A loan business image detection method, device, equipment and storage medium

Abstract

The present application discloses a loan business image detection method, device, equipment and storage medium, which relates to the field of image processing technology and is applied to the approval system, including: detecting a group photo image containing the facial information of the applicant of the loan business and the facial information of the corresponding account manager, and obtaining a probability result representing the probability of tampering of the group photo image; determining whether to trigger a face recognition operation for the group photo image based on the size relationship between the probability result and a preset probability threshold; if the face recognition operation is triggered, extracting the facial feature information in the group photo image, and calculating the similarity between the facial feature information and the facial feature information of the preset ID photo corresponding to the loan business, and obtaining a similarity result; judging whether the group photo image is legal or not according to the similarity result. In this way, the detection of the group photo image of the applicant and the account manager can be realized in the loan business, and the legality of the group photo image can be more accurately judged by the probability of tampering and the similarity.

Description

Technical Field

[0001] The present invention relates to the field of image processing technology, and in particular to a loan business image detection method, device, equipment and storage medium. Background Art

[0002] When applying for personal loans at financial institutions, customers are required to provide structured information such as their mobile phone number and ID number, as well as unstructured data such as facial recognition and document image processing. Images are a typical example of unstructured data, and these typically include liveness detection, facial comparison, dual recording, ID photos, and photos of the relationship manager and the customer. Market-wide image recognition technologies primarily include liveness detection (verifying whether the person is a real person), facial comparison (verifying whether the person is the same person), dual recording (audio and video recording), and ID photo recognition (which can utilize ORC (Optical Character Recognition)). During the loan approval process, some financial institutions require relationship managers to conduct an on-site visit to the customer's workplace or home and take a photo with the customer to demonstrate due diligence and confirm the customer's identity. However, in practice, some relationship managers, for various personal reasons, fail to conduct an on-site visit and instead fabricate photos with the customer to fulfill due diligence requirements, thereby ignoring the fraud risk of the customer's business or residential address.

[0003] It can be seen that how to effectively detect photos of account managers and customers during the loan approval process is a problem that needs to be solved in this field. Summary of the Invention

[0004] In view of this, the present invention aims to provide a loan business image detection method, device, equipment, and storage medium. This method can detect images of applicants and account managers in loan business applications, and more accurately determine the legitimacy of these images by using the probability of tampering and similarity. The specific solution is as follows:

[0005] In a first aspect, the present application provides a loan business image detection method, which is applied to an approval system, comprising:

[0006] Detecting a photo image containing facial information of a loan applicant and facial information of a corresponding account manager to obtain a probability result representing a probability that the photo image has been tampered with;

[0007] determining whether to trigger a face recognition operation for the group photo image based on a magnitude relationship between the probability result and a preset probability threshold;

[0008] If the face recognition operation is triggered, facial feature information of the person in the group photo is extracted through the face recognition operation, and similarity between the facial feature information and facial feature information of the pre-set ID photo corresponding to the loan business is calculated to obtain a similarity result;

[0009] Whether the group photo is legal is determined based on the similarity result.

[0010] Optionally, before detecting the photo image containing the facial information of the loan applicant and the facial information of the corresponding account manager, the process further includes:

[0011] Performing a consistency check on the personal information of the loan applicant and the liveness identification information of the applicant through the client to obtain a corresponding verification result;

[0012] If the verification result indicates that the information verification is passed, a photo image containing the facial information of the applicant and the facial information of the customer manager corresponding to the loan business is obtained through the client, so that the client can upload the photo image to the approval system.

[0013] Optionally, detecting a group photo image including facial information of a loan applicant and facial information of a corresponding account manager to obtain a probability result representing a probability of tampering of the group photo image includes:

[0014] Using a preset model, the differences between pixels in a photo containing the facial information of a loan applicant and the facial information of a corresponding account manager are detected to obtain abnormal features of the photo;

[0015] The proportion of the abnormal features in the group photo is determined as the probability that the group photo has been tampered with, and a probability result is obtained.

[0016] Optionally, the determining whether to trigger a face recognition operation for the group photo image based on a magnitude relationship between the probability result and a preset probability threshold includes:

[0017] If the probability result is less than the preset probability threshold, triggering a face recognition operation for the group photo image is allowed.

