Service information processing method and device, storage medium and electronic equipment

Abstract

The invention discloses a service information processing method and device, a storage medium and electronic equipment. The method relates to the field of artificial intelligence, and comprises the following steps: obtaining a scanning image of a to-be-audited paper business file of a target business; performing noise reduction processing on the scanned image through an image noise reduction model to obtain a first image; marking the to-be-audited information in the first image through an information identification model to obtain a second image; and comparing the annotation content in the second image with the service message information of the target service to obtain a comparison result, and determining an audit result for the paper service file based on the comparison result. Through application of the method and the device, the problem of low auditing efficiency in auditing whether the paper business file of the specific business is consistent with the online business message information or not in the prior art is solved.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence, and more specifically, to a method, device, storage medium, and electronic device for processing business information. Background Art

[0002] In the current fintech landscape, especially within financial institutions, remote authorization and review have become an integral part of daily operations. With the acceleration of digital transformation, a large number of paper business documents need to be converted to electronic data and verified for consistency with online business messages to ensure accuracy and compliance. Currently, this process in related technologies often relies on manual review, requiring auditors to individually verify that the information on paper documents matches electronic business records, resulting in low review efficiency.

[0003] Currently, no effective solution has been proposed to the above-mentioned problems in related technologies. Summary of the Invention

[0004] The main purpose of this application is to provide a business information processing method, device, storage medium and electronic device to solve the problem of low audit efficiency when reviewing whether the paper business documents of a specific business are consistent with the online business message information in the relevant technology.

[0005] To achieve the above-mentioned objectives, according to one aspect of the present application, a method for processing business information is provided. The method comprises: obtaining a scanned image of a paper business document to be reviewed for a target business; performing noise reduction processing on the scanned image using an image noise reduction model to obtain a first image; annotating the information to be reviewed in the first image using an information recognition model to obtain a second image; comparing the annotated content in the second image with the business message information of the target business to obtain a comparison result, and determining a review result for the paper business document based on the comparison result.

[0006] Furthermore, the business information processing method also includes: obtaining a first training sample set, wherein the training samples in the first training sample set include a first sample image and a second sample image, and the second sample image is obtained by adding noise to the first sample image; alternately training the discriminator and the generator in the generative adversarial network through the first training sample set until the target iteration condition is met, and determining the trained generator as the image denoising model.

[0007] Furthermore, the method for processing business information also includes: fixing the model parameters of the generator, performing denoising processing on the second sample image through the generator to obtain a third sample image, and training the discriminator based on the third sample image and the first sample image, wherein the discriminator is trained to distinguish between the real first sample image and the third sample image output by the generator; when the training of the discriminator meets the first iteration condition, fixing the model parameters of the discriminator, and training the generator based on the discrimination results of the discriminator for the third sample image and the first sample image, wherein the generator is trained to generate a third sample image whose similarity with the first sample image is greater than a preset threshold; when the training of the generator meets the second iteration condition, continuing to fix the model parameters of the generator and train the discriminator.

[0008] Furthermore, the method for processing business information also includes: obtaining a second training sample set, wherein the training sample of the second training sample set is a fourth sample image, the fourth sample image is obtained after noise reduction processing of the scanned image of the sample paper business document, and the true label of the second training sample set is the fourth sample image marked with information to be reviewed; training the initial information recognition model through the second training sample set to obtain an information recognition model.

[0009] Furthermore, the method for processing business information also includes: obtaining a business template associated with the target business, wherein the business template is used to record the attributes that need to be extracted from the paper business file of the target business; performing text recognition on the annotation content in the second image to obtain multiple text fields; matching the multiple text fields with the attributes in the business template to obtain attribute values ​​corresponding to the attributes in the business template; comparing the attribute values ​​corresponding to the attributes in the business template with the business message information of the target business to obtain a comparison result.

[0010] Furthermore, the method for processing business information also includes: for each attribute in the business template, determining the target attribute value corresponding to the attribute from the business message information; when the attribute value of each attribute is the same as the target attribute value of the attribute, determining that the comparison result representation annotation content is exactly the same as the business message information; when there is an attribute whose attribute value is different from the target attribute value, determining that the comparison result representation annotation content is different from the business message information.

[0011] Furthermore, the method for processing business information also includes: when the comparison result indicates that the marked content is exactly the same as the business message information, determining that the audit result indicates that the paper business document has passed the audit; when the comparison result indicates that the marked content is different from the business message information, determining that the audit result indicates that the paper business document has failed the audit.

[0012] To achieve the above-mentioned objectives, according to another aspect of the present application, a business information processing device is provided. The device includes: a first acquisition module for acquiring a scanned image of a paper business document to be reviewed for a target business; a first processing module for performing noise reduction processing on the scanned image using an image noise reduction model to obtain a first image; a second processing module for annotating the information to be reviewed in the first image using an information recognition model to obtain a second image; and a first determination module for comparing the annotated content in the second image with the business message information of the target business to obtain a comparison result, and determining the review result for the paper business document based on the comparison result.

[0013] Furthermore, the business information processing device also includes: a second acquisition module, used to obtain a first training sample set, wherein the training samples in the first training sample set include a first sample image and a second sample image, and the second sample image is obtained by adding noise to the first sample image; a second determination module, used to alternately train the discriminator and the generator in the generative adversarial network through the first training sample set until the target iteration condition is met, and then determine the trained generator as the image denoising model.

[0014] Furthermore, the second determination module also includes: a first training submodule, which is used to fix the model parameters of the generator, perform denoising on the second sample image through the generator to obtain a third sample image, and train the discriminator based on the third sample image and the first sample image, wherein the discriminator is trained to distinguish between the real first sample image and the third sample image output by the generator; a second training submodule, which is used to fix the model parameters of the discriminator when the training of the discriminator meets the first iteration condition, and train the generator based on the discrimination results of the discriminator for the third sample image and the first sample image, wherein the generator is trained to generate a third sample image whose similarity with the first sample image is greater than a preset threshold; a third training submodule, which is used to continue to fix the model parameters of the generator and train the discriminator when the training of the generator meets the second iteration condition.

