiency. The technical solution of this disclosure is as follows:

[0004] According to a first aspect of the present disclosure, an identity recognition method is provided, the method comprising:

[0005] In response to an identity verification request carrying an image of identification documents, the image features of the identification document image are obtained;

[0006] Based on the image features, the prediction models of multiple authentication platforms are invoked. The prediction models are used to predict the difference between the identity recognition results output by the authentication platforms and the actual recognition results.

[0007] Based on the prediction models of the multiple authentication platforms, prediction values ​​corresponding to the multiple authentication platforms are obtained respectively. The prediction values ​​are used to represent the predicted difference between the identity recognition result output by the authentication platform for the image feature and the actual recognition result of the image feature.

[0008] Select a target authentication platform from the multiple authentication platforms whose predicted value meets the criteria, and then perform identity verification.

[0009] According to a second aspect of the present disclosure, an identity recognition device is provided, the device comprising:

[0010] The first acquisition unit is configured to acquire image features of the document image in response to an identity recognition request carrying a document image;

[0011] The calling unit is configured to call the prediction models of multiple authentication platforms based on the image features. The prediction models are used to predict the difference between the identity recognition results output by the authentication platforms and the actual recognition results.

[0012] The first determining unit is configured to obtain prediction values ​​corresponding to the multiple authentication platforms based on their respective prediction models. The prediction values ​​are used to represent the predicted difference between the identity recognition result output by the authentication platform for the image feature and the actual recognition result of the image feature.

[0013] The identification unit is configured to select a target authentication platform from the plurality of authentication platforms whose predicted value meets the conditions, and to perform identity recognition.

[0014] In some embodiments, the identification unit is configured to: sort the plurality of authentication platforms based on the predicted values ​​corresponding to the plurality of authentication platforms respectively, wherein the order of each authentication platform is negatively correlated with the predicted value corresponding to the authentication platform; if the authentication platform ranked first is available, the authentication platform ranked first is used as the target authentication platform, and identity recognition is performed based on the target authentication platform.

[0015] In some embodiments, the identification unit is further configured to: if the first-ranked authentication platform is unavailable, use the next-ranked authentication platform as the target authentication platform and perform identity recognition based on the target authentication platform.

[0016] In some embodiments, the apparatus further includes:

[0017] The sending unit is configured to send a platform verification request carrying a platform identifier to the first-ranked authentication platform, the platform identifier being used to identify the target platform to which the platform verification request is sent;

[0018] The second determining unit is configured to determine that the top-ranked authentication platform is unavailable if it receives a prompt message from the top-ranked authentication platform indicating that the top-ranked authentication platform is unavailable. The top-ranked authentication platform is used to return a prompt message if it is determined that the platform identifier is not in the platform list based on the platform identifier. The platform list is used to store platform identifiers of platforms that can use the top-ranked authentication platform.

[0019] In some embodiments, the calling unit is configured to call the prediction model of each of the multiple authentication platforms other than the unavailable authentication platform if any of the multiple authentication platforms is unavailable.

[0020] In some embodiments, the apparatus further includes a rejection unit configured to refuse to provide services to the object that triggered the identity verification request if the identity verification result returned by the authentication platform indicates successful identity verification and the document image is obtained based on a document copy.

[0021] In some embodiments, the prediction model is a regression model with image features as independent variables and predicted values ​​as dependent variables.

[0022] In some embodiments, the apparatus further includes:

[0023] The second acquisition unit is configured to acquire multiple sets of sample data pairs for each of the authentication platforms. The sample data pairs include sample image features, sample identity recognition results output by the authentication platform for the sample image features, and sample real recognition results for the sample image features.

[0024] The establishment unit is configured to establish a regression model based on the result difference and sample image features corresponding to the multiple sets of sample data pairs, and obtain regression parameters and error parameters. The result difference is obtained based on the sample identity recognition result and the sample real recognition result in the sample data pair.

[0025] The third determining unit is configured to use the regression model determined based on the regression parameters and the error parameters as the prediction model of the identity verification platform.

[0026] According to a third aspect of the present disclosure, an electronic device is provided, the electronic device comprising:

[0027] One or more processors;

[0028] Memory used to store the executable program code of the processor;

[0029] The processor is configured to execute the program code to implement the aforementioned identity recognition method.

[0030] According to a fourth aspect of the present disclosure, a computer-readable storage medium is provided, which enables the electronic device to perform the above-described identity recognition method when the program code in the computer-readable storage medium is executed by a processor of an electronic device.

[0031] According to a fifth aspect of the present disclosure, a computer program product is provided, including a computer program that, when executed by a processor, implements the above-described identity recognition method.

[0032] This disclosure provides an identity recognition method. The method acquires image features of an ID card image and obtains prediction values ​​corresponding to each of the multiple identity verification platforms through their respective prediction models. Since the prediction value represents the predicted difference between the identity recognition result output by the identity verification platform for the image features and the actual recognition result, a target identity verification platform whose prediction value meets the conditions is selected from the multiple identity verification platforms for identity recognition. This maximizes the consistency between the identity recognition result output by the target identity verification platform and the actual recognition result, reduces the false positive rate when performing identity recognition through the identity verification platform, and thus improves the success rate of identity recognition, thereby improving the efficiency of identity recognition.

[0033] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this disclosure. Attached Figure Description

[0034] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this disclosure and, together with the description, serve to explain the principles of this disclosure, and are not intended to unduly limit this disclosure.

[0035] Figure 1 This is a schematic diagram illustrating an implementation environment according to an exemplary embodiment.

[0036] Figure 2 This is a flowchart illustrating an identity recognition method according to an exemplary embodiment.

[0037] Figure 3 This is a flowchart illustrating a predictive model generation method according to an exemplary embodiment.

[0038] Figure 4 This is a flowchart illustrating another identity recognition method according to an exemplary embodiment.

[0039] Figure 5 This is a flowchart illustrating another identity recognition method according to an exemplary embodiment.

[0040] Figure 6 This is a block diagram illustrating an identity recognition device according to an exemplary embodiment.

[0041] Figure 7 This is a block diagram illustrating a terminal according to an exemplary embodiment.

[0042] Figure 8This is a block diagram illustrating a server according to an exemplary embodiment. Detailed Implementation

[0043] To enable those skilled in the art to better understand the technical solutions of this disclosure, the technical solutions in the embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings.

[0044] It should be noted that the terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this disclosure are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this disclosure described herein can be implemented in orders other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this disclosure. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this disclosure as detailed in the appended claims.

[0045] It should be noted that all information (including but not limited to user device information, user personal information, etc.), data (including but not limited to data used for analysis, stored data, displayed data, etc.), and signals involved in this application have been authorized by the user or fully authorized by all parties, and the collection, use, and processing of related data must comply with the relevant laws, regulations, and standards of the relevant countries and regions. For example, the sample data involved in this application were all obtained under full authorization.

