Credential detection method, credential detection device, medium, equipment and product

Abstract

This specification provides a document detection method, document detection device, computer-readable storage medium, electronic device, and computer program product. The method includes: classifying document images to obtain one or more categories of document images, and setting a key element corresponding to each category. Further, based on the key element corresponding to each category, information extraction processing is performed on target document images belonging to that category to obtain key information of the target document images. Each target document image is then detected based on its key information.

Description

Technical Field

[0001] This specification relates to the field of image processing technology, and in particular to a method for detecting vouchers, a device for detecting vouchers, a computer-readable storage medium, an electronic device, and a computer program product. Background Technology

[0002] Identification information in digital images (such as stamps representing organizations and individuals) may be altered or used to create false information. To improve data security, there is an urgent need for a technical solution to detect the authenticity of identification information in images.

[0003] It should be noted that the information disclosed in the background section above is only used to enhance the understanding of the background of this specification, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0004] The purpose of this specification is to provide a method, device, computer-readable storage medium, electronic device, and computer program product for detecting credentials, which can effectively detect the authenticity of identification information in digital images and improve data security to at least a certain extent.

[0005] Other features and advantages of this specification will become apparent from the following detailed description, or may be learned in part by practice of this specification.

[0006] According to one aspect of this specification, a voucher detection method is provided, the method comprising: classifying voucher images to obtain N categories of voucher images, where N is a positive integer; determining a key element corresponding to the i-th category, where i takes the value of each integer between 1 and N, including 1 and N; performing information extraction processing on a first target voucher image belonging to the i-th category based on the key element corresponding to the i-th category to obtain key information of the first target voucher image; and detecting the first target voucher image based on the key information of the first target voucher image.

[0007] According to another aspect of this specification, a voucher detection device is provided, the device comprising: a classification module, a determination module, an extraction module, and a detection module.

[0008] The classification module is used to classify the voucher image to obtain N categories of voucher images, where N is a positive integer; the determination module is used to determine the key element corresponding to the i-th category, where i takes the value of each integer between 1 and N, including 1 and N; the extraction module is used to perform information extraction processing on the first target voucher image belonging to the i-th category based on the key element corresponding to the i-th category to obtain the key information of the first target voucher image; and the detection module is used to detect the first target voucher image based on the key information of the first target voucher image.

[0009] According to another aspect of this specification, an electronic device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the credential detection method as described in the above embodiments.

[0010] According to one aspect of this specification, a computer-readable storage medium is provided that stores instructions which, when executed on a computer or processor, cause the computer or processor to perform the credential verification method as described in the above embodiments.

[0011] According to another aspect of this specification, a computer program product containing instructions is provided that, when the computer program product is run on a computer or processor, causes the computer or processor to perform the credential verification method as described in the above embodiments.

[0012] The credential detection method, credential detection device, computer-readable storage medium, electronic device, and computer program product provided in the embodiments of this specification have the following technical effects:

[0013] In the exemplary embodiments provided in this specification, voucher images are classified to obtain one or more categories of voucher images. Each category has corresponding key elements. Further, based on the key elements corresponding to each category, information extraction processing is performed on the target voucher images belonging to that category, thereby obtaining the key information of the target voucher images. This scheme can selectively extract key information for each category of vouchers, thereby improving information extraction efficiency and accuracy. Furthermore, detecting each target voucher image based on its highly accurate key information results in high detection accuracy, which in turn helps improve the efficiency and accuracy of voucher review.

[0014] 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 specification. Attached Figure Description

[0015] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this specification and, together with the description, serve to explain the principles of this specification. It is obvious that the drawings described below are merely some embodiments of this specification, and those skilled in the art can obtain other drawings based on these drawings without any inventive effort.

[0016] Figure 1 This is a schematic flowchart of a credential detection method provided in one embodiment of this specification.

[0017] Figure 2 This is a flowchart illustrating a credential detection method provided in another embodiment of this specification.

[0018] Figure 3 This is a flowchart illustrating a credential detection method provided in another embodiment of this specification.

[0019] Figure 4 This is a schematic diagram of a credential image with a target object provided as an embodiment of this specification.

[0020] Figure 5 This is a flowchart illustrating a credential detection method provided in yet another embodiment of this specification.

[0021] Figure 6 This is a schematic flowchart of a credential detection method provided in one embodiment of this specification.

[0022] Figure 7 This is a flowchart illustrating a credential detection method provided in another embodiment of this specification.

[0023] Figure 8 This is a schematic flowchart of a document duplication detection method provided in one embodiment of this specification.

[0024] Figure 9 A schematic diagram of a credential verification device to which one embodiment of this specification can be applied is shown.

[0025] Figure 10 A schematic diagram of a credential verification device that can be applied to another embodiment of this specification is shown.

[0026] Figure 11 This is a schematic diagram of the structure of the electronic device provided in the embodiments of this specification. Detailed Implementation

[0027] To make the objectives, technical solutions, and advantages of this specification clearer, the embodiments of this specification will be described in further detail below with reference to the accompanying drawings.

[0028] In the following description, when referring to the accompanying drawings, the same numbers in different drawings denote the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this specification. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this specification as detailed in the appended claims.

[0029] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided to make this specification more comprehensive and complete, and to fully convey the concept of example embodiments to those skilled in the art. The described features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. In the following description, numerous specific details are provided to give a full understanding of the embodiments described herein. However, those skilled in the art will recognize that the technical solutions described herein may be practiced with one or more of the specific details omitted, or other methods, components, apparatus, steps, etc., may be employed. In other instances, well-known technical solutions are not shown or described in detail to avoid obscuring various aspects of this specification.