[0015] Furthermore, the business information processing device also includes: a third acquisition module, used to obtain a second training sample set, wherein the training sample of the second training sample set is a fourth sample image, the fourth sample image is obtained after noise reduction processing of the scanned image of the sample paper business document, and the true label of the second training sample set is the fourth sample image marked with information to be reviewed; a training module, used to train the initial information recognition model through the second training sample set to obtain an information recognition model.

[0016] Furthermore, the first determination module also includes: an acquisition sub-module, used to obtain a business template associated with the target business, wherein the business template is used to record the attributes that need to be extracted from the paper business documents of the target business; an identification sub-module, used to perform text recognition on the annotation content in the second image to obtain multiple text fields; a processing sub-module, used to match the multiple text fields with the attributes in the business template to obtain attribute values ​​corresponding to the attributes in the business template; a comparison sub-module, used to compare the attribute values ​​corresponding to the attributes in the business template with the business message information of the target business to obtain a comparison result.

[0017] Furthermore, the comparison submodule also includes: a first determination submodule, which is used to determine the target attribute value corresponding to the attribute from the business message information for each attribute in the business template; a second determination submodule, which is used to determine that the comparison result representation annotation content is exactly the same as the business message information when the attribute value of each attribute is the same as the target attribute value of the attribute; and a third determination submodule, which is used to determine that the comparison result representation annotation content is different from the business message information when there is an attribute whose attribute value is different from the target attribute value.

[0018] Furthermore, the first determination module also includes: a fourth determination sub-module, which is used to determine that the audit result indicates that the paper business document has passed the audit when the comparison result indicates that the marked content is exactly the same as the business message information; and a fifth determination sub-module, which is used to determine that the audit result indicates that the paper business document has failed the audit when the comparison result indicates that the marked content is different from the business message information.

[0019] In order to achieve the above-mentioned purpose, according to another aspect of the present application, a computer-readable storage medium is provided, which includes a stored executable program, wherein when the executable program is running, the device where the computer-readable storage medium is located is controlled to execute the above-mentioned business information processing method.

[0020] In order to achieve the above-mentioned purpose, according to another aspect of the present application, an electronic device is provided, which includes a memory storing an executable program; and a processor for running the program, wherein the above-mentioned business information processing method is executed when the program is running.

[0021] In order to achieve the above-mentioned purpose, according to another aspect of the present application, a computer program product is provided, including computer instructions, which implement the steps of the above-mentioned business information processing method when executed by a processor.

[0022] In an embodiment of the present application, by obtaining a scanned image of a paper business document to be audited for a target business and annotating the information to be audited in the image through an information recognition model, effective extraction of audit-related content in the paper business document is achieved, avoiding the impact of invalid information on the audit progress. By comparing the annotated content with the business message information of the target business to determine the audit result of the paper business document, automatic audit of the consistency between the paper business document and the business message information is achieved, thereby improving audit efficiency. In addition, using an image denoising model to perform noise reduction processing on the scanned image can effectively improve image quality, making key information easier to identify, thereby improving audit efficiency and audit accuracy.

[0023] It can be seen that the method provided by this application achieves the purpose of automatically reviewing whether paper business documents are consistent with online business message information, realizes the technical effect of improving review efficiency, and solves the technical problem of low review efficiency when reviewing whether paper business documents of specific business are consistent with online business message information in related technologies. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The accompanying drawings, which constitute part of this application, are intended to provide a further understanding of this application. The exemplary embodiments and descriptions of this application are intended to explain this application and do not constitute an improper limitation on this application. In the accompanying drawings:

[0025] Figure 1 This is a hardware structure block diagram of a computer terminal provided according to an embodiment of the present application;

[0026] Figure 2 This is the process of the method for processing business information provided by the embodiment of this application Figure 1 ;

[0027] Figure 3 is a structural diagram of an information recognition model provided according to an embodiment of the present application;

[0028] Figure 4 This is the process of the method for processing business information provided by the embodiment of this application Figure 2 ;

[0029] Figure 5 is a schematic diagram of a device for processing business information provided in an embodiment of the present application;

[0030] Figure 6 This is a structural block diagram of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0031] In order to enable those skilled in the art to better understand the present invention, the following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of this application.

[0032] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequential order. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in a sequence other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0033] It should be noted that the collected information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for display, data for analysis, etc.) involved in this application are information and data authorized by the user or fully authorized by all parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of the relevant data comply with relevant laws, regulations and standards, take necessary confidentiality measures, do not violate public order and good morals, and provide corresponding operation portals for users to choose to authorize or refuse. For example, an interface is set up between this system and relevant users or institutions to provide users with corresponding operation portals for users to choose to agree or refuse the automated decision-making results; if the user chooses to refuse, the expert decision-making process will be entered.

[0034] Example 1

[0035] According to an embodiment of the present application, an embodiment of a method for processing business information is also provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0036] The method embodiment provided in the first embodiment of the present application can be executed in a mobile terminal, a computer terminal or a similar computing device. Figure 1The hardware structure block diagram of a computer terminal (or mobile device) for implementing a method for processing business information is shown. Figure 1 As shown, the computer terminal 10 (or mobile device) may include one or more (illustrated as 102a, 102b, ..., 102n in the figure) processors 102 (the processor 102 may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA), a memory 104 for storing data, and a transmission device 106 for communication functions. In addition, it may also include: a display, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of the BUS bus), a network interface, a power supply and / or a camera. It will be understood by those skilled in the art that Figure 1 The structure shown is only for illustration and does not limit the structure of the above electronic device. Figure 1 More or fewer components than shown, or with Figure 1 Different configurations shown.