[0046] The identity recognition method provided in this embodiment is executed by an electronic device, which serves as a server. Figure 1 This is a schematic diagram of an implementation environment provided in this embodiment of the disclosure. See also: Figure 1 The implementation environment includes: terminal 101, server 102, and multiple authentication platforms 103. In this embodiment, server 102 is the backend server of terminal 101, and authentication platform 103 is a third-party service platform providing identity verification services. Terminal 101 is used to trigger an identity verification request, and server 102 is connected to the multiple authentication platforms 103 to perform identity verification based on the identity verification request through the multiple authentication platforms 103.

[0047] Terminal 101 can be at least one of the following devices: smartphone, smartwatch, desktop computer, laptop, virtual reality terminal, augmented reality terminal, wireless terminal, and laptop computer. Terminal 101 has communication capabilities and can access wired or wireless networks. Terminal 101 can refer to one of multiple terminals, and those skilled in the art will understand that the number of such terminals can be more or less. Server 102 can be an independent physical server, a server cluster composed of multiple physical servers, or a distributed file system. It can also be a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN (Content Delivery Network), and big data and artificial intelligence platforms. Authentication platform 103 can be an independent physical server, a server cluster composed of multiple physical servers, or a distributed file system. It can also be a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN (Content Delivery Network), and big data and artificial intelligence platforms. In some embodiments, server 102 and terminal 101 are directly or indirectly connected via wired or wireless communication, and this disclosure does not limit this. Optionally, the number of servers 102 may be more or less, and this disclosure does not limit this. Of course, server 102 may also include other functional servers to provide more comprehensive and diversified services. Server 102 undertakes the main computing work, and terminal 101 undertakes secondary computing work; or, server 102 undertakes secondary computing work, and terminal 101 undertakes the main computing work; or, server 102 or terminal 101 can each undertake computing work independently, and this disclosure does not limit this.

[0048] Figure 2 This is a flowchart illustrating an identity recognition method according to an exemplary embodiment, such as... Figure 2 As shown, this method is executed by the server and includes the following steps:

[0049] In step S201, the server responds to the identity recognition request carrying the image of the document and obtains the image features of the document image.

[0050] In this embodiment of the disclosure, the identity verification request is used to request identity verification based on the document image. The document image includes images of identity documents such as ID cards, passports, and driver's licenses that can be used for identity verification. The image features of the document image are feature vectors used to characterize the document image.

[0051] In step S202, the server calls the prediction models of multiple authentication platforms based on image features. These prediction models are used to predict the difference between the identity recognition result output by the authentication platform and the actual recognition result.

[0052] In this embodiment of the disclosure, the multiple authentication platforms are third-party service platforms that provide identity recognition services. The prediction model of any authentication platform is obtained based on the differences between the historical outputs of the authentication platform for multiple image features and the actual recognition results, as well as the multiple image features themselves.

[0053] In this embodiment, the identity recognition result includes any one of the following: a quality failure result, a successful recognition result, or a recognition failure result. The quality failure result includes, but is not limited to, issues such as glare, blurriness, incomplete borders, or counterfeit documents in the document image. The actual recognition result is the result of manual review of the document image.

[0054] In step S203, the server obtains the prediction values ​​corresponding to the multiple authentication platforms based on their respective prediction models. These prediction values ​​are used to represent the predicted difference between the identity recognition result output by the authentication platform for the image feature and the actual recognition result of the image feature.

[0055] In this embodiment of the disclosure, for each authentication platform, the server inputs the image features into the prediction model of that authentication platform to obtain the prediction value corresponding to that authentication platform. This prediction value is positively correlated with the prediction difference between the two recognition results; that is, the larger the prediction value, the larger the prediction difference, and vice versa.

[0056] In step S204, the server selects a target authentication platform from multiple authentication platforms whose predicted value meets the criteria, and performs identity verification.

[0057] In this embodiment of the disclosure, the predicted value meets the condition that the difference between the identity recognition result output by the identity verification platform and the actual recognition result is minimized.

[0058] This disclosure provides an identity recognition method. The method acquires image features of an ID card image and obtains prediction values ​​corresponding to each of the multiple identity verification platforms through their respective prediction models. Since the prediction value represents the predicted difference between the identity recognition result output by the identity verification platform for the image features and the actual recognition result, a target identity verification platform whose prediction value meets the conditions is selected from the multiple identity verification platforms for identity recognition. This maximizes the consistency between the identity recognition result output by the target identity verification platform and the actual recognition result, reduces the false positive rate when performing identity recognition through the identity verification platform, and thus improves the success rate of identity recognition, thereby improving the efficiency of identity recognition.

[0059] In some embodiments, selecting a target authentication platform whose predicted value meets the criteria from multiple authentication platforms for identity recognition includes: sorting the multiple authentication platforms based on their respective predicted values, wherein the order of each authentication platform is negatively correlated with its predicted value; if the authentication platform ranked first is available, the authentication platform ranked first is selected as the target authentication platform, and identity recognition is performed based on the target authentication platform.

[0060] In this embodiment of the disclosure, the identity verification platform ranked first is selected as the target identity verification platform for identity recognition. Since the predicted value of the identity verification platform ranked first is the smallest, the selected target identity verification platform ensures the consistency between its output identity recognition result and the actual recognition result to the greatest extent, reduces the false positive rate of the identity verification platform, and thus improves the identity recognition efficiency.

[0061] In some embodiments, the identity recognition method further includes: if the first-ranked identity verification platform is unavailable, using the next-ranked identity verification platform as the target identity verification platform, and performing identity recognition based on the target identity verification platform.

[0062] In this embodiment of the disclosure, if the first-ranked authentication platform is unavailable, the next-ranked authentication platform is selected as the target authentication platform. Since the next-ranked authentication platform is the authentication platform with the smallest predicted value other than the first-ranked one, this ensures the rationality of the selected authentication platform to the greatest extent and avoids repeated identification failures caused by continuously requesting an unavailable authentication platform for identity recognition.

[0063] In some embodiments, the process of determining that the top-ranked authentication platform is unavailable includes: sending a platform verification request carrying a platform identifier to the top-ranked authentication platform, the platform identifier being used to identify the target platform sending the platform verification request; if a prompt message indicating that the top-ranked authentication platform is unavailable is received from the top-ranked authentication platform, the process of determining that the top-ranked authentication platform is unavailable is determined, wherein the top-ranked authentication platform is used to return the prompt message if it is determined based on the platform identifier that the platform identifier is not in the platform list, and the platform list is used to store the platform identifiers of platforms that can use the top-ranked authentication platform.

[0064] In this embodiment of the disclosure, the use of the target platform is verified by sending a platform verification request to the authentication platform, thereby effectively determining whether the authentication platform is available.