[0030] Furthermore, the accompanying drawings are merely illustrative diagrams of this specification and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and therefore repeated descriptions of them will be omitted. Some block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities. These functional entities may be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor devices and / or microcontroller devices.

[0031] Some internet platforms or systems require users to submit supporting documents, such as identity verification and proof of factual circumstances. For example, in consumer finance scenarios, if a user is unable to repay on time due to objective factors, the consumer finance platform, during the repayment negotiation process, requires the user to submit valid proof of their inability to repay on time, such as proof of poverty. After receiving the supporting documents or vouchers provided by the user, the system needs to verify them for review. However, the following problems exist in the verification process provided by relevant technologies:

[0032] 1. For user-submitted vouchers (the supporting documents in this embodiment can be collectively referred to as vouchers), manual review of each item is inefficient, especially for large platforms and systems with high daily user submission volumes requiring significant manpower. 2. Verifying the authenticity of user-submitted vouchers is challenging. For example, valid vouchers often require official seals from relevant organizations, a process rife with forgery (e.g., using image processing software like Photoshop). Currently, manual verification is often inaccurate and difficult. 3. Many user-submitted supporting documents are also based on templates. Users simply modify their personal information before submission. This generally constitutes forged supporting documentation. Quickly and effectively identifying whether such documents have been modified or copied is a problem that needs to be addressed.

[0033] To address the aforementioned problems in related technologies, this specification provides a credential detection method, a credential detection device, a computer-readable storage medium, an electronic device, and a computer program product. Specifically, this specification first introduces embodiments of the credential detection method.

[0034] in, Figure 1 This is a flowchart illustrating a credential verification method provided in one embodiment of this specification. Exemplarily, the execution entity of the evaluation and verification method provided in this embodiment is a computing device, such as a server. (See reference...) Figure 1 The embodiment shown in the figure includes: S110-S140.

[0035] In S110, the voucher images are classified to obtain N categories of voucher images, where N is a positive integer.

[0036] The credential detection scheme provided in the embodiments of this specification can be executed on an electronic image of a credential, i.e., the aforementioned credential image. The credential can be evidentiary material used for identity verification, factual verification, etc.

[0037] For example, refer to Figure 2 The credential image 200 is input into a pre-trained image-text classification model 210, which then classifies the credential image 200. The credential image 200 can be one or more credential images belonging to the same user, such as identity document images and / or poverty certificate images belonging to the same user; it can also be multiple credential images belonging to different users. The categories corresponding to the image-text classification model correspond to the type of credential, such as identity document category, poverty certificate category, occupation certificate category, etc.

[0038] Suppose that the classification result of the above voucher image 200 is N (values ​​are positive integers) categories. (Reference) Figure 2 The classification result 220 output by the image-text classification model 210 includes: voucher images belonging to the first category are voucher image a, voucher image b, and voucher image c; voucher images belonging to the second category are voucher image h, voucher image j, ...; voucher images belonging to the Nth category are voucher image s, etc. Among them, voucher images a, b, c, h, j, ... s all belong to the aforementioned voucher images 200. If the aforementioned voucher image 200 is a single image, after processing by the image-text classification model 210, the classification result obtained might be that voucher image 200 belongs to the poverty certificate category.

[0039] Continue to refer to Figure 1 In S120, the key element corresponding to the i-th category is determined, where i takes the value of each integer between 1 and N, including 1 and N.

[0040] In exemplary embodiments, the key elements corresponding to different types of credentials are not entirely the same. For example, the key elements to be identified in a poverty certificate include identity information and the reason for poverty, while the key elements to be identified in an occupation certificate include identity information and occupation type. Therefore, for different types of credentials, users can set the key elements to be identified or retrieved according to their actual needs. For example, the key elements corresponding to the identity certificate type include: name, date of birth, address, etc. The key elements corresponding to the poverty certificate type include: name, telephone / contact information, poverty type, etc.

[0041] refer to Figure 2 In this embodiment, a key element / key element set 230 is determined for each category, including: key elements corresponding to the first category, ..., elements corresponding to the ith category, ..., key elements corresponding to the Nth category. Therefore, for a voucher belonging to a certain category, the information contained in the voucher is extracted according to the set key element class, and denoted as: the key information of the voucher.

[0042] This specification first categorizes the voucher images to intelligently determine the category of the voucher image to be detected. Furthermore, it identifies the key elements corresponding to that category, thereby specifically determining the information to be extracted from that category of voucher, which helps ensure the efficiency and accuracy of information extraction.

[0043] In S130, information extraction processing is performed on the first target voucher image belonging to the i-th category based on the key elements corresponding to the i-th category to obtain the key information of the first target voucher image. And in S140, the first target voucher image is detected based on the key information of the first target voucher image.

[0044] In an exemplary embodiment, the aforementioned first target voucher image (to distinguish it from the subsequent "second target voucher image") is any voucher image belonging to the i-th category. This embodiment uses the processing of the first target voucher image to illustrate the processing of any voucher image belonging to the i-th category.

[0045] refer to Figure 2 Based on the key elements corresponding to the first category, key information is obtained for target voucher images belonging to the first category. Synchronously or asynchronously, based on the key elements corresponding to the Nth category, key information is obtained for target voucher images belonging to the Nth category. Furthermore, by inputting the key information of each voucher image into the detection module 250, detection of each voucher image can be achieved.

[0046] Figure 1 In the illustrated embodiment of the voucher detection method, voucher images are classified to obtain one or more categories. Each category has corresponding key elements. Further, based on the key elements corresponding to each category, information extraction processing is performed on the target voucher images belonging to that category, thereby obtaining the key information of the target voucher image. This scheme allows for targeted extraction of key information for each category of vouchers, thus improving information extraction efficiency and accuracy. Furthermore, detecting each target voucher image based on its highly accurate key information results in high detection accuracy, which in turn helps improve the efficiency and accuracy of voucher verification.