[0037] It should be noted that the one or more processors 102 and / or other data processing circuits described above may generally be referred to herein as "data processing circuitry". The data processing circuitry may be embodied in whole or in part as software, hardware, firmware, or any other combination thereof. In addition, the data processing circuitry may be a single independent processing module, or may be incorporated in whole or in part into any of the other components of the computer terminal 10 (or mobile device). As described in the embodiments of the present application, the data processing circuitry serves as a processor control (e.g., selection of a variable resistor terminal path connected to an interface).

[0038] The memory 104 can be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the business information processing method in the embodiment of the present application. The processor 102 executes various functional applications and data processing by running the software programs and modules stored in the memory 104, that is, realizes the above-mentioned business information processing method. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include a memory remotely located relative to the processor 102, and these remote memories may be connected to the computer terminal 10 via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0039] The transmission device 106 is configured to receive or transmit data via a network. A specific example of the aforementioned network may include a wireless network provided by the communications provider of the computer terminal 10. In one embodiment, the transmission device 106 includes a network interface controller (NIC), which can be connected to other network devices via a base station to enable communication with the Internet. In another embodiment, the transmission device 106 may be a radio frequency (RF) module, which is configured to communicate with the Internet wirelessly.

[0040] The display may be, for example, a touch screen liquid crystal display (LCD) that enables a user to interact with a user interface of the computer terminal 10 (or mobile device).

[0041] Under the above operating environment, this application provides Figure 2 The processing method of the business information shown. Figure 2 This is the process of the method for processing business information provided by the embodiment of this application Figure 1 .

[0042] Step S201: obtaining a scanned image of a paper business document to be reviewed of a target business.

[0043] Optionally, electronic devices, application systems, servers and other devices may be used as the execution subject of the present application. In this embodiment, the target processing system is used as the execution subject to execute the above-mentioned business information processing method.

[0044] Optionally, the target business can be a business within a financial institution, specifically a business requiring remote authorization and review. For example, in a financial scenario, the target business could be an account opening application, a large-value transfer, or a loan approval. The paper business documents to be reviewed are original physical documents related to the target business, such as a customer's completed application form, photocopies of ID documents, transaction receipts, contracts, invoices, etc.

[0045] Optionally, the staff may convert the above-mentioned paper business documents into digital images (ie, the aforementioned scanned images) through a scanning device, and then upload the scanned images to the target processing system.

[0046] In an optional embodiment, after acquiring the scanned image, the target processing system may perform a preliminary inspection on the scanned image, such as checking the resolution, size and integrity verification, and then execute subsequent steps (i.e., steps S202-S204) if the preliminary inspection passes.

[0047] Step S202: performing noise reduction processing on the scanned image using an image noise reduction model to obtain a first image.

[0048] Optionally, the image denoising model may be a neural network model.

[0049] In an optional embodiment, the image denoising model is obtained based on generative adversarial network training, and the image denoising model may be a generator in a trained generative adversarial network.

[0050] In an optional embodiment, the image denoising model may also be another type of neural network model, such as a convolutional neural network model.

[0051] Alternatively, the scanned image can be input into an image denoising model, and the image output by the image denoising model is then determined as the denoised image, i.e., the first image. By performing denoising on the input image, image quality can be improved, providing a clear image foundation for subsequent feature recognition and annotation.

[0052] Step S203: annotate the information to be reviewed in the first image using an information recognition model to obtain a second image.

[0053] Optionally, the information recognition model may be a neural network model, for example, it may be a convolutional neural network model.

[0054] Optionally, the target processing system may input the first image into an information recognition model, which may identify the information to be reviewed in the first image and annotate the information to be reviewed, thereby outputting a second image. For example, the information recognition model may annotate the information to be reviewed by selecting a box, and the information types to be reviewed include, but are not limited to, text, numbers, and symbols.

[0055] Step S204 : comparing the annotated content in the second image with the business message information of the target business to obtain a comparison result, and determining the audit result for the paper business document based on the comparison result.

[0056] Optionally, the business message information of the target business refers to information generated online by the target business, for example, it may be an electronic application form, certificate image, electronic transaction application, electronic contract, electronic invoice, etc. associated with the target business.

[0057] For example, when the target business is a loan business, when a user applies for a loan, the user fills out a loan application form offline and simultaneously initiates a loan application online, or the staff of the financial institution fills out the loan information online based on the content of the paper application form. The target processing system can determine the loan application form filled out by the user offline as a paper business document to be reviewed, and determine the online loan application or loan information as the business message information of the target business.

[0058] For example, in the case of a transfer business of the target business, the user fills out a transfer application form offline and initiates the transfer application online. The target processing system can determine the transfer application form filled out by the user offline as a paper business document to be reviewed, and determine the application content of the online transfer application as the business message information of the target business.

[0059] Since paper business documents are easily tampered with, and when submitting online and offline information, there may be differences between the two due to typos, changes in intention, operational errors, etc., the target processing system can compare the annotated content in the second image with the business message information of the target business to obtain a comparison result. The comparison result is used to indicate whether the annotated content is the same as the business message information. Based on the comparison result, the audit result for the paper business document is determined. The audit result is used to indicate whether the paper business document has passed the audit, that is, it is used to indicate whether the paper business document has passed the consistency audit.

[0060] In an embodiment of the present application, by obtaining a scanned image of a paper business document to be audited for a target business and annotating the information to be audited in the image through an information recognition model, effective extraction of audit-related content in the paper business document is achieved, avoiding the impact of invalid information on the audit progress. By comparing the annotated content with the business message information of the target business to determine the audit result of the paper business document, automatic audit of the consistency between the paper business document and the business message information is achieved, thereby improving audit efficiency. In addition, using an image denoising model to perform noise reduction processing on the scanned image can effectively improve image quality, making key information easier to identify, thereby improving audit efficiency and audit accuracy.

[0061] It can be seen that the method provided by this application achieves the purpose of automatically reviewing whether paper business documents are consistent with online business message information, realizes the technical effect of improving review efficiency, and solves the technical problem of low review efficiency when reviewing whether paper business documents of specific business are consistent with online business message information in related technologies.