[0065] In some embodiments, based on image features, the prediction models of multiple authentication platforms are invoked, including: if any authentication platform is unavailable, the prediction models of the authentication platforms other than the unavailable authentication platform are invoked.

[0066] In this embodiment of the disclosure, only the available authentication platform is invoked, thus avoiding the waste of resources caused by invoking an unavailable authentication platform.

[0067] In some embodiments, the identity verification method further includes: if the identity verification result returned by the identity verification platform indicates that the identity verification was successful and the document image is obtained based on a copy of the document, refusing to provide services to the object that triggered the identity verification request.

[0068] In this embodiment of the disclosure, services are refused for document images obtained from document photocopies, thus ensuring the security and rigor of identity verification.

[0069] In some embodiments, the prediction model is a regression model with image features as independent variables and predicted values ​​as dependent variables.

[0070] In this embodiment of the disclosure, a regression model with image features as independent variables and predicted values ​​as dependent variables is used as the prediction model. Based on the image features, the prediction model can quickly obtain the predicted values, thus improving the efficiency of obtaining the predicted values.

[0071] In some embodiments, the identity recognition method further includes: for each authentication platform, acquiring multiple sets of sample data pairs, the sample data pairs including sample image features, sample identity recognition results output by the authentication platform for the sample image features, and sample real recognition results for the sample image features; establishing a regression model based on the result difference and sample image features corresponding to the multiple sets of sample data pairs respectively, and obtaining regression parameters and error parameters, the result difference being obtained based on the sample identity recognition results and sample real recognition results in the sample data pairs; and using the regression model determined based on the regression parameters and error parameters as the prediction model of the authentication platform.

[0072] In this embodiment of the disclosure, for each authentication platform, a regression model is established based on multiple sets of sample data for the corresponding sample image features and result differences. This yields a regression model with image features as independent variables and predicted values ​​as dependent variables, thus learning the relationship between image features and result differences. This regression model is then used as the prediction model for the authentication platform. Based on image features, this prediction model can predict the difference between the identity recognition result output by the authentication platform and the actual recognition result. Based on this difference, the optimal authentication platform can be selected, reducing the false positive rate when performing identity recognition through the authentication platform.

[0073] This disclosure provides an identity recognition method. The method acquires image features of an ID card image and obtains predicted values ​​for each of the multiple identity verification platforms using their respective prediction models. Since the predicted value represents the predicted difference between the identity recognition result output by the identity verification platform for the image features and the actual recognition result, a target identity verification platform whose predicted value meets the conditions is selected from the multiple identity verification platforms for identity recognition. This maximizes the consistency between the identity recognition result output by the target identity verification platform and the actual recognition result, reduces the false positive rate when the identity verification platform performs identity recognition, and thus improves the success rate of identity recognition, thereby improving the efficiency of identity recognition.

[0074] The above Figure 2 This is the basic process for identity recognition. The embodiments of this disclosure achieve identity recognition through a prediction model. The following is based on... Figure 3 The process of generating the prediction model for each authentication platform is explained. See [link / reference]. Figure 3 , Figure 3 This is a flowchart illustrating a predictive model generation method according to an exemplary embodiment. The method is executed by a server and includes the following steps:

[0075] In step S301, the server obtains multiple sets of sample data pairs for each of the multiple authentication platforms. Each sample data pair includes sample image features, the sample identity recognition result output by the authentication platform for the sample image features, and the sample real recognition result for the sample image features.

[0076] In this embodiment of the disclosure, multiple sets of sample data from any authentication platform are obtained based on the historical identity recognition results output by the authentication platform. Optionally, the server obtains the historical sample identity recognition results and the sample document images corresponding to the sample identity recognition results from the authentication platform, and then obtains the sample true recognition results based on the manual review of the sample document images, and extracts the sample image features of the sample document images to obtain a set of sample data pairs including sample image features, sample identity recognition results, and sample true recognition results.

[0077] In step S302, the server establishes a regression model based on the result difference and sample image features corresponding to multiple sets of sample data to obtain regression parameters and error parameters. The result difference is obtained based on the sample identity recognition result and the sample real recognition result in the sample data pair.

[0078] In this embodiment of the disclosure, the process of determining the result difference includes the following steps: the terminal extracts feature vectors for the sample identity recognition result and the sample real recognition result respectively, and the absolute value of the difference between the two extracted feature vectors is taken as the result difference. The two feature vectors are used to characterize the sample identity recognition result and the sample real recognition result respectively.

[0079] For example, the server builds a regression model based on the following formula (1).

[0080] s(x)=||s(sample real recognition result)-s(sample identity recognition result)|= θ0+θ1x1+ε(1)

[0081] Where s(x) represents the result difference, i.e., the dependent variable, and is the absolute value; s(sample true recognition result) represents the feature vector corresponding to the sample true recognition result; s(sample identity recognition result) represents the feature vector corresponding to the sample identity recognition result; θ0, θ1 represent the regression parameters to be determined; x1 represents the sample image features; and ε represents the error parameters to be determined.

[0082] In one implementation, the server determines the regression parameters and error parameters using the least squares method. Accordingly, the server determines the regression parameters and error parameters using the following formula (2).

[0083] Y = Xβ + ε(2)

[0084] in, This represents the difference in results, where n represents the vector dimension. Representing image features; This represents the regression parameters to be calculated, where β0 and β1 each represent a regression coefficient. This represents the error parameter to be calculated, which is a matrix of random variables of the error term.

[0085] The server substitutes the result difference and sample image features corresponding to multiple sets of sample data into the above formula (2) to establish a multiple regression linear equation and solve it to obtain the regression parameters and error parameters. The regression parameters represent the degree of influence of the independent variable on the dependent variable.

[0086] In step S303, the server uses the regression model determined based on the regression parameters and error parameters as the prediction model for the identity verification platform.

[0087] In this embodiment, the server substitutes the regression parameters and error parameters into the above formula (1) to obtain a regression model, which is used as the prediction model of the identity verification platform. That is, the prediction model is a regression model with image features as independent variables and predicted values ​​as dependent variables. The regression parameters represent the degree of influence of the independent variables on the dependent variables, that is, the regression parameters represent the degree of influence of image features on the predicted values.

[0088] For example, for any two authentication platforms, the first authentication platform and the second authentication platform, their prediction models are shown in formulas (3) and (4) respectively.

[0089] s(t)=|s(manual_result t )-s(provider_result t )|=t0+t1x t +ε t (3)

[0090] Where t represents the first authentication platform, s(t) represents the predicted value to be output by the first authentication platform, and s(provider_result) t ) represents the identity verification result output by the first identity verification platform, s(manual_result t ) represents the actual recognition result, t0 and t1 represent the regression parameters in the prediction model of the first identity verification platform, and ε t x represents the error parameter in the prediction model of the first authentication platform. t This represents the image features of the prediction model to be input into the first authentication platform.