[0047] In an exemplary embodiment, Figure 3 This is a schematic flowchart illustrating a credential verification method provided in another embodiment of this specification. (See reference) Figure 3 The embodiment shown in the figure includes: S310-S340 and S110.

[0048] In S310, it is determined whether the target object exists in the voucher image.

[0049] For example, the target object mentioned above could be a stamp, fingerprint, etc. (See reference) Figure 4The image of the document shown shows that the stamp 410 overlaps with the text information "Party A (Official Seal): Dongying XXX Co., Ltd.", while the fingerprint 420 overlaps with the text information "Legal (Authorized) Representative: YYY". It is evident that these target objects are typically located in key positions on the document (where key information is located). To improve the accuracy of subsequent key information identification and extraction, preprocessing of these target objects is necessary to reduce interference with key information identification and extraction.

[0050] In this embodiment, before executing S110 (classifying the voucher image), S310 is executed first to determine whether a target object exists in the voucher image. If it is determined that no target object exists in the voucher image, then S110 is executed directly to proceed as follows: Figure 1 The illustrated embodiment implements credential detection. If it is determined that a target object exists in the credential image, then steps S320-S340 are executed.

[0051] In S320, the authenticity of the target object is determined.

[0052] For example, the voucher image to be detected can be input into a pre-trained object detection model. This object detection model can be a deep learning model (e.g., Mantra-net) or a tree model based on target object features (e.g., grayscale distribution features and image edge features of the target object). In this embodiment, the output of the object detection model determines whether a target object exists in the voucher image, and if so, the authenticity of the target object.

[0053] If the target object is fake, then S340 is executed: determine the detection result of the voucher image and generate a reminder message to resubmit the voucher image. If the object detection model determines that the target object in the voucher image is a forged object, it means that the current voucher does not meet the requirements. Therefore, a reminder message to resubmit the voucher image can be generated to remind the relevant user to resubmit. It can be seen that by utilizing the target object before classifying the voucher image, a quick detection of some vouchers (where the target object exists and its authenticity is fake) can be achieved. At the same time, it can also reduce the need for... Figure 1 The embodiment shown reduces the number of credential images required for credential detection, thereby reducing the computational load to some extent.

[0054] If the target object is real, then S330 is executed: the target object is erased from the voucher image. For example, the erasure of the target object can be implemented using image processing software. Further, for the voucher image after erasing the target object, S110 is executed, that is, the voucher image is classified to obtain N categories of voucher images. In this embodiment, by performing the above-mentioned eraser preprocessing on the target object, interference with the identification and extraction of key information can be reduced, which is beneficial to improving the accuracy of subsequent key information identification and extraction.

[0055] visible, Figure 3 In the provided embodiments, target objects are used to quickly detect some vouchers (those containing target objects and whose authenticity is false) before classifying the voucher images. This also reduces the need for... Figure 1 The illustrated embodiment reduces the number of voucher images required for voucher detection, thus decreasing computational load to some extent. Furthermore, by performing preprocessing on the target objects, interference with key information identification and extraction is reduced, improving the accuracy of subsequent key information identification and extraction.

[0056] In an exemplary embodiment, Figure 5 This is a flowchart illustrating a credential verification method provided in yet another embodiment of this specification. (See reference) Figure 5 The embodiment shown in the figure includes: S110 and S510-S550.

[0057] After executing S110, multiple categories of credential images are identified. Among these categories, only one or a few categories may require the presence of the aforementioned target object. For example, identity verification credentials do not require the presence of the target object, while poverty verification credentials do, for instance, each poverty verification document requires fingerprints as per requirements. In this embodiment, the type requiring the presence of the target object is denoted as the target type.

[0058] If one or more voucher images to be detected are first classified in S110, and one of the results is that there is no voucher image of the target type among the currently classified voucher images, then there is no need to execute the corresponding embodiments of S510-S550 on the current one or more voucher images to be detected, thereby avoiding unnecessary calculation steps, saving computing resources, and improving voucher detection efficiency.

[0059] Specifically, in S510, for a second target credential image belonging to the target category, it is determined whether a target object exists.

[0060] If the target object is determined to exist in the voucher image, then S520-S550 are executed.

[0061] In S520, the authenticity of the target object is determined.

[0062] For example, the second target voucher image can be input into a pre-trained object detection model, and the output of the object detection model can be used to determine whether a target object exists in the second target voucher image, and if a target object exists in the second target voucher image, whether the target object is genuine or not.

[0063] If the target object is real, then S530 is executed: the target object is erased from the second target voucher image. For example, the erasure of the target object can be implemented using image processing software. Further, for the voucher image after erasing the target object, S540 is executed, that is, based on the key elements corresponding to the target category, information extraction processing is performed on the second target voucher image after erasing the target object to obtain the key information of the second target voucher image. S540 can be a specific implementation of S130. In this embodiment, by performing the above-mentioned erasure preprocessing on the target object, interference with key information identification and extraction can be reduced, which is beneficial to improving the accuracy of key information identification and extraction.

[0064] If the target object is not found in the voucher image, or if the target object is fake, then S550 is executed: determine the detection result of the second target voucher image and generate a reminder message to resubmit the voucher image. If the target object is not found in the second target voucher image, it means that the current voucher image does not meet the requirement of the target type requiring a stamp / fingerprint, and a reminder message to resubmit the voucher image is generated. If the target object in the voucher image is determined to be a forged object by the object detection model, it means that the current voucher does not meet the requirements, and therefore a reminder message to resubmit the voucher image also needs to be generated to remind the relevant user to resubmit.