[0062] Optionally, in the business information processing method provided in the embodiment of the present application, the image denoising model is obtained by: obtaining a first training sample set, wherein the training samples in the first training sample set include a first sample image and a second sample image, and the second sample image is obtained by adding noise to the first sample image; alternately training the discriminator and the generator in the generative adversarial network through the first training sample set until the target iteration condition is met, and determining the trained generator as the image denoising model.

[0063] In an optional embodiment, the image denoising model is trained using a generative adversarial network. During training, the results returned by the discriminator are incorporated into the generator's training, which then adjusts the weights of the displayed defects to improve the training effect. Since interference in scanned images is often caused by unclear areas, a self-attention mechanism is optionally introduced into the generator, allowing the network to focus on different areas of the image, thereby generating a more globally consistent image and better handling image noise.

[0064] Optionally, the target processing system can train a generative adversarial network based on a first training sample set. The training samples in the first training sample set include a first sample image and a second sample image, where the second sample image is obtained by denoising the first sample image. For example, the target processing system can obtain 5,000 images, 1,000 of which are high-definition virtual business scan images, and the remaining 4,000 are obtained by denoising the aforementioned high-definition virtual business document images. The denoising process can be to randomly add Gaussian noise and salt and pepper noise, and perform scaling and rotation to enhance the model's generalization and denoising capabilities for business images, enabling it to identify and remove noise and interference from a variety of images. The aforementioned 1,000 high-definition virtual business scan images are equivalent to the first sample images, and the remaining 4,000 are equivalent to the second sample images. The target processing system can combine a high-definition virtual business scan image and an image obtained by denoising it into a training sample.

[0065] After obtaining the first training sample set, the discriminator and generator in the generative adversarial network can be alternately trained using the first training sample set until the target iteration condition is met, and the trained generator is determined as the image denoising model. For example, the generator parameters are first fixed, and the second sample image is processed by the generator to obtain a de-noised image. These de-noised images are then used together with the first sample image to train the discriminator to improve its discrimination ability. Subsequently, after a certain number of iterations is met or a preset performance indicator (such as accuracy) is reached, the discriminator parameters are fixed and the generator is trained again. This alternating process is repeated until the target iteration condition is met (such as 10 rounds of alternating training, or the image quality of the de-noised image output by the generator meets a specific evaluation standard, such as a preset peak signal-to-noise ratio threshold), a trained generative adversarial network is obtained, and the trained generator is determined as the image denoising model.

[0066] It should be noted that the image denoising model obtained based on generative adversarial network training can effectively improve the denoising effect of the image denoising model, thereby improving the review efficiency and accuracy.

[0067] Optionally, in the business information processing method provided in the embodiment of the present application, the discriminator and the generator in the generative adversarial network are alternately trained using the first training sample set, including: fixing the model parameters of the generator, performing denoising processing on the second sample image through the generator to obtain a third sample image, and training the discriminator based on the third sample image and the first sample image, wherein the discriminator is trained to distinguish between the real first sample image and the third sample image output by the generator; when the training of the discriminator meets the first iteration condition, fixing the model parameters of the discriminator, and training the generator based on the discrimination results of the discriminator for the third sample image and the first sample image, wherein the generator is trained to generate a third sample image whose similarity with the first sample image is greater than a preset threshold; when the training of the generator meets the second iteration condition, continuing to fix the model parameters of the generator and train the discriminator.

[0068] Optionally, the target processing system can perform the following steps to train the generator and discriminator:

[0069] Step 1: Train the discriminator;

[0070] 1.1. Randomly extract a batch of training samples from the first training sample set. Optionally, all training samples can also be selected.

[0071] 1.2. Use the generator to process the noisy image (i.e., the second sample image) to obtain a denoised image (i.e., the third sample image).

[0072] 1.3. Input the denoised image and the real clear image (i.e., the first sample image) into the discriminator together, where the label of the real image can be "real" and the label of the denoised image can be "generated".

[0073] 1.4. The discriminator measures the gap between its prediction results and the true label by calculating the loss function (e.g., binary cross entropy loss), and updates the parameters of the discriminator through backpropagation so that it can better distinguish between real images and denoised images during the next training. For example, if the discriminator correctly classifies the real image as "real" and the denoised image as "generated", the loss will be lower, otherwise the loss will be higher.

[0074] 1.5. Repeat steps 1.1-1.4 until a first iteration condition is met, wherein the first iteration condition may be that the number of iterations of the current round of training (one alternating training round) reaches a first preset value.

[0075] Step 2: Train the generator;

[0076] 2.1. Randomly select a batch of training samples from the first training sample set.

[0077] 2.2. Use the generator to process the noisy image (i.e., the second sample image) to obtain a denoised image (i.e., the third sample image).

[0078] 2.3. Input the denoised image and the real clear image (i.e., the first sample image) into the discriminator together, where the label of the real image can be "real" and the label of the denoised image can be "generated".

[0079] 2.4. The generator's loss function is based on the output of the discriminator. For example, if the discriminator classifies the generated denoised image as "real", the generator's loss will be low, otherwise, the loss will be high. Optionally, the generator's total loss can be a weighted sum of the discriminator's loss and the L1 loss (absolute error loss), and the generator's parameters are updated by optimizing this total loss. The generator's parameters are updated through backpropagation so that the denoised images it generates during the next training can better deceive the discriminator, that is, closer to real images.

[0080] 2.5. Repeat steps 2.1-2.4 until a second iteration condition is met. The second iteration condition may be that the number of iterations in the current round of training (one alternating training round) reaches a second preset value. The first preset value may be the same as or different from the first preset value.

[0081] Step 3: Repeat steps 1 and 2, alternating between training the discriminator and generator until the target iteration condition is met. The target iteration condition can be 10 rounds of alternating training, or the target iteration condition can be that the image quality of the denoised image output by the generator meets a specific evaluation standard, such as a preset peak signal-to-noise ratio threshold.

[0082] It should be noted that through the above method, effective training of the generator and the discriminator is achieved, thereby improving the denoising effect of the image denoising model.