[0018] Optionally, the determining whether to trigger a face recognition operation for the group photo image based on a magnitude relationship between the probability result and a preset probability threshold includes:

[0019] If the probability result is not less than the preset probability threshold, the face recognition operation for the photo image is not allowed to be triggered, and the loan business is directly determined to be illegal.

[0020] Optionally, extracting facial feature information from the group photo through the facial recognition operation, and calculating the similarity between the facial feature information and facial feature information of a preset ID photo corresponding to the loan business to obtain a similarity result includes:

[0021] Determining a first facial region of the applicant and a second facial region of the account manager in the group photo, and performing a feature information extraction operation on the first facial region and the second facial region to obtain corresponding first facial feature information and second facial feature information;

[0022] The first facial feature information and the second facial feature information are respectively compared with the facial feature information of the applicant's ID photo and the facial feature information of the customer manager's ID photo that are pre-set corresponding to the loan business to obtain corresponding first similarity results and second similarity results.

[0023] Optionally, judging whether the group photo image is legal based on the similarity result includes:

[0024] If both the first similarity result and the second similarity result are greater than a preset similarity threshold, the group photo is determined to be legal;

[0025] If the first similarity result and / or the second similarity result is not greater than the preset similarity threshold, the group photo image is determined to be illegal.

[0026] In a second aspect, the present application provides a loan business image detection device, which is applied to an approval system and includes:

[0027] a tampering detection module configured to detect a photo containing the facial information of a loan applicant and the facial information of a corresponding account manager, and obtain a probability result representing a probability that the photo has been tampered with;

[0028] an operation triggering module, configured to trigger a face recognition operation on the group photo image based on a magnitude relationship between the probability result and a preset probability threshold;

[0029] a similarity calculation module, configured to extract facial feature information from the group photo through the facial recognition operation, and calculate the similarity between the facial feature information and the facial feature information of the pre-set ID photo corresponding to the loan business, to obtain a similarity result;

[0030] The judging module is used to judge whether the group photo image is legal according to the similarity result.

[0031] In a third aspect, the present application provides an electronic device, comprising:

[0032] Memory, used to store computer programs;

[0033] A processor is used to execute the computer program to implement the above-mentioned loan business image detection method.

[0034] In a fourth aspect, the present application provides a computer-readable storage medium for storing a computer program, which, when executed by a processor, implements the above-mentioned loan business image detection method.

[0035] As can be seen, the present application first detects the photo image containing the facial information of the loan applicant and the facial information of the corresponding account manager, and obtains a probability result representing the probability of tampering of the photo image; then triggers a face recognition operation on the photo image based on the relationship between the probability result and a preset probability threshold; then extracts facial feature information from the photo image through the face recognition operation, and calculates the similarity between the facial feature information and the facial feature information of the pre-set ID photo corresponding to the loan business, obtaining a similarity result; then determines whether the photo image is legitimate based on the similarity result. In this way, the present application can implement the detection operation of the photo image of the loan applicant and the corresponding account manager during the loan business approval process, first detecting the probability of tampering of the photo image, then identifying the similarity between the facial feature information in the photo image and the pre-set ID photo, and determining whether the photo image is legitimate based on the similarity; thus, in the loan business, the photo image of the applicant and the account manager can be detected, and the legitimacy of the photo image can be more accurately determined based on the tampering probability and similarity. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0037] Figure 1 This is a flow chart of a loan business image detection method disclosed in this application;

[0038] Figure 2 A flowchart of a specific loan business image detection method disclosed in this application;

[0039] Figure 3 This is a schematic diagram of the structure of a loan business image detection device disclosed in this application;

[0040] Figure 4 This is a structural diagram of an electronic device disclosed in this application. DETAILED DESCRIPTION

[0041] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0042] See also Figure 1 As shown, an embodiment of the present invention discloses a loan business image detection method, which is applied to an approval system, including:

[0043] Step S11: Detecting a group photo containing facial information of a loan applicant and facial information of a corresponding account manager to obtain a probability result representing a probability of tampering with the group photo.

[0044] In this embodiment of the present application, the approval system may receive loan application-related information uploaded by a client, including detailed application information and a related photo image. The photo image includes the facial information of the loan applicant and the facial information of the account manager associated with the loan application. The approval system first performs a tampering probability check on the received photo image, generating a probability result indicating the probability that the photo has been tampered with.

