Information input method and device, equipment and medium

Abstract

The invention relates to the technical field of computers, and discloses an information input method and device, equipment and a medium, and the method comprises the steps: obtaining to-be-recognized image information, analyzing the image information through a deep learning model, and obtaining a key text in the image information; performing structured processing on the key text to obtain the key text after the structured processing; performing classification matching on the key text after the structural processing according to a preset field to obtain classified classification information after classification; and filling the classification information into a preset webpage to obtain a target webpage after information input. The method can be applied to business management program systems such as financial science and technology, medical health, old-age care and the like, the information identification precision is improved, and the information input efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to an information entry method, device, equipment and medium. Background Art

[0002] With the rapid development of artificial intelligence (AI) technology, intelligent recognition and automated data entry are becoming increasingly common in the financial and healthcare industries. For example, in the financial sector, banks, insurance companies, and securities firms need to process large amounts of user identity information, contract data, and transaction records. In the healthcare industry, hospitals, clinics, and health insurance companies also face the need to enter large amounts of medical records, prescriptions, and test reports.

[0003] However, the application of current information entry technology in the financial and healthcare industries still faces numerous challenges. First, optical character recognition technology has low recognition rates for handwriting, low-contrast text, or content against complex backgrounds, impacting the accuracy of information entry. Second, related technologies struggle to automatically match field formats across different business systems, requiring manual intervention in information entry. Third, information formats differ across institutions, resulting in poor compatibility and limiting data entry efficiency. Therefore, improving both information recognition accuracy and efficiency is crucial. Summary of the Invention

[0004] The present invention provides an information entry method, apparatus, computer equipment and medium to solve the technical problems of low accuracy and low efficiency of information entry in related technologies.

[0005] In a first aspect, a method for inputting information is provided, comprising:

[0006] Obtain image information to be identified, and analyze the image information through a deep learning model to obtain key text in the image information;

[0007] Performing structural processing on the key text to obtain a structured key text;

[0008] Classify and match the key text after the structured processing according to preset fields to obtain classified classification information;

[0009] The classification information is filled into a preset web page to obtain a target web page after the information is entered.

[0010] In a second aspect, an information entry device is provided, comprising:

[0011] An acquisition module is used to acquire image information to be identified and analyze the image information through a deep learning model to obtain key text in the image information;

[0012] A structured processing module, configured to perform structured processing on the key text to obtain the structured key text;

[0013] A classification module is used to classify and match the key text after the structured processing according to preset fields to obtain classified classification information;

[0014] The information input module is used to fill the classification information into a preset web page to obtain a target web page after the information is input.

[0015] In a third aspect, a computer device is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above-mentioned information entry method when executing the computer program.

[0016] In a fourth aspect, a computer-readable storage medium is provided, which stores a computer program. When the computer program is executed by a processor, it implements the steps of the above-mentioned information entry method. In the scheme implemented based on the above-mentioned information entry method, device, computer equipment and storage medium, image information to be identified can be obtained, and the image information can be analyzed by a deep learning model to obtain key text in the image information. Furthermore, the key text can be structured to obtain the key text after structured processing, and the key text after structured processing can be classified and matched according to preset fields to obtain classified classification information. Thus, the classification information can be filled into a preset web page to obtain the target web page after information entry. In the present invention, the image information to be identified can be analyzed by a deep learning model, the key text therein can be accurately extracted, and the key text can be structured and classified and matched, thereby realizing efficient and intelligent information entry. When applied to the financial field, this method can significantly improve the automation level of information entry such as bill recognition, contract parsing, and identity authentication. For example, in bank account opening, loan approval, insurance claims, and other services, key information from documents such as ID cards, bank statements, and invoices can be automatically extracted and populated into the business system, improving the accuracy and compliance of information entry and reducing labor costs and error rates. Furthermore, this solution can optimize the data entry process, enabling financial institutions to obtain key information more quickly and improving information entry efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments of the present invention. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0018] Figure 1This is a schematic diagram of an application environment of an information entry method according to an embodiment of the present invention;