[0083] Optionally, in the business information processing method provided in the embodiment of the present application, the information recognition model is obtained by: obtaining a second training sample set, wherein the training sample of the second training sample set is a fourth sample image, the fourth sample image is obtained after noise reduction processing of the scanned image of the sample paper business document, and the true label of the second training sample set is the fourth sample image marked with the information to be reviewed; training the initial information recognition model through the second training sample set to obtain the information recognition model.

[0084] Optionally, the information recognition model may use a convolutional neural network model.

[0085] Optionally, the target processing system can train an information recognition model based on the second training sample set. The training sample of the second training sample set is a fourth sample image, which is obtained by denoising a scanned image of a sample paper business document. The true label of the second training sample set is the fourth sample image marked with the information to be reviewed. For example, a historical paper business document or a simulated paper business document is used as a sample paper document, and then a scanned image of the sample paper document is scanned and denoised using an image denoising model to obtain a fourth sample image. The fourth sample image can be manually annotated, for example, by using a marking box to mark the information to be reviewed in the fourth sample image, thereby obtaining a true label.

[0086] After obtaining the second training sample set, the initial information recognition model can be trained using the second training sample set to obtain an information recognition model.

[0087] In an optional embodiment, the image data input to the model may be normalized before training, cross entropy loss may be used as a loss function during training, and an optimizer may be used to optimize model parameters.

[0088] In an optional embodiment, Figure 3 is a structural diagram of the information recognition model provided in the embodiment of the present application, such as Figure 3 As shown, the information recognition model can include a fully connected layer, a "convolutional layer + pooling layer" combination, an empty convolutional layer, a "convolutional layer + pooling layer" combination, and a normalized output layer. By adding a dilated convolution layer (i.e., an empty convolutional layer), the model can appropriately expand the receptive field and achieve more detailed segmentation. The "convolutional layer + pooling layer" combination includes at least one convolutional layer and at least one pooling layer. The number and arrangement of convolutional and pooling layers can be set according to actual needs.

[0089] It should be noted that, through the above method, effective training of the initial information recognition model is achieved, thereby improving the processing accuracy of the trained recognition model.

[0090] Optionally, in the business information processing method provided in the embodiment of the present application, the annotated content in the second image and the business message information of the target business are compared to obtain a comparison result, including: obtaining a business template associated with the target business, wherein the business template is used to record the attributes that need to be extracted from the paper business document of the target business; performing text recognition on the annotated content in the second image to obtain multiple text fields; matching the multiple text fields with the attributes in the business template to obtain attribute values ​​corresponding to the attributes in the business template; comparing the attribute values ​​corresponding to the attributes in the business template with the business message information of the target business to obtain a comparison result.

[0091] Optionally, the business template is used to record the attributes that need to be extracted from the paper business documents of the target business. For example, the attributes can be user name, transaction amount, date, ID number, etc. The business template can also define attribute value characteristics of the attributes (such as the format of the attribute value, the length of the attribute value, the type of the attribute value, etc.) or regular expressions. The attribute value characteristics or regular expressions are used to match the annotation content and attributes. For example, when processing a loan application, the corresponding business template will list the attributes to be extracted, such as customer ID, loan amount, loan time, etc. The business template can also list the attribute value characteristics of these attributes. For example, the value of the customer ID is a 12-digit code, and the value of the loan time is 8 digits, and it contains the words year, month, and day.

[0092] Optionally, the target processing system may use optical character recognition technology to perform text recognition on the annotation content in the second image to obtain multiple text fields.

[0093] In an optional embodiment, after obtaining multiple text fields, the target processing system can match the multiple text fields with the attributes in the business template based on the attribute value features or regular expressions in the business template to obtain attribute values ​​corresponding to the attributes in the business template.

[0094] After obtaining the attribute value corresponding to the attribute in the service template, a comparison may be performed based on the attribute value corresponding to the attribute in the service template and the service message information of the target service to obtain a comparison result.

[0095] It should be noted that, through the above-mentioned method, the attributes of each information in the annotation content can be effectively determined, thereby improving the accuracy of the comparison result.

[0096] Optionally, in the business information processing method provided in the embodiment of the present application, a comparison is performed based on the attribute value corresponding to the attribute in the business template and the business message information of the target business to obtain a comparison result, including: for each attribute in the business template, determining the target attribute value corresponding to the attribute from the business message information; when the attribute value of each attribute is the same as the target attribute value of the attribute, determining that the comparison result represents that the annotation content is exactly the same as the business message information; when there is an attribute whose attribute value is different from the target attribute value, determining that the comparison result represents that the annotation content is different from the business message information.

[0097] Optionally, for each attribute in the business template, the target processing system can find the attribute from the business message information, and then determine the text located at a preset position of the attribute (such as to the right of the attribute, below the attribute, etc.) as the target attribute value.

[0098] After obtaining the target attribute value, the target processing system can compare the target attribute value and the attribute value of the attribute. If the target attribute value and the attribute value of a certain attribute are the same, the image content corresponding to the attribute value in the second image can be marked green. If the target attribute value and the attribute value of a certain attribute are different, the image content corresponding to the attribute value in the second image can be marked red.

[0099] Optionally, when the attribute value of each attribute is the same as the target attribute value of the attribute, it is determined that the comparison result represents that the annotation content is exactly the same as the business message information; when there is an attribute whose attribute value is different from the target attribute value, it is determined that the comparison result represents that the annotation content is different from the business message information.

[0100] It should be noted that by determining the comparison result between the annotation content in the second image and the business message information of the target business based on the comparison result of each of the above attributes, the accuracy of the determined comparison result is improved, thereby improving the audit accuracy.

[0101] Optionally, in the business information processing method provided in the embodiment of the present application, the audit result for the paper business document is determined based on the comparison result, including: when the comparison result indicates that the marked content is exactly the same as the business message information, determining that the audit result indicates that the paper business document has passed the review; when the comparison result indicates that the marked content is different from the business message information, determining that the audit result indicates that the paper business document has failed the review.