[0065] visible, Figure 5 In the provided embodiments, after classifying the voucher images, it is determined whether a voucher image belonging to the aforementioned target type exists based on the classification. If no such image exists, steps S510-S550 do not need to be executed, thereby reducing computational load and improving detection efficiency. If a voucher image belonging to the target type exists, the preprocessing provided by steps S510-S550 for erasing the target object can reduce interference with key information identification and extraction, thus improving the accuracy of key information identification and extraction.

[0066] In an exemplary embodiment, Figure 6 This is a schematic flowchart illustrating a credential verification method provided in one embodiment of this specification. (Reference) Figure 6 The embodiment shown in the figure includes:

[0067] Before executing S110, execute S610: determine whether the analysis result of the text recognition accuracy of the voucher image meets the first preset value.

[0068] In the embodiments of this specification, optical character recognition (OCR) is required during the extraction of key information from the voucher image. In order to improve the recognition accuracy, the analysis result of the text recognition degree is determined before classifying the voucher image. This can be determined by a preset analysis model.

[0069] If the text recognition accuracy analysis result is not greater than the first preset value, it indicates that the text recognition effect of the voucher image is poor. Then, execute S620: generate a reminder message for re-examining the voucher image. The first preset value can be determined according to actual needs; for example, it can be 95%. A value greater than or meeting the first preset value indicates that the recognition accuracy of the voucher image is greater than 95%.

[0070] If the text recognition accuracy analysis result is greater than the first preset value, it indicates that the text recognition accuracy of the voucher image is high. Classification processing is then performed in S110 to determine the category to which the voucher image belongs. Further, S1302 is executed: text information recognition is performed on the target voucher image belonging to the i-th category to obtain the text information of the target voucher image; and S1304 is executed: based on the key elements corresponding to the i-th category, the text information of the target voucher image is extracted to obtain the key information of the target voucher image.

[0071] exist Figure 6 In the embodiment shown, before classifying the voucher images, the voucher images are first screened based on the analysis results of text recognition, and images with text that do not meet the requirements or are not clearly captured are removed, thereby improving the quality of the voucher images used for classification and facilitating subsequent information recognition and extraction.

[0072] In an exemplary embodiment, Figure 7 This is a flowchart illustrating a credential verification method according to another embodiment of this specification, which can be specifically used as a concrete implementation of S140. (See reference...) Figure 7 The credential detection types provided in the embodiment shown in the figure include: information integrity detection, information authenticity detection, and duplicate detection.

[0073] Specifically, refer to Figure 7The key information 710 of the target credential image belonging to the i-th category is compared with the key element 720 corresponding to the i-th category to perform information integrity detection on the target credential image belonging to the i-th category. For example, the i-th category is the identity verification type, and the key elements of the identity verification type include, for example, name, date of birth, and address. However, if the key information extracted from the identity verification image submitted by the user does not include the date of birth, it can be determined that the credential image information is incomplete.

[0074] refer to Figure 7 The key information 730 of the target credential image is compared with the authentication information 740 corresponding to the target credential image to verify the authenticity of the information in the target credential image. For example, if the user's name is extracted from the credential image provided by the user as YYY, and the user's name in the system's authentication information is YYy, then the authenticity of the credential image information can be determined to be fake.

[0075] refer to Figure 7 By comparing the key information (after removing identity information) 750 of the target credential image belonging to the i-th category with the i-th database of descriptive information about the i-th category, duplicate detection of the aforementioned target credential image can be achieved. For example, Figure 8 This is a flowchart illustrating a method for detecting duplicate vouchers according to an embodiment of this specification. (Refer to...) Figure 8 The embodiment shown in the figure includes:

[0076] S1402, filter out the identity information from the key information of the target credential image belonging to the i-th category; S1404, obtain the i-th database of descriptive information about the i-th category, wherein the i-th database contains descriptive information about the i-th category for multiple users respectively; S1406, compare the key information with the identity information filtered out with the descriptive information about the i-th category for each user in the i-th database.

[0077] The duplicate detection in this specification primarily targets cases of falsified content in real-name authentication certificates. This mainly manifests as users using the same template and simply modifying their personal information before submitting to the system. For example, in poverty certificates and serious illness certificates, the descriptions of the reasons for poverty and the reasons for serious illness are essentially the same. Submissions using the same template will yield largely similar content, thus requiring a duplicate check on the content submitted by the user.

[0078] To implement information duplication detection, a database is set up for each category requiring duplication detection. This database stores descriptive information about that type from different users, with the similarity between them being less than a preset value. For example, for the poverty certificate category, the database corresponding to this category stores poverty certificate descriptions from multiple users, and the descriptions from different users are not duplicated. That is, the poverty certificate descriptions from each user stored in this database are considered authentic and reliable.

[0079] Continue to refer to Figure 8 S1408, determine whether the similarity is greater than the second preset value.

[0080] For example, characters extracted from the voucher image can be compared with characters in the description information of different users in the corresponding database. The similarity is determined based on the degree of repetition or intersection ratio between the characters. This method is relatively accurate and concise and efficient. Alternatively, the description information of different users in the database can be stored as sample semantic vectors. Furthermore, the similarity is calculated between the semantic vector of key information extracted from the voucher image and the aforementioned sample semantic vectors. The second preset value is determined according to actual needs.

[0081] If the similarity between the key information of the identity information and the description information of any user in the i-th category in the i-th database is greater than the second preset value, it means that the description information of the current user in the i-th category is duplicated with the description information of at least one user in the database. In other words, there is a high possibility of content forgery. Then, execute S14010: generate a reminder message to resubmit the credential image.