[0045] In a specific embodiment, before detecting the photo image containing the facial information of the loan applicant and the corresponding account manager, the process may further include: performing a consistency check on the personal information of the loan applicant and the applicant's liveness recognition information via a client to obtain a corresponding verification result; if the verification result indicates that the verification is successful, then obtaining a photo image containing the facial information of the applicant and the corresponding account manager via the client, so that the client can upload the photo image to the approval system. Specifically, before uploading loan-related information to the approval system, the client needs to perform a consistency check on the applicant's personal information and the applicant's liveness recognition information to obtain a verification result indicating whether the corresponding information is consistent; further, when processing the loan, the client can use the liveness recognition information to verify in real time whether the applicant and their corresponding personal information are consistent, thereby performing a consistency check on the information. If the corresponding verification result indicates that the information is consistent, then the client can further collect a photo image of the loan applicant and the corresponding account manager, where the photo image contains the facial information of both the applicant and the account manager. Correspondingly, if the verification result indicates that the information is inconsistent, the loan application for the applicant can be directly terminated.

[0046] In another specific embodiment, the detection of the group photo image containing the facial information of the applicant for the loan business and the facial information of the corresponding account manager to obtain a probability result representing the probability of the group photo image being tampered with may include: detecting the differences between the pixels of the group photo image containing the facial information of the applicant for the loan business and the facial information of the corresponding account manager through a preset model to obtain abnormal features of the group photo image; determining the proportion of the abnormal features in the group photo image as the probability that the group photo image has been tampered with to obtain a probability result. Specifically, when the approval system detects the group photo image containing the facial information of the applicant for the loan business and the facial information of the corresponding account manager uploaded by the client, it may perform feature detection on the group photo image through a preset network model to obtain abnormal features in the group photo image; then, based on the proportion of the abnormal features detected in the group photo image, it may calculate the probability that the group photo image has been tampered with to obtain a corresponding probability result.

[0047] Step S12: determining whether to trigger a face recognition operation for the group photo image based on a magnitude relationship between the probability result and a preset probability threshold.

[0048] In this embodiment, the above steps can detect the group photo image and obtain a corresponding probability result indicating the probability of tampering. Subsequently, based on the relationship between the probability result and a preset probability threshold, a determination is made as to whether to trigger a facial recognition operation for the group photo image. In a specific embodiment, determining whether to trigger a facial recognition operation for the group photo image based on the relationship between the probability result and the preset probability threshold may include: if the probability result is less than the preset probability threshold, allowing the facial recognition operation for the group photo image to be triggered. Specifically, if the probability result is less than the preset probability threshold, then the number of pixel features indicating an abnormality in the group photo image is relatively small, and it can be determined that the abnormality is caused by transmission. In this case, facial feature recognition can continue for the group photo image, i.e., triggering a facial recognition operation for the group photo image. Accordingly, in another specific embodiment, determining whether to trigger a facial recognition operation for the group photo image based on the relationship between the probability result and the preset probability threshold may include: if the probability result is not less than the preset probability threshold, not allowing the facial recognition operation for the group photo image to be triggered, and directly determining that the loan business is illegal. It can be understood that if the probability result representing the probability of tampering corresponding to the group photo image is not less than the preset probability threshold, it can be determined that the group photo image has been tampered with, such as being processed by image editing software; at this time, the group photo image is unreliable, and the relevant loan business can be directly determined to be illegal, and the relevant process of the loan business can be terminated.

[0049] Step S13: If the face recognition operation is triggered, facial feature information in the group photo is extracted through the face recognition operation, and the similarity between the facial feature information and the facial feature information of the preset ID photo corresponding to the loan business is calculated to obtain a similarity result.

[0050] In this embodiment, after the above steps, when a facial recognition operation is triggered for the group photo, facial feature information from the group photo can be extracted through the facial recognition operation. Furthermore, the facial feature information in the group photo can be compared with the ID photo associated with the loan application. It is understood that the ID photo of the loan applicant and the ID photo of the relevant account manager can be pre-stored in the approval system or obtained online. By comparing the facial feature information, the similarity between the facial feature information in the group photo and the facial feature information in the pre-set ID photo can be determined, resulting in a similarity result.