[0091] s(a)=|s(manual_result a )-s(provider_resulta )|=a0+a1x a +ε a (4)

[0092] Where 'a' represents the second authentication platform, 's(a)' represents the predicted value to be output by the second authentication platform, and 's(provider_result)' represents the predicted value to be output by the second authentication platform. a ) represents the identity verification result output by the second identity verification platform, s(manual_result a ) represents the actual recognition result, a0 and a1 represent the regression parameters in the prediction model of the second identity verification platform, and ε a x represents the error parameter in the prediction model of the second authentication platform. a This represents the image features of the prediction model to be input into the second authentication platform.

[0093] It should be noted that the image features of any document image can characterize the quality features of that document image, including at least one of brightness, sharpness, and border integrity. Accordingly, the multiple prediction models generated in this embodiment correspond to a class of document images that have the same quality features, meaning that each identity verification platform is suitable for a class of document images with specific quality features. Thus, for any document image, after extracting its image features, the most suitable identity verification platform can be selected from the multiple prediction models for identity recognition.

[0094] In this embodiment of the disclosure, the server obtains the prediction models for each of the multiple authentication platforms based on the above steps S301-S303. It should be noted that the execution entity server in the above steps S301-S303 can be an authentication platform or a platform requesting authentication; that is, each authentication platform can generate its own prediction model, or the platform requesting authentication can generate the prediction models for the multiple authentication platforms.

[0095] In this embodiment of the disclosure, for each authentication platform, a regression model is established based on multiple sets of sample data for the corresponding sample image features and result differences. This yields a regression model with image features as independent variables and predicted values ​​as dependent variables, thus learning the relationship between image features and result differences. This regression model is then used as the prediction model for the authentication platform. Based on image features, this prediction model can predict the difference between the identity recognition result output by the authentication platform and the actual recognition result. Based on this difference, the optimal authentication platform can be selected, reducing the false positive rate when performing identity recognition through the authentication platform.

[0096] The above Figure 3 This is the process of generating predictive models. Figure 2 This is the basic process for identity verification, and the following is based on... Figure 4 The process of identity verification will be further elaborated. Among other things, Figure 2 and Figure 4 Each embodiment is passed Figure 3 The resulting prediction model is implemented. See also... Figure 4 , Figure 4 This is a flowchart illustrating an identity recognition method according to an exemplary embodiment, the method being executed by a server, and the method includes the following steps.

[0097] In step S401, the server responds to the identity recognition request carrying the image of the document and obtains the image features of the document image.

[0098] In some embodiments, the authentication request is triggered based on a target application installed on the terminal. This target application includes a target function that requires real-name authentication to function, and the authentication request is triggered based on that target function. For example, the target function could be a ticketing function, a payment function, etc.

[0099] In this embodiment of the disclosure, the server extracts feature vectors from the document image to obtain the image features of the document image.

[0100] In step S402, the server, based on the image features, calls the prediction models of multiple authentication platforms. These prediction models are used to predict the difference between the identity recognition results output by the authentication platform and the actual recognition results.

[0101] In this embodiment of the disclosure, some of the multiple authentication platforms may be currently unavailable. To avoid wasting resources, in some embodiments, if all authentication platforms are available, the server calls the prediction model of each of the multiple authentication platforms; if any of the multiple authentication platforms is unavailable, the server calls the prediction model of each of the multiple authentication platforms other than the unavailable authentication platform. In this way, only the available authentication platforms are called, avoiding the resource waste caused by calling the unavailable authentication platforms.

[0102] In some embodiments, the server determines unavailable authentication platforms based on the previous identity verification process. When performing identity verification based on any authentication platform, if it is unavailable, the authentication platform will return an unavailable message to the server. Reasons for unavailability include the server not having granted permission to use the authentication platform, reaching the maximum number of uses of the authentication platform, or insufficient resource balance for using the authentication platform.

[0103] The previous identity verification process can be the identity verification process of the target number most recent time, the last identity verification process, or multiple identity verification processes within the target time period most recent time period. By controlling the scope, the system avoids continuously deeming a certain identity verification platform unavailable, thus ensuring the timeliness of the data.

[0104] In one implementation, the server stores the prediction models for each of the multiple authentication platforms in a decision tree model. The server performs predictions by calling multiple prediction models within this decision tree model. Optionally, the server inputs the image features and the platform identifiers of authentication platforms other than the unavailable authentication platform into the decision tree model. The decision tree model then calls the available authentication platforms based on these platform identifiers, thus improving the convenience and efficiency of calling authentication platforms. In this embodiment, the server pre-acquires and stores these multiple prediction models to improve the efficiency of subsequently selecting an authentication platform based on the prediction models.

[0105] In step S403, the server obtains prediction values ​​corresponding to each of the multiple authentication platforms based on their respective prediction models. These prediction values ​​are used to represent the predicted difference between the identity recognition results for image features output by the authentication platform and the actual recognition results for image features.

[0106] In this embodiment of the disclosure, the server inputs the image features into the prediction models of multiple authentication platforms respectively, and outputs the prediction values ​​corresponding to the multiple authentication platforms through the multiple prediction models.

[0107] In this embodiment of the disclosure, the prediction model is a regression model with image features as independent variables and predicted values ​​as dependent variables. By using a regression model with image features as independent variables and predicted values ​​as dependent variables as the prediction model, the predicted values ​​can be obtained quickly based on the image features, thus improving the efficiency of obtaining the predicted values.

[0108] For example, for any two authentication platforms among multiple authentication platforms, the first authentication platform and the second authentication platform, their prediction models are as shown in formulas (3) and (4) above, respectively. After inputting the image features into the prediction models of these two authentication platforms, the predicted values ​​are obtained as shown in formulas (5) and (6) below.

[0109] s(t)=|s(manual_result t )-s(provider_result t )|=t0+t1pic1+ε t (5)

[0110] s(a)=|s(manual_result a )-s(provider_result a )|=a0+a1pic1+ε a (6)

[0111] Where pic1 represents the image features, and s(t) and s(a) represent the predicted values ​​obtained, respectively.

[0112] In step S404, the server sorts the multiple authentication platforms based on the predicted values ​​corresponding to each authentication platform, with the order of each authentication platform being negatively correlated with the predicted value corresponding to that authentication platform.

[0113] In this embodiment of the disclosure, the order of each authentication platform is negatively correlated with the predicted value corresponding to that authentication platform; that is, the smaller the predicted value of any authentication platform, the earlier it is in the order, and the larger the predicted value, the later it is in the order. For example, taking s(t) and s(a) as examples, if s(t) > s(a), then the first authentication platform is ordered after the second authentication platform.