[0082] If the similarity between the key information of the identity information and the description information of any user in the i-th category in the i-th database is not greater than the second preset value, it means that the description information of the current user in the i-th category is not repeated with the description information of any user in the database. In other words, the possibility of content forgery is small. Then, S14012 is executed: the key information of the target credential image is added to the i-th database to enrich the database of this type and facilitate the subsequent information duplication detection.

[0083] The credential detection scheme provided by the above embodiments in this specification can detect credentials belonging to any category in multiple categories (multiple scenarios). The detection result is either to accept the credential image or not to accept the credential image. If the credential image is not accepted, the user can be reminded to re-upload the corresponding credential image.

[0084] The exemplary embodiments provided in this specification offer an intelligent identification method for real-name credentials. This method can automatically perform verification of the integrity, accuracy, authenticity, and duplication of credential information submitted by users, thereby improving the accuracy of credential detection and review, increasing the efficiency of credential review, and reducing the false judgment rate.

[0085] It should be noted that the above figures are merely illustrative of the processes included in the methods according to exemplary embodiments of this specification, and are not intended to be limiting. It is readily understood that the processes shown in the above figures do not indicate or limit the temporal order of these processes. Furthermore, it is readily understood that these processes may, for example, be executed synchronously or asynchronously in multiple modules.

[0086] The following are embodiments of the apparatus described in this specification, which can be used to execute the embodiments of the methods described in this specification. For details not disclosed in the apparatus embodiments of this specification, please refer to the embodiments of the methods described in this specification.

[0087] in, Figure 9 A schematic diagram of a credential verification device applicable to one embodiment of this specification is shown. Please refer to... Figure 9 The credential verification device shown in the figure can be implemented as all or part of an electronic device through software, hardware, or a combination of both. It can also be integrated as an independent module on a server or as an independent module in an electronic device.

[0088] The voucher detection device 900 described in the embodiments of this specification includes: a classification module 910, a determination module 920, an extraction module 930, and a detection module 940.

[0089] The classification module 910 is used to classify the voucher image to obtain N categories of voucher images, where N is a positive integer; the determination module 920 is used to determine the key element corresponding to the i-th category, where i is any integer between 1 and N, including 1 and N; the extraction module 930 is used to perform information extraction processing on the first target voucher image belonging to the i-th category based on the key element corresponding to the i-th category to obtain the key information of the first target voucher image; and the detection module 940 is used to detect the first target voucher image based on the key information of the first target voucher image.

[0090] In an exemplary embodiment, Figure 10 A schematic diagram illustrating the structure of a credential verification device according to another exemplary embodiment of this specification is provided. See also... Figure 10 :

[0091] In an exemplary embodiment, based on the foregoing scheme, the aforementioned credential detection device 900 further includes a judgment module 950.

[0092] The aforementioned judgment module 950 is used to determine whether a target object exists for a second target voucher image belonging to the target category after classifying the voucher image, wherein the target category is the category in which the voucher image must contain a target object; the aforementioned judgment module 950 is also used to determine the authenticity of the target object if the target object exists in the second target voucher image.

[0093] Specifically, if the authenticity of the target object is true, the second target voucher image is used to extract key information based on the key elements corresponding to the target category; if the authenticity of the target object is false or the target object does not exist in the second target voucher image, the detection result of the second target voucher image is determined and a reminder message to resubmit the voucher image is generated.

[0094] In an exemplary embodiment, based on the foregoing scheme, the aforementioned credential detection device 900 further includes an erasure module 960.

[0095] The erasure module 960 is used to erase the target object from the second target voucher image when the authenticity of the target object is true; the extraction module 930 is specifically used to: perform information extraction processing on the second target voucher image with the erased target object based on the key elements corresponding to the target category, and obtain the key information of the second target voucher image.

[0096] In an exemplary embodiment, based on the foregoing scheme, the judgment module 950 is further configured to: determine whether a target object exists in the voucher image before classifying the voucher image; the judgment module 950 is further configured to: perform authenticity detection on the target object for the voucher image containing the target object.

[0097] Where the target object is not present in the aforementioned voucher image or the authenticity of the target object is true, the aforementioned voucher image is used for classification; where the authenticity of the target object is false, the detection result of the aforementioned voucher image is determined.

[0098] In an exemplary embodiment, based on the foregoing solution, the erasure module 960 is further configured to: erase the target object from the voucher image if the authenticity of the target object is true.

[0099] The classification module 910 is specifically used to classify the voucher images with the target object erased and the voucher images without the target object, to obtain N categories of voucher images.

[0100] In an exemplary embodiment, based on the aforementioned scheme, the determining module 920 is specifically used to: determine the key element corresponding to the i-th category based on the attribute characteristics of the credential information of the i-th category.

[0101] In an exemplary embodiment, based on the foregoing scheme, the aforementioned credential detection device 900 further includes an analysis module 970.

[0102] The analysis module 970 is used to perform text recognition analysis on the voucher image before classifying the voucher image.

[0103] For voucher images whose text recognition results do not meet the first preset value, a reminder message to resubmit the voucher image is generated; for voucher images whose text recognition results meet the first preset value, they are classified. The extraction module 930 is specifically used for: performing text information recognition on the first target voucher image belonging to the i-th category to obtain the text information of the first target voucher image; and extracting the text information of the first target voucher image based on the key elements corresponding to the i-th category to obtain the key information of the first target voucher image.

[0104] In an exemplary embodiment, based on the aforementioned scheme, the detection module 940 is specifically used to: compare the key information of the first target credential image with the key elements corresponding to the i-th category to detect the information integrity of the first target credential image; and / or compare the key information of the first target credential image with the authentication information corresponding to the first target credential image to detect the authenticity of the information of the first target credential image.