[0051] In a specific embodiment, extracting facial feature information from the group photo through the facial recognition operation and calculating the similarity between the facial feature information and the facial feature information of the ID photo preset corresponding to the loan business to obtain a similarity result may include: determining a first facial region of the applicant and a second facial region of the account manager in the group photo, performing a feature information extraction operation on the first facial region and the second facial region to obtain corresponding first facial feature information and second facial feature information; and comparing the first facial feature information and the second facial feature information with the facial feature information of the ID photo of the applicant and the ID photo preset corresponding to the loan business, respectively, to obtain corresponding first similarity results and second similarity results. Specifically, in the process of performing facial feature recognition on the group photo, it is necessary to perform feature information extraction operations on the first facial region where the applicant's face is located and the second facial region where the account manager's face is located in the group photo, respectively, to obtain the first facial feature information corresponding to the applicant and the second facial feature information corresponding to the account manager. Afterwards, the extracted first facial feature information and the second facial feature information are compared with the facial feature information in the preset relevant ID photo to obtain a first similarity result for the first facial feature information and a second similarity result for the second facial feature information, respectively.

[0052] Step S14: judging whether the group photo is legal based on the similarity result.

[0053] In this embodiment, after obtaining the similarity result of the facial features between the group photo image of the loan business and the ID photo through the above steps, the legality of the group photo image can be determined based on the similarity result; in a specific embodiment, the legality of the group photo image can be determined based on the similarity result, which may include: if the first similarity result and the second similarity result are both greater than a preset similarity threshold, then the group photo image is determined to be legal; if the first similarity result and / or the second similarity result are not greater than the preset similarity threshold, then the group photo image is determined to be illegal. Specifically, if the first similarity result and the second similarity result between the group photo image of the loan business and the pre-set ID photo are both greater than the preset similarity threshold, then the group photo image can be determined to be legal, that is, the loan business can proceed with the subsequent process normally; correspondingly, if the first similarity result and / or the second similarity result are not greater than the preset similarity threshold, then it means that the group photo image is illegal, and the relevant loan business should also be terminated.

[0054] It can be seen from this that the present application can realize the detection operation of the group photo image of the loan business applicant and the corresponding account manager during the loan business approval process. First, the probability of tampering of the group photo image is detected, and then the similarity between the facial feature information in the group photo image and the pre-set ID photo is identified, and the legality of the group photo image is judged based on the similarity. In this way, in the loan business, the detection of the group photo image of the applicant and the account manager can be realized, and the legality of the group photo image can be more accurately judged by the probability of tampering and the similarity. Moreover, the technical solution of the present application can realize automatic detection of group photos and output of detection results, which greatly releases manpower and improves the scope and efficiency of image detection.

[0055] like Figure 2 As shown, the embodiment of the present application discloses a loan business image detection method, which is applied to the approval system and specifically includes:

[0056] In this embodiment, when submitting a loan application, a customer (applicant) submits relevant personal information. Liveness recognition is used on the client to verify the consistency of the customer's ID information in real time. If the liveness recognition verification finds any inconsistency, a prompt is provided to re-verify the information. If multiple inconsistencies are found, the process is terminated. Accordingly, after the liveness recognition information is verified to be consistent, the relevant account manager can collect loan-related information, as well as a photo of the account manager and the loan applicant, through the client, and upload this photo to the approval system.

[0057] Furthermore, the approval system can use a pre-configured authentication detection model to detect the possibility of forgery in group photos. Specifically, the authentication detection model first analyzes the photo's content, such as verifying whether the entire photo's Exif (Exchangeable Image File) information has been edited using image editing software such as Photoshop (an image processing software) or GIMP (an open-source image processing software). It also detects noise or spectral artifacts to capture fine-grained visual differences. It provides direct feedback on tampering signs such as erasure, reprinted text, copy-and-move, and splicing, thereby deriving a probability result that the photo has been tampered with. In some cases, the group photo may also require normalization, such as resizing and normalizing pixel values, to meet the input requirements of the authentication detection model. Furthermore, data augmentation operations such as rotation, flipping, and scaling can be performed to improve the model's generalization capabilities. If the probability exceeds a certain threshold, the image is deemed tampered; otherwise, the group photo can continue to be tested. It can be understood that the authentication detection model extracts abnormal features by analyzing the differences in pixels or image quality across different unit areas in the image. These abnormal features reflect the differences in pixels or image quality across different unit areas in the image, providing a basis for authentication results. The authentication detection model can be trained based on sample images and their corresponding authentication labels. During training, the model gradually optimizes its parameters by learning from a large number of sample images and their corresponding authentication labels, improving authentication accuracy and robustness. Ultimately, a fully trained authentication detection model can effectively authenticate images to be authenticated in real-world applications.