[0102] Optionally, after obtaining the audit results, the second image with color markings (i.e., the red and green markings described above), the audit results, and the attribute values ​​corresponding to the attributes in the determined business template can be integrated together, for example, into a single document to obtain a target document. The target document can then be displayed to a target recipient, such as an administrator at a financial institution. This ensures the accuracy and traceability of the output images and information for subsequent inquiries.

[0103] In some embodiments, after determining that the audit result indicates that the paper business document has passed the audit, the target entity of the financial institution may proceed to the next step. The next step may involve the target entity confirming whether to approve the requested content in the business message. For example, an administrator may confirm whether to grant a loan to the customer, whether to allow the user to transfer funds, or whether to allow the user to open an account.

[0104] It should be noted that the above method improves the accuracy of the determined audit results.

[0105] In an optional embodiment, Figure 4 This is the process of the method for processing business information provided by the embodiment of this application Figure 2 ,according to Figure 4 An optional application process for processing business information is described. Figure 4 As shown, the target processing system can obtain a scanned image of a paper business document for a target business to be reviewed. It then uses an image noise reduction model to perform noise reduction on the scanned image, generating a first image. It then uses an information recognition model to annotate the information to be reviewed in the first image, generating a second image. Subsequently, based on the business template and the annotated content, it determines the attribute values ​​corresponding to the attributes in the business template, extracting the template information. Finally, it compares the attribute values ​​corresponding to the attributes in the business template with the business message information of the target business to obtain a comparison result, and then determines the review result for the paper business document based on the comparison result.

[0106] It can be seen that the method provided by this application achieves the purpose of automatically reviewing whether paper business documents are consistent with online business message information, realizes the technical effect of improving review efficiency, and solves the technical problem of low review efficiency when reviewing whether paper business documents of specific business are consistent with online business message information in related technologies.

[0107] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0108] Example 2

[0109] The embodiment of the present application further provides a device for processing business information. It should be noted that the device for processing business information in the embodiment of the present application can be used to execute the method for processing business information provided in the embodiment of the present application. The following introduces the device for processing business information provided in the embodiment of the present application.

[0110] According to an embodiment of the present application, a device for implementing the above-mentioned method for processing business information is also provided. Figure 5 As shown, the device includes:

[0111] A first acquisition module 501 is used to acquire a scanned image of a paper business document to be reviewed of a target business;

[0112] A first processing module 502 is configured to perform noise reduction processing on the scanned image using an image noise reduction model to obtain a first image;

[0113] The second processing module 503 is configured to mark the information to be reviewed in the first image using an information recognition model to obtain a second image;

[0114] The first determination module 504 is configured to compare the annotation content in the second image with the business message information of the target business to obtain a comparison result, and determine a review result for the paper business document based on the comparison result.

[0115] In an embodiment of the present application, by obtaining a scanned image of a paper business document to be audited for a target business and annotating the information to be audited in the image through an information recognition model, effective extraction of audit-related content in the paper business document is achieved, avoiding the impact of invalid information on the audit progress. By comparing the annotated content with the business message information of the target business to determine the audit result of the paper business document, automatic audit of the consistency between the paper business document and the business message information is achieved, thereby improving audit efficiency. In addition, using an image denoising model to perform noise reduction processing on the scanned image can effectively improve image quality, making key information easier to identify, thereby improving audit efficiency and audit accuracy.

[0116] It can be seen that the method provided by this application achieves the purpose of automatically reviewing whether paper business documents are consistent with online business message information, realizes the technical effect of improving review efficiency, and solves the technical problem of low review efficiency when reviewing whether paper business documents of specific business are consistent with online business message information in related technologies.

[0117] Optionally, in the business information processing device provided in the embodiment of the present application, the business information processing device also includes: a second acquisition module, used to obtain a first training sample set, wherein the training samples in the first training sample set include a first sample image and a second sample image, and the second sample image is obtained by adding noise to the first sample image; a second determination module, used to alternately train the discriminator and the generator in the generative adversarial network through the first training sample set until the target iteration condition is met, and the trained generator is determined as the image denoising model.

[0118] Optionally, in the business information processing device provided in the embodiment of the present application, the second determination module also includes: a first training sub-module, used to fix the model parameters of the generator, perform denoising on the second sample image through the generator to obtain a third sample image, and train the discriminator based on the third sample image and the first sample image, wherein the discriminator is trained to distinguish between the real first sample image and the third sample image output by the generator; a second training sub-module, used to fix the model parameters of the discriminator when the training of the discriminator meets the first iteration condition, and train the generator based on the discrimination results of the discriminator for the third sample image and the first sample image, wherein the generator is trained to generate a third sample image whose similarity with the first sample image is greater than a preset threshold; a third training sub-module, used to continue to fix the model parameters of the generator and train the discriminator when the training of the generator meets the second iteration condition.

[0119] Optionally, in the business information processing device provided in the embodiment of the present application, the business information processing device also includes: a third acquisition module, used to obtain a second training sample set, wherein the training sample of the second training sample set is a fourth sample image, and the fourth sample image is obtained after noise reduction processing of the scanned image of the sample paper business document, and the true label of the second training sample set is the fourth sample image marked with information to be reviewed; a training module, used to train the initial information recognition model through the second training sample set to obtain an information recognition model.

[0120] Optionally, in the business information processing device provided in the embodiment of the present application, the first determination module also includes: an acquisition sub-module, used to obtain a business template associated with the target business, wherein the business template is used to record the attributes that need to be extracted from the paper business document of the target business; an identification sub-module, used to perform text recognition on the annotation content in the second image to obtain multiple text fields; a processing sub-module, used to match the multiple text fields with the attributes in the business template to obtain attribute values ​​corresponding to the attributes in the business template; a comparison sub-module, used to compare the attribute values ​​corresponding to the attributes in the business template with the business message information of the target business to obtain a comparison result.