[0105] In an exemplary embodiment, based on the aforementioned scheme, the detection module 940 includes: an acquisition unit 9402, a screening unit 9404, a comparison unit 9406, and a detection unit 9408.

[0106] The acquisition unit 9402 is used to: filter out identity information from the key information of the first target credential image; the filtering unit 9404 is used to: acquire the i-th database of description information about the i-th category, wherein the i-th database contains description information about the i-th category for multiple users respectively; the comparison unit 9406 is used to: compare the key information from which the identity information has been filtered out with the description information about the i-th category for each user in the i-th database; and the detection unit 9408 is used to: detect the information duplication of the first target credential image based on the comparison results.

[0107] In an exemplary embodiment, based on the foregoing scheme, the detection module 940 further includes a generation unit 94010 and an addition unit 94012.

[0108] The generation unit 94010 is used to: after comparing the key information for filtering out the identity information with the description information of each user in the i-th category in the i-th database, if the comparison result is that the similarity between the key information for filtering out the identity information and the description information of any user in the i-th database regarding the i-th category is greater than a second preset value, then generate a reminder message to resubmit the credential image.

[0109] The aforementioned adding unit 94012 is used to add the key information of the first target credential image to the i-th database when the comparison result is that the similarity between the key information of the aforementioned identity information and the description information of any user in the i-th database regarding the i-th category is not greater than the aforementioned second preset value.

[0110] It should be noted that the above embodiment of the voucher detection device is only illustrated by the division of the above functional modules when executing the voucher detection method. In actual application, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above.

[0111] Furthermore, the voucher detection device and voucher detection method embodiments provided in the above embodiments belong to the same concept. Therefore, for details not disclosed in the device embodiments of this specification, please refer to the voucher detection method embodiments described above, which will not be repeated here.

[0112] The example numbers in this specification are for descriptive purposes only and do not represent the superiority or inferiority of the examples.

[0113] This specification also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of any of the methods described above.

[0114] Figure 11 This schematic diagram illustrates the structure of an electronic device according to an exemplary embodiment of this specification. Please refer to... Figure 11 As shown, the electronic device 1100 includes a processor 1101 and a memory 1102.

[0115] In this embodiment, processor 1101 is the control center of the computer system and can be a processor of a physical machine or a processor of a virtual machine. Processor 1101 may include one or more processing cores, such as a 4-core processor or an 8-core processor. Processor 1101 may be implemented using at least one hardware form selected from Digital Signal Processing (DSP), Field-Programmable Gate Array (FPGA), and Programmable Logic Array (PLA). Processor 1101 may also include a main processor and a coprocessor; the main processor is used to process data in the wake-up state, and the coprocessor is a low-power processor used to process data in the standby state.

[0116] In the embodiments described in this specification, the processor 1101 is specifically used for:

[0117] The voucher images are classified to obtain N categories of voucher images, where N is a positive integer; the key element corresponding to the i-th category is determined, where i is any integer between 1 and N, including 1 and N; based on the key element corresponding to the i-th category, information extraction processing is performed on the first target voucher image belonging to the i-th category to obtain the key information of the first target voucher image; based on the key information of the first target voucher image, the first target voucher image is detected.

[0118] Furthermore, the aforementioned processor 1101 is specifically used for:

[0119] After classifying the voucher images as described above, for the second target voucher image belonging to the target category, it is determined whether a target object exists, wherein the target category is the category in which the voucher image must contain a target object; if the target object exists in the second target voucher image, the authenticity of the target object is determined.

[0120] Specifically, if the authenticity of the target object is true, the second target voucher image is used to extract key information based on the key elements corresponding to the target category; if the authenticity of the target object is false or the target object does not exist in the second target voucher image, the detection result of the second target voucher image is determined and a reminder message to resubmit the voucher image is generated.

[0121] Furthermore, the aforementioned processor 1101 is specifically used for:

[0122] If the authenticity of the target object is true, the target object is erased from the second target voucher image. The information extraction processing of the first target voucher image belonging to the i-th category based on the key elements corresponding to the i-th category to obtain the key information of the first target voucher image includes: the information extraction processing of the second target voucher image after erasing the target object based on the key elements corresponding to the target category to obtain the key information of the second target voucher image.

[0123] Furthermore, the aforementioned processor 1101 is specifically used for:

[0124] Before classifying the voucher images as described above, it is determined whether a target object exists in the voucher images; for voucher images containing the target object, the authenticity of the target object is detected; wherein, if the target object does not exist in the voucher image or if the authenticity of the target object is true, the voucher image is used for classification; if the authenticity of the target object is false, the detection result of the voucher image is determined.

[0125] Furthermore, determining the key element corresponding to the i-th category includes: determining the key element corresponding to the i-th category based on the attribute characteristics of the voucher information of the i-th category.

[0126] Furthermore, the aforementioned processor 1101 is specifically used for:

[0127] Before classifying the voucher images as described above, text recognition analysis is performed on the voucher images. For voucher images whose text recognition analysis results do not meet the first preset value, a reminder message to resubmit the voucher image is generated. For voucher images whose text recognition analysis results meet the first preset value, classification is performed. The information extraction processing of the first target voucher image belonging to the i-th category based on the key elements corresponding to the i-th category to obtain the key information of the first target voucher image includes: performing text information recognition on the first target voucher image belonging to the i-th category to obtain the text information of the first target voucher image; and extracting the text information of the first target voucher image based on the key elements corresponding to the i-th category to obtain the key information of the first target voucher image.