[0058] Furthermore, after determining to continue detecting the group photo through the above steps, the face region can be located in the image using a face detection algorithm, such as MTCNN (Multi-task Convolutional Neural Network), SSD (Single Shot MultiBox Detector), or YOLO (You Only Look Once). These algorithms can accurately detect the location and size of the face. Effective feature information is extracted from the located face region. These features can include the location information of key points such as the eyes, nose, and mouth, as well as the shape information of the facial contour. Common feature extraction methods include convolutional neural networks (CNN) and local binary patterns (LBP). In the technical solution of this application, since counterfeit detection technology (tampering probability) is already used to ensure the authenticity of the image, deep neural networks can be further utilized to extract more complex and abstract features. The facial features extracted from the group photo are then compared with facial templates (such as ID photos) in a pre-set database. Various matching algorithms, such as Euclidean distance and cosine similarity, can be used to calculate the similarity between the features. Based on the similarity, the input image is compared with the corresponding face in the database, thereby confirming the identity. Furthermore, based on the identity comparison results, the recognized identity information is output. If a match is successful, the corresponding identity label is output; otherwise, a prompt message such as "Unknown" or "Unrecognizable" is output. The output of the authentication detection model can then be fed back through manual confirmation to optimize the model.

[0059] It can be seen from this that the present application can realize the detection operation of the group photo image of the loan business applicant and the corresponding account manager during the loan business approval process. First, the probability of tampering of the group photo image is detected, and then the similarity between the facial feature information in the group photo image and the pre-set ID photo is identified, and the legality of the group photo image is judged based on the similarity. In this way, in the loan business, the detection of the group photo image of the applicant and the account manager can be realized, and the legality of the group photo image can be more accurately judged through the probability of tampering and the similarity. Moreover, the technical solution of the present application can realize automatic detection of group photos and output of detection results, which greatly releases manpower, improves the scope and efficiency of image detection, and can intercept and prevent impact-type fraud risks to avoid risk losses.

[0060] like Figure 3 As shown, the embodiment of the present application discloses a loan business image detection device, which is applied to an approval system, including:

[0061] a tampering detection module 11 for detecting a photo containing the facial information of a loan applicant and the facial information of a corresponding account manager, and obtaining a probability result representing a probability that the photo has been tampered with;

[0062] An operation triggering module 12 is configured to trigger a face recognition operation for the group photo image based on a magnitude relationship between the probability result and a preset probability threshold;

[0063] A similarity calculation module 13 is configured to extract facial feature information from the group photo through the facial recognition operation, and calculate the similarity between the facial feature information and the facial feature information of the pre-set ID photo corresponding to the loan business, thereby obtaining a similarity result;

[0064] The judgment module 14 is configured to judge whether the group photo is legitimate based on the similarity result.

[0065] It can be seen from this that the present application can realize the detection operation of the group photo image of the loan business applicant and the corresponding account manager during the loan business approval process. First, the probability of tampering of the group photo image is detected, and then the similarity between the facial feature information in the group photo image and the pre-set ID photo is identified, and the legality of the group photo image is judged based on the similarity. In this way, in the loan business, the detection of the group photo image of the applicant and the account manager can be realized, and the legality of the group photo image can be more accurately judged by the probability of tampering and the similarity.

[0066] In a specific embodiment, the device may further include:

[0067] An information verification module is used to perform information consistency verification on the personal information of a loan applicant and the liveness recognition information of the applicant through the client to obtain a corresponding verification result;

[0068] The group photo image acquisition module is used to obtain a group photo image containing the facial information of the applicant and the facial information of the customer manager corresponding to the loan business through the client when the verification result representation information verification is passed, so that the client can upload the group photo image to the approval system.