[0019] Figure 2 This is a flow chart of an information entry method according to an embodiment of the present invention;

[0020] Figure 3 yes Figure 1 A schematic flow chart of a specific implementation of step S10;

[0021] Figure 4 It is a structural diagram of an information entry device in one embodiment of the present invention;

[0022] Figure 5 is a structural diagram of a computer device in one embodiment of the present invention;

[0023] Figure 6 FIG. 2 is another structural diagram of a computer device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0024] 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 them. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0025] The information entry method provided by the embodiment of the present invention can be applied to Figure 1In an application environment, a client communicates with a server via a network. The server can obtain image information to be identified through the client, analyze the image information using a deep learning model, obtain key text in the image information, perform structured processing on the key text, obtain structured key text, classify and match the structured key text according to preset fields, obtain classified information, populate the classified information into a preset web page, obtain a target web page after information entry, and finally feed the target web page back to the client. In the present invention, the image information to be identified can be parsed using a deep learning model, key text can be accurately extracted, and structured and classified and matched, thereby achieving efficient and intelligent information entry. When applied to the financial field, this method can significantly improve the automation level of information entry such as bill recognition, contract parsing, and identity authentication. For example, in bank account opening, loan approval, insurance claims, and other businesses, key information from documents such as ID cards, bank statements, and invoices can be automatically extracted and populated into the business system, improving the accuracy and compliance of information entry and reducing labor costs and error rates. Furthermore, this solution can optimize the data entry process, enabling financial institutions to more quickly access key information and improve information entry efficiency. The client can be, but is not limited to, various personal computers, laptops, smartphones, tablets, and portable wearable devices. The server can be implemented as a standalone server or a server cluster consisting of multiple servers. The present invention is described in detail below using specific embodiments.

[0026] See also Figure 2 As shown, Figure 2 A flowchart of an information entry method provided in an embodiment of the present invention includes the following steps:

[0027] S10: Obtain image information to be identified, and analyze the image information through a deep learning model to obtain key text in the image information.

[0028] For example, image information to be identified may be obtained, such as bank checks, invoices, transaction documents, or identification documents submitted by customers, etc., which are not limited in this application. The above-mentioned image information may be obtained from mobile phone photography, scanners, or other digital means. Furthermore, deep learning models (such as OCR recognition technology, natural language processing technology, etc.) may be used to analyze these image information and extract key text data, such as amount, transaction time, account information, or contract terms.

[0029] Among them, such as Figure 3 As shown, in step S10, that is, analyzing the image information through a deep learning model to obtain key text in the image information, the following steps are included:

[0030] S11: performing a preprocessing operation on the image information to obtain preprocessed image information.

[0031] The preprocessing operation includes grayscale operation, denoising operation, edge detection and image enhancement operation.

[0032] S12: Perform text detection on the pre-processed image information using a deep learning network to extract a text area.

[0033] S13: performing text analysis on the text area to extract key text from the image information.

[0034] In steps S11-S13, the image information must first be preprocessed to improve the accuracy of subsequent text recognition. Preprocessing includes grayscaling, which converts color images to grayscale to reduce computational complexity; denoising, which removes noise from the scanning or shooting process, such as watermarks, shadows, and background interference; edge detection, which highlights the edges of text in the image to make the text areas clearer; and image enhancement, which adjusts brightness and contrast to improve text readability. These operations optimize image quality and provide better input data for subsequent text detection and analysis.

[0035] Furthermore, after image preprocessing, a deep learning network can be used to perform text detection on the image, accurately locating the text area within a complex background. This allows the position of the text within the image to be identified and a coordinate frame for the text area to be generated. For example, in a bank bill review scenario, key information areas such as the account name, amount, and date on the bill can be automatically detected, while ignoring any interference from seals or watermarks in the background.