[0114] In this embodiment of the disclosure, the authentication platform ranked first has the smallest predicted value, therefore it is preferentially selected for identity verification. However, in some cases, the authentication platform ranked first may be unavailable. Accordingly, if the authentication platform ranked first is available, step S405 is executed; if the authentication platform ranked first is unavailable, step S406 is executed.

[0115] In step S405, if the authentication platform ranked first is available, the server will use the authentication platform ranked first as the target authentication platform and perform identity recognition based on the target authentication platform.

[0116] For example, continuing with s(t) and s(a), if the multiple authentication platforms include only the first authentication platform and the second authentication platform, and s(t) > s(a), it means that the predicted value corresponding to the first authentication platform is greater than the predicted value corresponding to the second authentication platform. Then the authentication platform ranked first is the second authentication platform. If the second authentication platform is available, then the second authentication platform will be used as the target authentication platform for identity recognition.

[0117] In this embodiment of the disclosure, the above steps S404-S405 realize the process of selecting a target authentication platform with a predictive value that meets the conditions from multiple authentication platforms for identity recognition. This embodiment uses the authentication platform that is ranked first and available as the target authentication platform for identity recognition. Since the authentication platform ranked first has the smallest predicted value, the selected target authentication platform ensures the consistency between its output identity recognition result and the actual recognition result to the greatest extent, reduces the false positive rate of the authentication platform, and thus improves the efficiency of identity recognition.

[0118] In this embodiment of the disclosure, the process of the server performing identity verification based on the target authentication platform includes the following steps: the server sends an identity verification request carrying an image of an identification document to the target authentication platform; the target authentication platform performs identity verification based on the image of the identification document and returns the identity verification result to the server. Optionally, the target authentication platform performs identity verification based on OCR (Optical Character Recognition) technology.

[0119] In this embodiment, the process by which the server determines that the top-ranked authentication platform is available includes the following steps: the server sends an authentication request carrying a platform identifier to the top-ranked authentication platform. This platform identifier identifies the target platform sending the platform authentication request, i.e., the server itself. If the server receives confirmation information from the top-ranked authentication platform confirming its availability, it determines that the top-ranked authentication platform is available. This top-ranked authentication platform returns confirmation information if it determines that the platform identifier is in the platform list based on that identifier. The platform list stores the platform identifiers of platforms capable of using the top-ranked authentication platform. This embodiment verifies whether the target platform has permission to use the authentication platform by sending a platform authentication request to the authentication platform, thereby effectively determining whether the authentication platform is available.

[0120] In some embodiments, when determining whether any authentication platform is available, the server sends a platform verification request carrying the platform identifier to the authentication platform. After receiving confirmation information from the authentication platform, the server then sends an identity recognition request carrying an image of the identification document. This avoids wasting transmission resources due to the authentication platform being unavailable after sending the identification document image. In other embodiments, the server sends an identity recognition request carrying both the platform identifier and the identification document image to the authentication platform. If the target authentication platform is available, identity recognition can be performed directly based on the identification document image, avoiding back-and-forth data transmission and improving identity recognition efficiency.

[0121] In step S406, if the authentication platform ranked first is unavailable, the server will use the next ranked authentication platform as the target authentication platform and perform identity verification based on that target authentication platform.

[0122] In this embodiment of the disclosure, if the first-ranked authentication platform is unavailable, the next-ranked authentication platform is selected as the target authentication platform. Since the next-ranked authentication platform is the authentication platform with the smallest predicted value other than the first-ranked one, this ensures the rationality of the selected authentication platform to the greatest extent and avoids repeated identification failures caused by continuously requesting an unavailable authentication platform for identity recognition.

[0123] In this embodiment, the process by which the server determines that the top-ranked authentication platform is unavailable includes the following steps: the server sends a platform verification request carrying a platform identifier to the top-ranked authentication platform; if the server receives a prompt message from the top-ranked authentication platform indicating that the top-ranked authentication platform is unavailable, the server determines that the top-ranked authentication platform is unavailable. This top-ranked authentication platform is used to return the prompt message when it is determined that the platform identifier is not in the platform list based on the platform identifier. This embodiment verifies whether the target platform has permission to use the authentication platform by sending a platform verification request to the authentication platform, thereby effectively determining whether the authentication platform is available.

[0124] In some embodiments, the prompt message also carries the reason why the top-ranked authentication platform is unavailable. Accordingly, if the top-ranked authentication platform determines that its platform identifier is not in the platform list, it further queries multiple unavailable platform lists based on the platform identifier. These multiple unavailable platform lists are used to store platform identifiers of different types of platforms that are unavailable to the top-ranked authentication platform, such as multiple unavailable platform lists used to store platform identifiers of platforms that have not been granted usage rights, have reached their usage limit, or have insufficient balance.

[0125] It should be noted that the availability of any authentication platform to be used must first be confirmed; that is, the availability of the next authentication platform in the ranking must also be confirmed first. If the next authentication platform in the ranking is available, authentication is performed based on it. If it is unavailable, the above steps are repeated for the next authentication platform selected, until an available authentication platform is selected for authentication.

[0126] In some embodiments, the server only supports identity verification based on document images obtained from the original document itself. Accordingly, if the identity verification platform returns an identity verification result indicating successful verification and the document image is obtained from a copy of the document, the server refuses to provide services to the object triggering the identity verification request, thus ensuring the security and rigor of the identity verification process. In this embodiment, the verification is not limited to document copies; for any process where identity verification is not based on the original document, the server refuses to provide services to the object triggering the identity verification request. For example, if identity verification is performed using a fake or temporary document obtained through techniques such as Photoshop, the server refuses to provide services to the object triggering the identity verification request. The server only provides services to the object triggering the identity verification request if the identity verification platform returns an identity verification result indicating successful verification and the document image is obtained from the original document.

[0127] In this embodiment, if the identity verification platform can determine the source of the document image, it indicates that the platform supports anti-counterfeiting capabilities. The server then determines whether to provide service to the object triggering the identity verification request based on this source. In other embodiments, if the identity verification platform does not support anti-counterfeiting capabilities, the server can determine the source of the document image by performing its own anti-counterfeiting capabilities on the image, even if the identity verification result returned by the platform indicates successful verification but does not indicate the source of the document image. Service will only be provided to the object triggering the identity verification request if the document image originates from the original document.

[0128] In this embodiment, the anti-counterfeiting capability is first identified through an identity verification platform. Only when the identity verification platform does not support the identification of anti-counterfeiting capabilities is the anti-counterfeiting capability identified by itself. This effectively utilizes the resources of the identity verification platform, reduces the service pressure on the server, and improves the flexibility of identifying anti-counterfeiting capabilities.

[0129] In this embodiment of the disclosure, in order to improve the success rate of real-name authentication of objects and reduce real-name problems caused by false identity recognition, multiple identity verification platforms are connected, and a multiple linear regression model is used to predict the difference between the identity recognition results of different identity verification platforms and the actual recognition results, so as to select the best identity verification platform.