[0121] Optionally, in the business information processing device provided in the embodiment of the present application, the comparison submodule also includes: a first determination submodule, used to determine, for each attribute in the business template, the target attribute value corresponding to the attribute from the business message information; a second determination submodule, used to determine that the comparison result representation annotation content is exactly the same as the business message information when the attribute value of each attribute is the same as the target attribute value of the attribute; and a third determination submodule, used to determine that the comparison result representation annotation content is different from the business message information when there is an attribute whose attribute value is different from the target attribute value.

[0122] Optionally, in the business information processing device provided in the embodiment of the present application, the first determination module also includes: a fourth determination sub-module, used to determine that the audit result indicates that the paper business document has passed the audit when the comparison result indicates that the marked content is exactly the same as the business message information; and a fifth determination sub-module, used to determine that the audit result indicates that the paper business document has failed the audit when the comparison result indicates that the marked content is different from the business message information.

[0123] It should be noted that the first acquisition module 501, the first processing module 502, the second processing module 503, and the first determination module 504 correspond to steps S201 to S204 in Example 1. The examples and application scenarios implemented by the four modules and the corresponding steps are the same, but are not limited to the contents disclosed in the above-mentioned Example 1. It should be noted that the above-mentioned modules or units can be hardware components or software components stored in a memory (e.g., memory 104) and processed by one or more processors (e.g., processors 102a, 102b, ..., 102n). The above-mentioned modules can also be part of the device and can be run in the computer terminal 10 provided in Example 1.

[0124] Example 3

[0125] An embodiment of the present application may provide an electronic device, Figure 6 This is a structural block diagram of an electronic device according to an embodiment of the present application. Figure 6 As shown, the electronic device may include: one or more ( Figure 6 Only one is shown) processor 1002, memory 1004, storage controller, and peripheral interface, wherein the peripheral interface is connected to the radio frequency module, audio module and display.

[0126] Among them, the memory can be used to store software programs and modules, such as program instructions / modules corresponding to the methods and devices in the embodiments of the present application. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory, that is, implementing the above-mentioned method. The memory may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory may further include a memory remotely arranged relative to the processor, and these remote memories may be connected to the terminal via a network. Examples of the above-mentioned network include but are not limited to the Internet, an intranet, a local area network, a mobile communication network and a combination thereof.

[0127] The processor can call the information and application stored in the memory through the transmission device to perform the following steps: obtain a scanned image of the paper business document to be reviewed of the target business; perform noise reduction processing on the scanned image through the image noise reduction model to obtain a first image; mark the information to be reviewed in the first image through the information recognition model to obtain a second image; compare the marked content in the second image with the business message information of the target business to obtain a comparison result, and determine the review result for the paper business document based on the comparison result.

[0128] The processor can also call the information and application stored in the memory through the transmission device to perform the following steps: obtaining a first training sample set, wherein the training samples in the first training sample set include a first sample image and a second sample image, and the second sample image is obtained by adding noise to the first sample image; alternately training the discriminator and the generator in the generative adversarial network through the first training sample set until the target iteration condition is met, and determining the trained generator as the image denoising model.

[0129] The processor can also call the information and application stored in the memory through the transmission device to perform the following steps: fix the model parameters of the generator, perform denoising on the second sample image through the generator to obtain a third sample image, and train the discriminator based on the third sample image and the first sample image, wherein the discriminator is trained to distinguish between the real first sample image and the third sample image output by the generator; when the training of the discriminator meets the first iteration condition, fix the model parameters of the discriminator, and train the generator based on the discrimination results of the discriminator for the third sample image and the first sample image, wherein the generator is trained to generate a third sample image whose similarity with the first sample image is greater than a preset threshold; when the training of the generator meets the second iteration condition, continue to fix the model parameters of the generator and train the discriminator.

[0130] The processor can call the information and application stored in the memory through the transmission device to perform the following steps: obtain a second training sample set, wherein the training sample of the second training sample set is a fourth sample image, the fourth sample image is obtained after noise reduction processing of the scanned image of the sample paper business document, and the true label of the second training sample set is the fourth sample image marked with the information to be reviewed; train the initial information recognition model through the second training sample set to obtain an information recognition model.

[0131] The processor can call the information and application stored in the memory through the transmission device to perform the following steps: obtain a business template associated with the target business, wherein the business template is used to record the attributes that need to be extracted from the paper business documents of the target business; perform text recognition on the annotation content in the second image to obtain multiple text fields; match the multiple text fields with the attributes in the business template to obtain attribute values ​​corresponding to the attributes in the business template; compare the attribute values ​​corresponding to the attributes in the business template with the business message information of the target business to obtain a comparison result.

[0132] The processor can call the information and application stored in the memory through the transmission device to perform the following steps: for each attribute in the business template, determine the target attribute value corresponding to the attribute from the business message information; when the attribute value of each attribute is the same as the target attribute value of the attribute, determine that the comparison result representation annotation content is exactly the same as the business message information; when there is an attribute whose attribute value is different from the target attribute value, determine that the comparison result representation annotation content is different from the business message information.

[0133] The processor can call the information and application stored in the memory through the transmission device to perform the following steps: when the comparison result indicates that the marked content is exactly the same as the business message information, determine that the audit result indicates that the paper business document has passed the audit; when the comparison result indicates that the marked content is different from the business message information, determine that the audit result indicates that the paper business document has failed the audit.

[0134] In an embodiment of the present application, by obtaining a scanned image of a paper business document to be audited for a target business and annotating the information to be audited in the image through an information recognition model, effective extraction of audit-related content in the paper business document is achieved, avoiding the impact of invalid information on the audit progress. By comparing the annotated content with the business message information of the target business to determine the audit result of the paper business document, automatic audit of the consistency between the paper business document and the business message information is achieved, thereby improving audit efficiency. In addition, using an image denoising model to perform noise reduction processing on the scanned image can effectively improve image quality, making key information easier to identify, thereby improving audit efficiency and audit accuracy.