[0036] For example, after completing the detection of text areas, these areas can be further parsed to extract and identify key text information in the image. For example, optical character recognition technology can be used to parse the text area line by line to identify the specific character content. In financial applications, this step can be used to extract information such as invoice numbers, contract terms, transaction details, and match them with the database. For example, in the automated reimbursement process, the merchant name, amount, and tax number can be extracted from the invoice image and compared with the company's financial system to ensure the accuracy and compliance of the data, thereby greatly improving the efficiency and accuracy of financial services.

[0037] S20: Performing structural processing on the key text to obtain a key text after structural processing.

[0038] In some embodiments, the key text is structured to obtain the structured key text, including: deformatting the key text to obtain the processed key text; analyzing the processed key text and splitting the processed text according to a preset format to obtain the structured key text.

[0039] It should be understood that structuring key text can improve data readability and facilitate subsequent analysis. For example, when processing documents such as bank transaction receipts, invoices, and contracts, the initially extracted key text often contains redundant formatting information, such as line breaks, spaces, special symbols, or noise characters. Therefore, the text must first be deformatted, that is, irrelevant characters are removed and the text format is unified. For another example, transaction records extracted from bank statements may contain extra spaces or special symbols, so they can be standardized first to ensure the neatness of the data.

[0040] For example, after deformatting, the text content can be analyzed and split, and stored as structured data according to a preset format. For example, when processing bank statements, the extracted text may contain "2025 / 03 / 10 transfer 5,000 yuan Industrial and Commercial Bank of China", so it can be split into date (2025-03-10), transaction type (transfer), amount (5,000 yuan), counterparty (Industrial and Commercial Bank of China) according to the preset format and stored in a database or form. For another example, for contract parsing, key terms such as contract number, party A and party B information, signing date, and payment method can be extracted and split, thereby supporting financial institutions in automated compliance reviews, data reconciliation, or risk assessments. Through this structured processing, the originally scattered text data becomes clear and usable, greatly improving the intelligence and automation level of financial services.

[0041] S30: Classify and match the structured key text according to preset fields to obtain classified classification information.

[0042] In some embodiments, the classification and matching of the structured key text according to preset fields includes: determining the similarity value between the preset field and the structured key text by means of cosine similarity; if the similarity value is less than a preset threshold, determining that the classification information is that the structured key text matches the preset field.

[0043] For example, the key text after structured processing can be classified and matched, thereby improving data organization and business process automation. For example, in scenarios such as bank transaction flow, invoice review, and contract analysis, the extracted key text needs to be mapped to the corresponding business fields, such as transaction type, payment method, account opening bank, invoice tax rate, etc. In order to achieve accurate matching, the cosine similarity can be used to calculate the similarity value between the text and the preset field. For example, in an enterprise financial system, the extracted transaction description may be "POS consumption-catering-500 yuan", and the similarity between the text and the preset fields (such as "catering", "transportation", and "office supplies") can be calculated to determine that the transaction belongs to the "catering" category.

[0044] For example, if the calculated similarity value is greater than a preset threshold (for example, the preset threshold is 80%, 90%, etc., which is not limited in this embodiment), the classification result can be determined; if the calculated similarity value is less than the preset threshold, it can be further processed, such as manual review or secondary calculation. In addition, for contract parsing scenarios, similarity calculations can be performed on the extracted clause content, and automatically matched to preset fields such as "payment terms", "breach of contract liability", and "contract term", thereby helping financial institutions to efficiently classify contract content and reduce the workload of manual review. Through this intelligent matching method based on cosine similarity, a large amount of unstructured data can be automatically classified, improving the accuracy and efficiency of data management.

[0045] S40: Filling the classification information into a preset web page to obtain a target web page after the information is entered.

[0046] For example, in scenarios such as bank loan approval, invoice reimbursement, or account opening, classified information can be automatically populated into pre-set web pages, such as loan application pages, financial reimbursement systems, or contract management systems. This allows fields that previously required manual entry (such as customer name, loan amount, transaction category, invoice number, etc.) to be intelligently populated by the system, reducing human input errors and accelerating business processes.