[0128] Furthermore, the detection of the first target credential image based on the key information of the first target credential image includes: comparing the key information of the first target credential image with the key elements corresponding to the i-th category to detect the information integrity of the first target credential image; and / or comparing the key information of the first target credential image with the authentication information corresponding to the first target credential image to detect the authenticity of the information of the first target credential image.

[0129] Furthermore, the detection of the first target credential image based on the key information of the first target credential image includes: filtering out identity information from the key information of the first target credential image; obtaining an i-th database of descriptive information about the i-th category, wherein the i-th database contains descriptive information about the i-th category for each user; comparing the key information with the identity information filtered out with the descriptive information about the i-th category for each user in the i-th database; and detecting the information duplication of the first target credential image based on the comparison result.

[0130] Furthermore, the aforementioned processor 1101 is specifically used for:

[0131] After comparing the key information for filtering out the aforementioned identity information with the description information of each user in the i-th category in the i-th database, if the comparison result is that the similarity between the key information for filtering out the aforementioned identity information and the description information of any user in the i-th category in the i-th database is greater than a second preset value, then a reminder message to resubmit the credential image is generated; if the comparison result is that the similarity between the key information for filtering out the aforementioned identity information and the description information of any user in the i-th database for the i-th category is not greater than the second preset value, then the key information of the first target credential image is added to the i-th database.

[0132] Memory 1102 may include one or more computer-readable storage media, which may be non-transitory. Memory 1102 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 of this specification, the non-transitory computer-readable storage media in memory 1102 is used to store at least one instruction for execution by processor 1101 to implement the methods in the embodiments of this specification.

[0133] In some embodiments, the electronic device 1100 further includes a peripheral device interface 1103 and at least one peripheral device. The processor 1101, memory 1102, and peripheral device interface 1103 can be connected via a bus or signal line. Each peripheral device can be connected to the peripheral device interface 1103 via a bus, signal line, or circuit board. Specifically, the peripheral device includes at least one of a display screen 1104, a camera 1105, and an audio circuit 1106.

[0134] Peripheral interface 1103 can be used to connect at least one input / output (I / O) related peripheral device to processor 1101 and memory 1102. In some embodiments of this specification, processor 1101, memory 1102, and peripheral interface 1103 are integrated on the same chip or circuit board; in some other embodiments of this specification, any one or two of processor 1101, memory 1102, and peripheral interface 1103 can be implemented on separate chips or circuit boards. This specification does not specifically limit the embodiments in this regard.

[0135] Display screen 1104 is used to display a user interface (UI). The UI may include graphics, text, icons, videos, and any combination thereof. When display screen 1104 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 1101 for processing. In this case, display screen 1104 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 of this specification, there may be one display screen 1104, which serves as the front panel of electronic device 1100; in other embodiments, there may be at least two display screens 1104, respectively disposed on different surfaces of electronic device 1100 or in a folded design; in still other embodiments, display screen 1104 may be a flexible display screen, disposed on a curved or folded surface of electronic device 1100. Furthermore, display screen 1104 may also be configured as a non-rectangular irregular shape, i.e., a non-rectangular screen. The display screen 1104 can be made of materials such as liquid crystal display (LCD) and organic light-emitting diode (OLED).

[0136] Camera 1105 is used to capture images or videos. Optionally, camera 1105 includes a front-facing camera and a rear-facing camera. Typically, the front-facing camera is located on the front panel of the electronic device, and the rear-facing camera is located on the back of the electronic device. 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, virtual reality (VR) shooting, or other fusion shooting functions. In some embodiments of this specification, camera 1105 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.

[0137] The audio circuit 1106 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 that are input to the processor 1101 for processing. For stereo sound acquisition or noise reduction purposes, there may be multiple microphones, each located in a different part of the electronic device 1100. The microphone may also be an array microphone or an omnidirectional microphone.

[0138] Power supply 1107 is used to supply power to various components in electronic device 1100. Power supply 1107 can be alternating current, direct current, a disposable battery, or a rechargeable battery. When power supply 1107 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 that is charged via a wired line, and a wireless rechargeable battery is a battery that is charged via a wireless coil. The rechargeable battery can also be used to support fast charging technology.

[0139] The block diagrams of the electronic device shown in the embodiments of this specification do not constitute a limitation on the electronic device 1100. The electronic device 1100 may include more or fewer components than shown, or combine certain components, or use different component arrangements.

[0140] In the description of this specification, it should be understood that the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance. Those skilled in the art can understand the specific meaning of these terms in this specification based on the specific circumstances. Furthermore, in the description of this specification, unless otherwise stated, "multiple" means two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship.

[0141] This specification also provides a computer-readable storage medium storing instructions that, when executed on a computer or processor, cause the computer or processor to perform one or more steps in the above embodiments. If the constituent modules of the above-described credential verification device are implemented as software functional units and sold or used as independent products, they can be stored in the aforementioned computer-readable storage medium.

[0142] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions. When these computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this specification are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in or transmitted through a computer-readable storage medium. The computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, Digital Subscriber Line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium accessible to a computer or a data storage device such as a server or data center that integrates one or more available media. The aforementioned available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., Digital Versatile Discs (DVDs)), or semiconductor media (e.g., Solid State Disks (SSDs)).

[0143] It should be noted that the above description describes specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims may be performed in a different order than that shown in the embodiments and still achieve the desired results. Furthermore, the processes depicted in the drawings do not necessarily require the specific or sequential order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0144] The above description is merely a specific embodiment of this specification, but the scope of protection of this specification is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this specification should be included within the scope of protection of this specification. Therefore, equivalent variations made in accordance with the claims of this specification are still within the scope of this specification.