[0135] It can be seen that the method provided by this application achieves the purpose of automatically reviewing whether paper business documents are consistent with online business message information, realizes the technical effect of improving review efficiency, and solves the technical problem of low review efficiency when reviewing whether paper business documents of specific business are consistent with online business message information in related technologies.

[0136] It can be understood by those skilled in the art that Figure 6 The structure shown is for illustration only, and the electronic device may also be a smart phone (such as an Android phone, an iOS phone, etc.), a tablet computer, a PDA, a mobile Internet device (Mobile Internet Devices, MID), a PAD, or other terminal devices. Figure 6 It does not limit the structure of the above electronic device. For example, the electronic device may also include Figure 6 More or fewer components (such as network interfaces, display devices, etc.) shown in, or with Figure 6 Different configurations shown.

[0137] A person skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing the hardware related to the terminal device through a program, and the program can be stored in a computer-readable storage medium, which may include: a flash drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0138] Example 4

[0139] The embodiment of the present application further provides a storage medium. Optionally, in this embodiment, the storage medium can be used to store the program code executed by the method for processing business information provided in the first embodiment.

[0140] Optionally, in this embodiment, the above-mentioned storage medium may be located in any computer terminal in a computer terminal group in a computer network, or in any mobile terminal in a mobile terminal group.

[0141] The present application also provides a computer program product, which, when executed on a data processing device, is suitable for executing the steps of the method for processing business information.

[0142] The serial numbers of the above embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.

[0143] In the above embodiments of the present application, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, please refer to the relevant description of other embodiments.

[0144] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0145] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0146] In addition, the functional units in the various embodiments of the present application may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0147] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for enabling a computer device (which can be a personal computer, a server or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk or an optical disk.

[0148] The above is only a preferred embodiment of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A method for processing business information, characterized in that: include: Obtain scanned images of paper business documents to be reviewed for the target business; Performing noise reduction processing on the scanned image using an image noise reduction model to obtain a first image; Annotating the information to be reviewed in the first image using an information recognition model to obtain a second image; The annotated content in the second image is compared with the business message information of the target business to obtain a comparison result, and a review result for the paper business document is determined based on the comparison result.

2. The method according to claim 1, characterized in that The image denoising model is obtained by: Acquire a first training sample set, wherein the training samples in the first training sample set include a first sample image and a second sample image, and the second sample image is obtained by adding noise to the first sample image; The discriminator and the generator in the generative adversarial network are alternately trained using the first training sample set until the target iteration condition is met, and the trained generator is determined as the image denoising model.

3. The method according to claim 2, characterized in that The discriminator and the generator in the generative adversarial network are alternately trained using the first training sample set, including: Fixing the model parameters of the generator, performing denoising on the second sample image using the generator to obtain a third sample image, and training the discriminator based on the third sample image and the first sample image, wherein the discriminator is trained to distinguish between the true first sample image and the third sample image output by the generator; When the training of the discriminator satisfies a first iteration condition, fixing the model parameters of the discriminator, and training the generator based on the discrimination result of the discriminator on the third sample image and the first sample image, wherein the generator is trained to generate a third sample image having a similarity with the first sample image greater than a preset threshold; When the training of the generator meets the second iteration condition, the model parameters of the generator continue to be fixed and the discriminator is trained.

4. The method according to claim 1, wherein The information recognition model is obtained in the following way: Obtain a second training sample set, wherein the training sample of the second training sample set is a fourth sample image, the fourth sample image is obtained by performing noise reduction processing on a scanned image of a sample paper business document, and the true label of the second training sample set is the fourth sample image marked with information to be reviewed; The initial information recognition model is trained using the second training sample set to obtain the information recognition model.

5. The method according to claim 1, characterized in that Comparing the annotated content in the second image with the service message information of the target service to obtain a comparison result includes: Acquire a business template associated with the target business, wherein the business template is used to record attributes that need to be extracted from a paper business document of the target business; performing text recognition on the annotation content in the second image to obtain a plurality of text fields; Matching the multiple text fields with the attributes in the business template to obtain attribute values ​​corresponding to the attributes in the business template; The comparison result is obtained based on comparing the attribute value corresponding to the attribute in the service template with the service message information of the target service.

6. The method according to claim 5, characterized in that Comparing the attribute value corresponding to the attribute in the service template with the service message information of the target service to obtain the comparison result includes: For each attribute in the service template, determining a target attribute value corresponding to the attribute from the service message information; When the attribute value of each attribute is identical to the target attribute value of the attribute, determining that the comparison result indicates that the annotation content is completely identical to the service message information; In the case that there is an attribute whose attribute value is different from the target attribute value, determining the comparison result indicates that there is a difference between the annotation content and the service message information.

7. The method according to claim 1, characterized in that Determining the audit results for the paper business document based on the comparison results includes: If the comparison result indicates that the marked content is completely identical to the business message information, determining that the review result indicates that the paper business document has passed the review; When the comparison result indicates that the marked content is different from the business message information, it is determined that the review result indicates that the paper business document has failed the review.

8. A device for processing business information, characterized in that: include: A first acquisition module is used to acquire a scanned image of a paper business document to be reviewed of a target business; A first processing module, configured to perform noise reduction processing on the scanned image using an image noise reduction model to obtain a first image; A second processing module is configured to mark the information to be reviewed in the first image using an information recognition model to obtain a second image; The first determination module is configured to compare the annotation content in the second image with the business message information of the target business to obtain a comparison result, and determine a review result for the paper business document based on the comparison result.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium includes a stored executable program, wherein when the executable program is run, the device where the computer-readable storage medium is located is controlled to execute the business information processing method according to any one of claims 1 to 7.

10. An electronic device, characterized in that: include: a memory storing an executable program; A processor is used to run the program, wherein the program executes the business information processing method according to any one of claims 1 to 7 when running.

11. A computer program product comprising computer instructions, characterized in that When the computer instructions are executed by a processor, the steps of the business information processing method according to any one of claims 1 to 7 are implemented.

Images