[0047] In some embodiments, after filling the classification information into a preset web page and obtaining the target web page after the information is entered, it also includes: displaying the entered information through the target web page so that the user can verify the entered information; if the user confirms that the entered information is correct, an entry log is generated.

[0048] For example, after completing the information entry, the entered data can be displayed on the target webpage, allowing the user to verify and confirm the information. If the user confirms that the information is correct, an entry log is automatically generated, recording the details of the entry, including timestamp, user ID, data source, etc., to meet the audit and compliance requirements of financial institutions. If the user discovers an error, they can manually correct it or trigger a re-identification process. This automatic entry + manual verification model not only improves the accuracy of information entry, but also takes into account user operational flexibility and compliance.

[0049] In some embodiments, the method further includes: obtaining a training data set and a pre-trained model; wherein the training data set includes historical image information; annotating the training data set to obtain an annotation result, wherein the annotation result includes key text corresponding to the historical image information; training the pre-trained model through the training data set and the annotation result to obtain the deep learning model.

[0050] Specifically, several historical image information images can be obtained to serve as training datasets for training. Furthermore, each historical image information can be annotated to obtain annotation results including key text corresponding to the historical image information. The annotation results are then used as labels for the set of input data. Each set of labeled training datasets is then input into a pre-trained model for supervised learning. Training ends when a training termination condition is met, such as when the number of training times reaches a threshold or when the model's output accuracy reaches a threshold, resulting in a fully trained deep learning model.

[0051] In an embodiment of the present application, the training data set and the annotation results can be input into a pre-training model for supervised learning, and then a deep learning model can be trained.

[0052] Based on the above embodiment, after obtaining the deep learning model, it also includes: iteratively training the deep learning model based on the training data set and the annotation results to extract data features, and calculating the loss function; using a preset method to iteratively train the loss function for the purpose of reducing the loss function value until the loss function value is less than the expected threshold; based on the loss function after iterative training, obtaining the iterative deep learning model.

[0053] It is understandable that in order to train a deep learning model with higher accuracy, the deep learning model can be repeatedly trained iteratively to continuously reduce the loss function until the loss function meets the expected threshold requirement.

[0054] It should be noted that this application does not limit the above-mentioned preset method and expected threshold. For example, the preset method can be a gradient descent algorithm, a batch gradient descent algorithm, a stochastic gradient descent algorithm, etc. This application takes the gradient descent algorithm as an example for explanation.

[0055] The purpose of the gradient descent algorithm is to find the minimum value of the loss function or converge to the minimum value through iteration. In a geometric sense, the gradient descent algorithm is that the gradient decreases fastest in the direction opposite to the vector where the function changes the fastest, so it is easier to find the minimum value of the function. Based on this, in the embodiment of the present application, the event extraction model can be repeatedly iteratively trained using the gradient descent algorithm so that the loss function is continuously reduced, thereby reducing the error of the calculation result.

[0056] In an embodiment of the present application, the gradient descent algorithm is used to repeatedly iteratively train the deep learning model so that the loss function is continuously reduced to obtain the iterative deep learning model, and then a more accurate analysis result can be obtained based on the iterative deep learning model.

[0057] It can be seen that in the above solution, the image information to be identified can be analyzed through a deep learning model, the key text can be accurately extracted, and it can be structured and classified and matched, thereby achieving efficient and intelligent information entry. When applied to the financial field, this method can significantly improve the automation level of information entry such as bill recognition, contract parsing, and identity authentication. For example, in bank account opening, loan approval, insurance claims and other businesses, key information from documents such as identity cards, bank statements, invoices, etc. can be automatically extracted and filled into the business system, improving the accuracy and compliance of information entry and reducing labor costs and error rates. In addition, this solution can also optimize the data input link of information entry, allowing financial institutions to obtain key information more quickly and improve information entry efficiency.