[0069] In a specific implementation, the tamper detection module 11 may include:

[0070] a pixel detection unit configured to detect differences between pixels in a photo containing the facial information of a loan applicant and the facial information of a corresponding account manager using a preset model, thereby obtaining abnormal features of the photo;

[0071] The probability determination unit is used to determine the proportion of the abnormal features in the group photo as the probability that the group photo has been tampered with, and obtain a probability result.

[0072] In a specific embodiment, the operation triggering module 12 may include:

[0073] The first operation triggering unit is configured to allow triggering of a face recognition operation on the group photo image when the probability result is less than the preset probability threshold.

[0074] In another specific embodiment, the operation triggering module 12 may include:

[0075] The second operation triggering unit is configured to disallow triggering of the face recognition operation for the group photo image when the probability result is not less than a preset probability threshold, and directly determine that the loan business is illegal.

[0076] In a specific embodiment, the similarity calculation module 13 may include:

[0077] a feature extraction unit, configured to determine a first facial region of the applicant and a second facial region of the account manager in the group photo, and perform feature information extraction operations on the first facial region and the second facial region to obtain corresponding first facial feature information and second facial feature information;

[0078] The feature comparison unit is used to compare the first facial feature information and the second facial feature information with the facial feature information of the applicant's ID photo and the facial feature information of the customer manager's ID photo that are pre-set corresponding to the loan business, so as to obtain a corresponding first similarity result and a second similarity result.

[0079] In a specific embodiment, the judgment module 14 may include:

[0080] a first determination unit, configured to determine that the group photo is legitimate when both the first similarity result and the second similarity result are greater than a preset similarity threshold;

[0081] The second determination unit is configured to determine that the group photo image is illegal when the first similarity result and / or the second similarity result is not greater than the preset similarity threshold.

[0082] Furthermore, the embodiment of the present application also discloses an electronic device, Figure 4 This is a structural diagram of an electronic device 20 according to an exemplary embodiment. The content in the diagram should not be considered as any limitation to the scope of application of the present application.

[0083] Figure 4This is a schematic diagram of the structure of an electronic device 20 provided in an embodiment of the present application. The electronic device 20 may specifically include: at least one processor 21, at least one memory 22, a power supply 23, a communication interface 24, an input / output interface 25, and a communication bus 26. The memory 22 is used to store a computer program, which is loaded and executed by the processor 21 to implement the relevant steps of the loan business image detection method disclosed in any of the aforementioned embodiments. Furthermore, the electronic device 20 in this embodiment may specifically be an electronic computer.

[0084] In this embodiment, the power supply 23 is used to provide operating voltage for each hardware device on the electronic device 20; the communication interface 24 can create a data transmission channel between the electronic device 20 and the external device. The communication protocol it follows is any communication protocol that can be applied to the technical solution of this application and is not specifically limited here; the input and output interface 25 is used to obtain external input data or output data to the outside world. Its specific interface type can be selected according to specific application needs and is not specifically limited here.

[0085] In addition, the memory 22, as a carrier for resource storage, can be a read-only memory, random access memory, disk or CD, etc. The resources stored thereon can include an operating system 221, a computer program 222, etc., and the storage method can be temporary storage or permanent storage.

[0086] The operating system 221 is used to manage and control the hardware devices and computer program 222 on the electronic device 20. The operating system 221 may be Windows Server, NetWare, Unix, Linux, etc. In addition to including computer programs capable of implementing the loan business image detection method performed by the electronic device 20 as disclosed in any of the aforementioned embodiments, the computer program 222 may further include computer programs capable of performing other specific tasks.

[0087] Furthermore, this application discloses a computer-readable storage medium for storing a computer program; wherein, when executed by a processor, the computer program implements the aforementioned loan business image detection method. The specific steps of this method can be found in the corresponding content disclosed in the aforementioned embodiments and will not be further described here.

[0088] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from the other embodiments. Reference can be made to the descriptions of the identical or similar parts between the various embodiments. For the devices disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the descriptions are relatively simple, and the relevant parts can be referred to the descriptions of the methods.

[0089] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the above description has generally described the components and steps of each example according to their functions. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0090] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein may be implemented directly using hardware, a software module executed by a processor, or a combination of the two. The software module may be placed in random access memory (RAM), internal memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0091] Finally, it should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of additional identical elements in the process, method, article, or device comprising the element.