[0130] See Figure 5 , Figure 5This is a flowchart illustrating an identity recognition process according to an exemplary embodiment. When an object enters a real-name authentication platform requiring verification, the server inputs the image features of the identification document and the platform identifiers of available authentication platforms into a decision tree model. This decision tree model includes prediction models for multiple authentication platforms. Based on the image features and platform identifiers, the decision tree model calls the prediction models of the available authentication platforms to obtain predicted values ​​for each available authentication platform, and then outputs a selection result for the authentication platform based on these predicted values. If the selected authentication platform is available, an identity recognition result is output based on that platform; otherwise, the process returns to the step of selecting an authentication platform, and a new authentication platform is selected for identity recognition.

[0131] In this embodiment, an identity verification platform is selected based on a predictive model for identity recognition, improving the success rate of real-name authentication and reducing real-name issues caused by false identification of documents. Furthermore, when one identity verification platform is unavailable, another available platform can be automatically selected for identity recognition, instead of continuously retrying with a fixed platform, thus improving the availability of the identity verification platform and consequently increasing the recognition rate of correct document images. This method selects the optimal identity verification platform in real time based on the document image and allows for platform switching, thereby achieving dynamic selection of the identity verification platform.

[0132] In this embodiment, a multiple linear regression model is used to select the identity verification platform suitable for the document image, thereby reducing real-name authentication issues caused by false positives from the identity verification platform. Furthermore, by expanding to multiple identity verification platforms and selecting the one with the lowest predicted value, the impact of false positives from a single platform can be minimized. Additionally, when an identity verification platform becomes unavailable, a fallback mechanism is implemented, i.e., a new identity verification platform is selected, rather than the system exiting abnormally, thus improving the user experience.

[0133] In this embodiment, only a multiple linear regression model is used as an example for illustration. In some embodiments, the prediction model can also be a deep learning model obtained through reinforcement learning, which is used to select the authentication platform and obtain the optimal choice. However, the selection process based on the deep learning model may take a little longer, meaning that the prediction model in this embodiment can reduce the time spent selecting the authentication platform.

[0134] This disclosure provides an identity recognition method. The method acquires image features of an ID card image and obtains prediction values ​​corresponding to each of the multiple identity verification platforms through their respective prediction models. Since the prediction value represents the predicted difference between the identity recognition result output by the identity verification platform for the image features and the actual recognition result, the identity verification platform with the smallest prediction value is selected from the multiple identity verification platforms for identity recognition. This maximizes the consistency between the identity recognition result output by the identity verification platform and the actual recognition result, reduces the false positive rate when the identity verification platform performs identity recognition, and thus improves the success rate of identity recognition, thereby improving the efficiency of identity recognition.

[0135] Figure 6 This is a block diagram illustrating an identity recognition device according to an exemplary embodiment. (Refer to...) Figure 6 The device includes:

[0136] The first acquisition unit 601 is configured to acquire image features of an ID image in response to an identity recognition request carrying an ID image.

[0137] Calling unit 602 is configured to call the prediction models of multiple authentication platforms based on image features. The prediction models are used to predict the difference between the identity recognition results output by the authentication platforms and the actual recognition results.

[0138] The first determining unit 603 is configured to obtain prediction values ​​corresponding to the multiple authentication platforms based on their respective prediction models. The prediction values ​​are used to represent the predicted difference between the identity recognition result output by the authentication platform for image features and the actual recognition result of the image features.

[0139] The identification unit 604 is configured to select a target authentication platform from multiple authentication platforms whose predicted value meets the criteria, and to perform identity recognition.

[0140] In some embodiments, the identification unit 604 is configured to: sort multiple authentication platforms based on the predicted values ​​corresponding to each authentication platform, wherein the order of each authentication platform is negatively correlated with the predicted value corresponding to the authentication platform; if the authentication platform ranked first is available, the authentication platform ranked first is used as the target authentication platform, and identity recognition is performed based on the target authentication platform.

[0141] In some embodiments, the identification unit 604 is further configured to: if the first-ranked authentication platform is unavailable, use the next-ranked authentication platform as the target authentication platform and perform identity recognition based on the target authentication platform.

[0142] In some embodiments, the apparatus further includes:

[0143] The sending unit is configured to send a platform verification request carrying a platform identifier to the first-ranked authentication platform. The platform identifier is used to identify the target platform to which the platform verification request is sent.

[0144] The second determining unit is configured to determine that the top-ranked authentication platform is unavailable if it receives a prompt message from the top-ranked authentication platform indicating that the top-ranked authentication platform is unavailable. The top-ranked authentication platform is used to return a prompt message if it is determined that the platform identifier is not in the platform list based on the platform identifier. The platform list is used to store the platform identifiers of platforms that can use the top-ranked authentication platform.

[0145] In some embodiments, the calling unit 602 is configured to call the prediction model of each of the multiple authentication platforms other than the unavailable authentication platform if any authentication platform among the multiple authentication platforms is unavailable.

[0146] In some embodiments, the apparatus further includes a rejection unit configured to refuse to provide services to the object that triggered the identity verification request if the identity verification result returned by the authentication platform indicates that the identity verification was successful and the document image is obtained based on a copy of the document.

[0147] In some embodiments, the prediction model is a regression model with image features as independent variables and predicted values ​​as dependent variables.

[0148] In some embodiments, the apparatus further includes:

[0149] The second acquisition unit is configured to acquire multiple sets of sample data pairs for each authentication platform. The sample data pairs include sample image features, sample identity recognition results output by the authentication platform for the sample image features, and sample real recognition results for the sample image features.

[0150] The unit is configured to establish a regression model based on the result difference and sample image features corresponding to multiple sets of sample data pairs, and obtain regression parameters and error parameters. The result difference is obtained based on the sample identity recognition result and the sample real recognition result in the sample data pair.

[0151] The third determining unit is configured to use the regression model determined based on regression parameters and error parameters as the predictive model for the identity verification platform.

[0152] This disclosure provides an identity recognition device that acquires image features of an ID card image and obtains predicted values ​​for each of the multiple identity verification platforms using their respective prediction models. Since the predicted value represents the predicted difference between the identity recognition result output by the identity verification platform for the image feature and the actual recognition result, a target identity verification platform whose predicted value meets the conditions is selected for identity recognition. This maximizes the consistency between the identity recognition result output by the target identity verification platform and the actual recognition result, reduces the false positive rate when the identity verification platform performs identity recognition, and thus improves the success rate of identity recognition, thereby improving the efficiency of identity recognition.

[0153] Regarding the apparatus in the above embodiments, the specific manner in which each unit performs its operation has been described in detail in the embodiments related to the method, and will not be elaborated upon here.