[0058] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0059] In one embodiment, an information entry device is provided, which corresponds one-to-one to the information entry method in the above embodiment. Figure 4 As shown, the information entry device includes an acquisition module 101, a structured processing module 102, a classification module 103, and an information entry module 104. The functional modules are described in detail as follows:

[0060] An acquisition module 101 is configured to acquire image information to be identified and analyze the image information using a deep learning model to obtain key text in the image information;

[0061] A structured processing module 102 is used to perform structured processing on the key text to obtain the structured key text;

[0062] The classification module 103 is used to classify and match the key text after the structured processing according to preset fields to obtain classified classification information;

[0063] The information input module 104 is used to fill the classification information into a preset web page to obtain a target web page after the information is input.

[0064] In one embodiment, the acquisition module 101 is specifically configured to:

[0065] Performing a preprocessing operation on the image information to obtain preprocessed image information; wherein the preprocessing operation includes grayscale operation, denoising operation, edge detection and image enhancement operation;

[0066] Performing text detection on the pre-processed image information using a deep learning network to extract a text area;

[0067] The text area is subjected to text analysis to extract key text from the image information.

[0068] In one embodiment, the structured processing module 102 is specifically configured to:

[0069] Performing deformatting processing on the key text to obtain processed key text;

[0070] The processed key text is analyzed and split according to a preset format to obtain the structured key text.

[0071] In one embodiment, the classification module 103 is further configured to:

[0072] Determining the similarity between the preset field and the key text after the structured processing by means of cosine similarity;

[0073] If the similarity value is less than a preset threshold, it is determined that the classification information matches the key text after the structured processing and the preset field.

[0074] In one embodiment, the information entry module 104 is specifically configured to:

[0075] Displaying the entered information on the target webpage so that the user can verify the entered information;

[0076] If the user confirms that the entered information is correct, an entry log is generated.

[0077] In one embodiment, the acquisition module 101 is specifically configured to:

[0078] Obtaining a training data set and a pre-trained model; wherein the training data set includes historical image information;

[0079] Annotating the training data set to obtain an annotation result, wherein the annotation result includes key text corresponding to the historical image information;

[0080] The pre-training model is trained using the training data set and the annotation results to obtain the deep learning model.

[0081] In one embodiment, the acquisition module 101 is specifically configured to:

[0082] Iteratively train the deep learning model based on the training data set and the annotation results to extract data features, and calculate the loss function;

[0083] Iteratively training the loss function using a preset method for the purpose of reducing the loss function value until the loss function value is less than an expected threshold;

[0084] Based on the loss function after iterative training, an iterative deep learning model is obtained.

[0085] The present invention provides an information entry device that can analyze the image information to be identified through a deep learning model, accurately extract the key text therein, and perform structured processing and classification matching on the information, thereby realizing efficient and intelligent information entry. When applied to the financial field, this method can significantly improve the automation level of information entry such as bill recognition, contract parsing, and identity authentication. For example, in bank account opening, loan approval, insurance claims and other businesses, key information in documents such as identity cards, bank statements, invoices, etc. can be automatically extracted and filled into the business system, thereby improving the accuracy and compliance of information entry and reducing labor costs and error rates. In addition, the solution can also optimize the data input link of information entry, allowing financial institutions to obtain key information more quickly and improve information entry efficiency.

[0086] For the specific definition of the information entry device, please refer to the definition of the information entry method above and will not be repeated here. Each module in the above-mentioned information entry device can be implemented in whole or in part by software, hardware, or a combination thereof. Each of the above-mentioned modules can be embedded in or independent of the processor in the computer device in hardware form, or can be stored in the memory of the computer device in software form, so that the processor can call and execute the operations corresponding to each of the above modules.

[0087] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 5 As shown. The computer device includes a processor, memory, network interface and database connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes non-volatile and / or volatile storage media and internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external client via a network connection. When the computer program is executed by the processor, it implements the functions or steps on the server side of an information entry method.

[0088] In one embodiment, a computer device is provided. The computer device may be a client, and its internal structure diagram may be as follows: Figure 6 As shown. The computer device includes a processor, memory, network interface, display screen and input device connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external server via a network connection. When the computer program is executed by the processor, it realizes the functions or steps of the client side of an information entry method.