[0092] The above is a detailed introduction to the technical solution provided by the present application. Specific examples are used herein to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea. At the same time, for those skilled in the art, according to the ideas of the present application, there may be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A loan business image detection method, characterized in that: Applicable to approval systems, including: Detecting a group photo image containing facial information of a loan business applicant and facial information of a corresponding account manager to obtain a probability result representing a probability of tampering with the group photo image; Determining whether to trigger a face recognition operation for the group photo image based on a magnitude relationship between the probability result and a preset probability threshold; If the face recognition operation is triggered, facial feature information in the group photo is extracted through the face recognition operation, and the similarity between the facial feature information and the facial feature information of the pre-set ID photo corresponding to the loan business is calculated to obtain a similarity result; Whether the group photo image is legal is determined based on the similarity result.

2. The loan business image detection method according to claim 1, characterized in that: Before detecting the photo image containing the facial information of the applicant for the loan business and the facial information of the corresponding account manager, it also includes: Performing information consistency verification on the personal information of the applicant for the loan business and the liveness recognition information of the applicant through the client to obtain a corresponding verification result; If the verification result indicates that the information verification is passed, a group photo image including the facial information of the applicant and the facial information of the account manager corresponding to the loan business is obtained through the client, so that the client can upload the group photo image to the approval system.

3. The loan business image detection method according to claim 1, characterized in that: The detecting of the group photo image including the facial information of the loan business applicant and the facial information of the corresponding account manager to obtain a probability result representing the probability of the group photo image being tampered with includes: The difference between each pixel of the group photo image containing the facial information of the loan business applicant and the facial information of the corresponding account manager is detected by a preset model to obtain abnormal features of the group photo image; The proportion of the abnormal features in the group photo is determined as the probability that the group photo has been tampered with, and a probability result is obtained.

4. The loan business image detection method according to claim 1, characterized in that: The determining whether to trigger a face recognition operation for the group photo image based on a magnitude relationship between the probability result and a preset probability threshold includes: If the probability result is less than the preset probability threshold, triggering a face recognition operation for the group photo image is allowed.

5. The loan business image detection method according to claim 1, characterized in that: The determining whether to trigger a face recognition operation for the group photo image based on a magnitude relationship between the probability result and a preset probability threshold includes: If the probability result is not less than the preset probability threshold, triggering the face recognition operation for the group photo image is not allowed, and the loan business is directly determined to be illegal.

6. The loan business image detection method according to any one of claims 1 to 4, characterized in that: The facial feature information in the group photo is extracted through the facial recognition operation, and the similarity between the facial feature information and the facial feature information of the preset ID photo corresponding to the loan business is calculated to obtain a similarity result, including: Determine a first facial region of the applicant and a second facial region of the account manager in the group photo, and perform feature information extraction operations on the first facial region and the second facial region to obtain corresponding first facial feature information and second facial feature information; The first facial feature information and the second facial feature information are respectively compared with the facial feature information of the applicant's ID photo and the facial feature information of the account manager's ID photo that are pre-set corresponding to the loan business to obtain corresponding first similarity results and second similarity results.

7. The loan business image detection method according to claim 6, characterized in that: The determining whether the group photo image is legal according to the similarity result includes: If both the first similarity result and the second similarity result are greater than a preset similarity threshold, then the group photo is determined to be legal; If the first similarity result and / or the second similarity result is not greater than the preset similarity threshold, the group photo image is determined to be illegal.

8. A loan business image detection device, characterized in that: Applicable to approval systems, including: A tampering detection module is used to detect the group photo image containing the facial information of the applicant of the loan business and the facial information of the corresponding account manager, and obtain a probability result representing the probability of the group photo image being tampered with; An operation triggering module, configured to trigger a face recognition operation for the group photo image based on a magnitude relationship between the probability result and a preset probability threshold; A similarity calculation module, used to extract facial feature information from the group photo through the facial recognition operation, and calculate the similarity between the facial feature information and the facial feature information of the pre-set ID photo corresponding to the loan business to obtain a similarity result; The judging module is used to judge whether the group photo image is legal according to the similarity result.

9. An electronic device, characterized in that: include: Memory, used to store computer programs; A processor, used to execute the computer program to implement the loan business image detection method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: Used to store a computer program, which, when executed by a processor, implements the loan business image detection method as described in any one of claims 1 to 7.

Images