Claims

1. A method of detecting a credential, wherein, The method includes: The voucher images are classified to obtain N categories of voucher images, where N is a positive integer. The steps of classifying the voucher images include: inputting the voucher images into a pre-trained image-text classification model to classify the voucher images through the image-text classification model; Determine the key element corresponding to the i-th category, where i takes the value of any integer between 1 and N, including 1 and N; Based on the key elements corresponding to the i-th category, information extraction processing is performed on the first target voucher image belonging to the i-th category to obtain the key information of the first target voucher image; Based on the key information of the first target voucher image, the first target voucher image is detected; The detection of the first target voucher image based on key information of the first target voucher image includes: The key information of the first target voucher image is compared with the key elements corresponding to the i-th category to detect the information integrity of the first target voucher image; The key information of the first target credential image is compared with the authentication information corresponding to the first target credential image to detect the authenticity of the information in the first target credential image; and The identity information in the key information of the first target credential image is filtered out; the i-th database of descriptive information about the i-th category is obtained, wherein the i-th database contains descriptive information about the i-th category for multiple users respectively; the key information from which the identity information is filtered out is compared with the descriptive information about the i-th category for each user in the i-th database; and the information duplication of the first target credential image is detected based on the comparison results.

2. The method of claim 1, wherein, After classifying the voucher images, the method further includes: For a second target voucher image belonging to the target category, determine whether a target object exists, wherein the target category is the category in which the voucher image must contain a target object; If the target object exists in the second target voucher image, determine the authenticity of the target object; Specifically, if the authenticity of the target object is true, the second target voucher image is used to extract key information based on the key elements corresponding to the target category; if the authenticity of the target object is false or the target object does not exist in the second target voucher image, the detection result of the second target voucher image is determined and a reminder message to resubmit the voucher image is generated.

3. The method of claim 2, wherein, The method further includes: If the authenticity of the target object is true, the target object is erased from the second target voucher image; The information extraction process is performed on the first target voucher image belonging to the i-th category based on the key elements corresponding to the i-th category to obtain the key information of the first target voucher image, including: Based on the key elements corresponding to the target category, information extraction processing is performed on the second target voucher image of the erased target object to obtain the key information of the second target voucher image.

4. The method of claim 1, wherein, Prior to classifying the voucher images, the method further includes: Determine whether the target object exists in the voucher image; For the credential image containing the target object, perform authenticity detection on the target object; Wherein, if the target object is not present in the voucher image or if the authenticity of the target object is true, the voucher image is used for classification; if the authenticity of the target object is false, the detection result of the voucher image is determined.

5. The method of claim 4, wherein, The method further includes: If the authenticity of the target object is true, the target object is erased from the voucher image; The process of classifying the voucher images to obtain N categories of voucher images includes: The credential images of the target object that have been erased and the credential images of the target object that do not exist are classified into N categories.

6. The method according to any one of claims 1 to 5, wherein, Prior to classifying the voucher images, the method further includes: Perform text recognition analysis on the voucher image; For voucher images whose text recognition results do not meet the first preset value, a reminder message to resubmit the voucher image is generated; for voucher images whose text recognition results meet the first preset value, they are classified. The information extraction process is performed on the first target voucher image belonging to the i-th category based on the key elements corresponding to the i-th category to obtain the key information of the first target voucher image, including: Text information recognition is performed on the first target voucher image belonging to the i-th category to obtain the text information of the first target voucher image; Based on the key elements corresponding to the i-th category, the text information of the first target voucher image is extracted and processed to obtain the key information of the first target voucher image.

7. The method according to claim 1, wherein, After comparing the key information for filtering out the identity information with the description information of each user in the i-th database regarding the i-th category, the method further includes: If the comparison result is that the similarity between the key information of the identity information to be filtered out and the description information of any user in the i-th database about the i-th category is greater than a second preset value, then a reminder message to resubmit the credential image is generated. If the comparison result is that the similarity between the key information of the identity information to be filtered out and the description information of any user in the i-th database regarding the i-th category is not greater than the second preset value, then the key information of the first target credential image is added to the i-th database.

8. A voucher detection device, wherein, The device includes: The classification module is used to classify voucher images to obtain N categories of voucher images, where N is a positive integer; The determination module is used to: determine the key element corresponding to the i-th category, where i takes the value of each integer between 1 and N, including 1 and N; The extraction module is used to: perform information extraction processing on the first target voucher image belonging to the i-th category based on the key elements corresponding to the i-th category, and obtain the key information of the first target voucher image; The detection module is used to: detect the first target voucher image based on key information of the first target voucher image; The classification module is specifically used to: input the voucher image into a pre-trained image-text classification model, so as to classify the voucher image through the image-text classification model; The detection module is specifically used to: compare the key information of the first target voucher image with the key elements corresponding to the i-th category in order to detect the information integrity of the first target voucher image; The key information of the first target credential image is compared with the authentication information corresponding to the first target credential image to detect the authenticity of the information in the first target credential image; and The identity information in the key information of the first target credential image is filtered out; the i-th database of descriptive information about the i-th category is obtained, wherein the i-th database contains descriptive information about the i-th category for multiple users respectively; the key information from which the identity information is filtered out is compared with the descriptive information about the i-th category for each user in the i-th database; and the information duplication of the first target credential image is detected based on the comparison results.

9. A computer-readable storage medium storing instructions that, when executed on a computer or processor, cause the computer or processor to perform the credential verification method as described in any one of claims 1 to 7.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein, When the processor executes the computer program, it implements the credential detection method as described in any one of claims 1 to 7.

11. A computer program product comprising instructions that, when run on a computer or processor, causes the computer or processor to perform the credential verification method as described in any one of claims 1 to 7.

Images