[0089] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the following steps are performed:

[0090] Obtain image information to be identified, and analyze the image information through a deep learning model to obtain key text in the image information;

[0091] Performing structural processing on the key text to obtain a structured key text;

[0092] Classify and match the key text after the structured processing according to preset fields to obtain classified classification information;

[0093] The classification information is filled into a preset web page to obtain a target web page after the information is entered.

[0094] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the following steps are implemented:

[0095] Obtain image information to be identified, and analyze the image information through a deep learning model to obtain key text in the image information;

[0096] Performing structural processing on the key text to obtain a structured key text;

[0097] Classify and match the key text after the structured processing according to preset fields to obtain classified classification information;

[0098] The classification information is filled into a preset web page to obtain a target web page after the information is entered.

[0099] It should be noted that the above functions or steps that can be implemented by the computer-readable storage medium or computer device can be found in the relevant descriptions of the server side and the client side in the aforementioned method embodiment. To avoid repetition, they will not be described one by one here.

[0100] Those skilled in the art will appreciate that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchl ink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0101] Those skilled in the art will clearly understand that for the sake of convenience and brevity of description, only the division of the above-mentioned functional units and modules is used as an example. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0102] The embodiments described above are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention, and should all be included in the scope of protection of the present invention.

Claims

1. A method for inputting information, characterized in that: The method comprises: Obtain image information to be identified, and analyze the image information through a deep learning model to obtain key text in the image information; Performing structural processing on the key text to obtain a structured key text; Classify and match the key text after the structured processing according to preset fields to obtain classified classification information; The classification information is filled into a preset web page to obtain a target web page after the information is entered.

2. The method according to claim 1, characterized in that The process of parsing the text in the image information using a deep learning model to obtain key text information in the image information includes: Performing a preprocessing operation on the image information to obtain preprocessed image information; wherein the preprocessing operation includes grayscale operation, denoising operation, edge detection and image enhancement operation; Performing text detection on the pre-processed image information using a deep learning network to extract a text area; The text area is subjected to text analysis to extract key text from the image information.

3. The method according to claim 1, characterized in that The structural processing of the key text to obtain the key text after structural processing includes: Performing deformatting processing on the key text to obtain processed key text; The processed key text is analyzed and split according to a preset format to obtain the structured key text.

4. The method according to claim 1, wherein The classifying and matching the key text after the structured processing according to the preset fields includes: Determining the similarity between the preset field and the key text after the structured processing by means of cosine similarity; If the similarity value is less than a preset threshold, it is determined that the classification information matches the key text after the structured processing and the preset field.

5. The method according to claim 1, wherein After filling the classification information into a preset web page and obtaining a target web page after the information is entered, the method further includes: Displaying the entered information on the target webpage so that the user can verify the entered information; If the user confirms that the entered information is correct, an entry log is generated.

6. The method according to claim 1, characterized in that The method further comprises: Obtaining a training data set and a pre-trained model; wherein the training data set includes historical image information; Annotating the training data set to obtain an annotation result, wherein the annotation result includes key text corresponding to the historical image information; The pre-training model is trained using the training data set and the annotation results to obtain the deep learning model.

7. The method according to claim 6, characterized in that After obtaining the deep learning model, the method further includes: Iteratively train the deep learning model based on the training data set and the annotation results to extract data features, and calculate the loss function; Iteratively training the loss function using a preset method for the purpose of reducing the loss function value until the loss function value is less than an expected threshold; Based on the loss function after iterative training, an iterative deep learning model is obtained.

8. An information entry device, characterized in that: include: An acquisition module is used to acquire image information to be identified and analyze the image information through a deep learning model to obtain key text in the image information; A structured processing module, configured to perform structured processing on the key text to obtain the structured key text; A classification module is used to classify and match the key text after the structured processing according to preset fields to obtain classified classification information; The information entry module is used to fill the classification information into a preset web page to obtain a target web page after the information is entered.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the steps of the information entry method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the information entry method according to any one of claims 1 to 7 are implemented.

Images