[0154] Figure 7 A structural block diagram of a terminal 700 provided in an exemplary embodiment of this disclosure is shown. The terminal 700 may be a smartphone, tablet computer, MP3 player (Moving Picture Experts Group Audio Layer III), MP4 player (Moving Picture Experts Group Audio Layer IV), laptop computer, or desktop computer. The terminal 700 may also be referred to as a user device, portable terminal, laptop terminal, desktop terminal, or other names.

[0155] Typically, terminal 700 includes a processor 701 and a memory 702.

[0156] Processor 701 may include one or more processing cores, such as a quad-core processor, an octa-core processor, etc. Processor 701 may be implemented using at least one hardware form selected from DSP (Digital Signal Processing), FPGA (Field-Programmable Gate Array), and PLA (Programmable Logic Array). Processor 701 may also include a main processor and a coprocessor. The main processor, also known as a CPU (Central Processing Unit), is used to process data in the wake-up state; the coprocessor is a low-power processor used to process data in the standby state. In some embodiments, processor 701 may integrate a GPU (Graphics Processing Unit), which is responsible for rendering and drawing the content to be displayed on the screen. In some embodiments, processor 701 may also include an AI (Artificial Intelligence) processor, which is used to handle computational operations related to machine learning.

[0157] The memory 702 may include one or more computer-readable storage media, which may be non-transitory. The memory 702 may also include high-speed random access memory and non-volatile memory, such as one or more disk storage devices or flash memory devices. In some embodiments, the non-transitory computer-readable storage media in the memory 702 are used to store at least one program code, which is executed by the processor 701 to implement the identity recognition method provided in the method embodiments of this disclosure.

[0158] In some embodiments, the terminal 700 may also optionally include a peripheral device interface 703 and at least one peripheral device. The processor 701, memory 702, and peripheral device interface 703 can be connected via a bus or signal line. Each peripheral device can be connected to the peripheral device interface 703 via a bus, signal line, or circuit board. Specifically, the peripheral device includes at least one of the following: a radio frequency circuit 704, a display screen 705, a camera assembly 706, an audio circuit 707, and a power supply 708.

[0159] Peripheral device interface 703 can be used to connect at least one I / O (Input / Output) related peripheral device to processor 701 and memory 702. In some embodiments, processor 701, memory 702 and peripheral device interface 703 are integrated on the same chip or circuit board; in some other embodiments, any one or two of processor 701, memory 702 and peripheral device interface 703 can be implemented on separate chips or circuit boards, which is not limited in this embodiment.

[0160] The radio frequency (RF) circuit 704 is used to receive and transmit RF (Radio Frequency) signals, also known as electromagnetic signals. The RF circuit 704 communicates with communication networks and other communication devices via electromagnetic signals. The RF circuit 704 converts electrical signals into electromagnetic signals for transmission, or converts received electromagnetic signals back into electrical signals. Optionally, the RF circuit 704 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, etc. The RF circuit 704 can communicate with other terminals via at least one wireless communication protocol. This wireless communication protocol includes, but is not limited to: metropolitan area networks (MANs), various generations of mobile communication networks (2G, 3G, 4G, and 5G), wireless local area networks (WLANs), and / or WiFi (Wireless Fidelity) networks. In some embodiments, the RF circuit 704 may also include circuitry related to NFC (Near Field Communication), which is not limited in this disclosure.

[0161] Display screen 705 is used to display a UI (User Interface). This UI may include graphics, text, icons, videos, and any combination thereof. When display screen 705 is a touch display screen, it also has the ability to collect touch signals on or above its surface. These touch signals can be input as control signals to processor 701 for processing. In this case, display screen 705 can also be used to provide virtual buttons and / or a virtual keyboard, also known as soft buttons and / or a soft keyboard. In some embodiments, there may be one display screen 705, which serves as the front panel of terminal 700; in other embodiments, there may be at least two display screens 705, respectively disposed on different surfaces of terminal 700 or in a folded design; in still other embodiments, display screen 705 may be a flexible display screen, disposed on a curved or folded surface of terminal 700. Furthermore, display screen 705 may be configured as a non-rectangular, irregular shape, i.e., a non-rectangular screen. Display screen 705 may be made of materials such as LCD (Liquid Crystal Display) or OLED (Organic Light-Emitting Diode).

[0162] The camera assembly 706 is used to acquire images or videos. Optionally, the camera assembly 706 includes a front-facing camera and a rear-facing camera. Typically, the front-facing camera is located on the front panel of the terminal, and the rear-facing camera is located on the back of the terminal. In some embodiments, there are at least two rear-facing cameras, which are any one of a main camera, a depth-sensing camera, a wide-angle camera, and a telephoto camera, to achieve background blurring by fusion of the main camera and the depth-sensing camera, panoramic shooting by fusion of the main camera and the wide-angle camera, VR (Virtual Reality) shooting, or other fusion shooting functions. In some embodiments, the camera assembly 706 may also include a flash. The flash can be a single-color temperature flash or a dual-color temperature flash. A dual-color temperature flash refers to a combination of a warm-light flash and a cool-light flash, which can be used for light compensation at different color temperatures.

[0163] The audio circuit 707 may include a microphone and a speaker. The microphone is used to collect sound waves from the user and the environment, converting the sound waves into electrical signals that are input to the processor 701 for processing, or input to the radio frequency circuit 704 for voice communication. For stereo sound acquisition or noise reduction purposes, multiple microphones may be used, each located at a different part of the terminal 700. The microphone may also be an array microphone or an omnidirectional microphone. The speaker is used to convert the electrical signals from the processor 701 or the radio frequency circuit 704 into sound waves. The speaker may be a conventional diaphragm speaker or a piezoelectric ceramic speaker. When the speaker is a piezoelectric ceramic speaker, it can convert electrical signals not only into audible sound waves but also into inaudible sound waves for purposes such as distance measurement. In some embodiments, the audio circuit 707 may also include a headphone jack.

[0164] Power supply 708 is used to power the various components in terminal 700. Power supply 708 can be AC ​​power, DC power, a disposable battery, or a rechargeable battery. When power supply 708 includes a rechargeable battery, the rechargeable battery can support wired charging or wireless charging. The rechargeable battery can also be used to support fast charging technology.

[0165] Those skilled in the art will understand that Figure 7 The structure shown does not constitute a limitation on terminal 700, and may include more or fewer components than shown, or combine certain components, or use different component arrangements.

[0166] Figure 8 This is a schematic diagram of a server structure according to an embodiment of this application. The server 800 can vary considerably due to different configurations or performance. It may include one or more Central Processing Units (CPUs) 801 and one or more memories 802. The memories 802 are used to store executable program code, and the processors 801 are configured to execute the executable program code to implement the identity recognition methods provided in the various method embodiments described above. Of course, the server may also have wired or wireless network interfaces, a keyboard, and input / output interfaces for input and output. The server may also include other components for implementing device functions, which will not be elaborated here.

[0167] In an exemplary embodiment, a computer-readable storage medium including instructions is also provided, such as a memory including instructions, which can be executed by a terminal's processor to perform the aforementioned identity verification method. Optionally, the computer-readable storage medium may be a ROM, random access memory (RAM), CD-ROM, magnetic tape, floppy disk, or optical data storage device, etc.

[0168] In an exemplary embodiment, a computer program product is also provided, including a computer program that, when executed by a processor, implements the aforementioned identity recognition method. In some embodiments, the computer program product involved in this application embodiment may be deployed and executed on a single computer device, or on multiple computer devices located in one location, or on multiple computer devices distributed across multiple locations and interconnected via a communication network. These multiple computer devices distributed across multiple locations and interconnected via a communication network may constitute a blockchain system.

[0169] Other embodiments of this disclosure will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this disclosure are indicated by the claims.

[0170] It should be understood that this disclosure is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this disclosure is limited only by the appended claims.

Claims

1. An identity recognition method, characterized in that, Applied to a server, the method includes: In response to an identity verification request carrying an image of identification documents, image features are acquired to characterize the quality features of the image of identification documents, the quality features including at least one of brightness, sharpness, and border integrity; and historically available identity verification platforms to be tested are determined from multiple identity verification platforms based on previous identity verification processes. The image features and the platform identifier of the identity verification platform to be tested are input into the decision tree model, so that the decision tree model calls the prediction model of each identity verification platform to be tested based on the platform identifier. The prediction model of any identity verification platform is used to predict the difference between the identity recognition result output by the identity verification platform and the actual recognition result; each prediction model corresponds to a type of document image, and a type of document image has the same quality features. The image features are input into the prediction models of each of the identity verification platforms under test. Through multiple prediction models, the predicted values ​​for each identity verification platform under test are output. The predicted values ​​represent the predicted difference between the identity recognition result output by the corresponding identity verification platform for the image features and the actual recognition result of the image features. The predicted values ​​are positively correlated with the predicted differences. Based on the predicted values, the identity verification platforms under test are sorted. The order of each identity verification platform is negatively correlated with the predicted value corresponding to the identity verification platform. A test authentication platform is selected as the target platform. The availability of the target platform is confirmed. If available, the target platform is used for identity verification. If unavailable, the next test authentication platform is selected, and the availability confirmation process is repeated until a usable test authentication platform is selected. Platforms ranked higher in the test authentication platform list are selected with priority. The availability of the target platform is determined based on information returned by the target platform. If the server identifier of the current server is not in the target platform's platform list, the target platform is determined to be unavailable for the server. Then, multiple unavailable platform lists are queried based on the server identifier. Each unavailable platform list stores server identifiers corresponding to a specific reason for unavailability, and the returned information includes the reason for unavailability. A prediction model for an identity verification platform is generated as follows: The platform acquires historical output sample identity recognition results and corresponding sample document images; the platform determines the actual identification result and image features of the sample document images, resulting in a set of sample data pairs including the image features, the identity recognition result, and the actual identification result; a regression model is established based on the result differences and image features corresponding to multiple sets of sample data pairs, yielding regression parameters and error parameters representing the degree of influence of independent variables on dependent variables. The result differences are obtained based on the identity recognition result and the actual identification result in the sample data pairs; the regression model determined based on the regression parameters and the error parameters is used as the prediction model.

2. The identity recognition method according to claim 1, characterized in that, The method further includes: If a message indicating that the top-ranked authentication platform is unavailable is received, it is determined that the top-ranked authentication platform is unavailable. The top-ranked authentication platform is used to return a message when it is determined that the platform identifier is not in the platform list based on the platform identifier.

3. The identity recognition method according to claim 1, characterized in that, The method further includes: If the identity verification platform returns an identity verification result indicating successful identity verification and the document image is obtained based on a copy of the document, services will be refused to be provided to the object that triggered the identity verification request.

4. An identity recognition device, characterized in that, The device includes: The first acquisition unit is configured to, in response to an identity recognition request carrying an image of identification documents, acquire image features characterizing the quality features of the identification document image, the quality features including at least one of brightness, sharpness, and border integrity; and The module is configured to perform the following steps: determine the historically available authentication platform to be tested from multiple authentication platforms based on the previous identity recognition process; The calling unit is configured to input the image features and the platform identifier of the identity verification platform under test into the decision tree model, so that the decision tree model can call the prediction model of the identity verification platform under test based on the platform identifier. The prediction model of any identity verification platform is used to predict the difference between the identity recognition result output by the identity verification platform and the actual recognition result. Each prediction model corresponds to a type of document image, and a type of document image has the same quality features. The first determining unit is configured to input the image features into the prediction models of each of the identity verification platforms to be tested, and output prediction values ​​corresponding to each of the identity verification platforms to be tested through multiple prediction models. The prediction values ​​are used to represent the predicted difference between the identity recognition result of the corresponding identity verification platform for the image features and the actual recognition result of the image features. The prediction values ​​are positively correlated with the prediction difference. The identification unit is configured to sort the identity verification platforms under test based on the predicted values, wherein the order of each identity verification platform is negatively correlated with the predicted value corresponding to the identity verification platform. The module is used to perform the following steps: Select an authentication platform to be tested as the target platform; confirm whether the target platform is available; if available, use the target platform as the target authentication platform for identity verification; if unavailable, continue to select the next authentication platform to be tested as the target platform, and repeat the confirmation step until an available authentication platform to be tested is selected; wherein, the earlier the authentication platform to be tested is ranked, the higher its priority is selected; whether the target platform is available is determined based on the information fed back by the target platform; the target platform is used to determine that the target platform is unavailable to the server when it is determined that the server identifier of the current server is not in the platform list of the target platform; then, it continues to query multiple unavailable platform lists based on the server identifier, each unavailable platform list stores the server identifier corresponding to a reason for unavailability, and the feedback information includes the reason for unavailability; A prediction model for an identity verification platform is generated as follows: The platform acquires historical output sample identity recognition results and corresponding sample document images; the platform determines the actual identification result and image features of the sample document images, resulting in a set of sample data pairs including the image features, the identity recognition result, and the actual identification result; a regression model is established based on the result differences and image features corresponding to multiple sets of sample data pairs, yielding regression parameters and error parameters representing the degree of influence of independent variables on dependent variables. The result differences are obtained based on the identity recognition result and the actual identification result in the sample data pairs; the regression model determined based on the regression parameters and the error parameters is used as the prediction model.

5. An electronic device, characterized in that, include: processor; Memory used to store the processor's executable instructions; The processor is configured to execute the instructions to implement the identity recognition method as described in any one of claims 1 to 3.

6. A computer-readable storage medium, wherein instructions in the computer-readable storage medium, when executed by a processor of an electronic device, enable the electronic device to perform the identification method as claimed in any one of claims 1 to 3.

Images