Integrated financial intelligent terminal receipt machine

Abstract

The invention provides an integrated financial intelligent terminal receipt machine, which comprises the following steps: an autonomous receipt delivery assembly receives reimbursement receipts, scans and identifies the reimbursement receipts, and extracts first key information of the reimbursement receipts; the information processing component extracts second key information of the reimbursement receipt, and processes the first key information and the second key information through a data processing engine; meanwhile, performing electronic filing on the processed first key information and second key information, automatically printing a corresponding approval sheet and filing an accounting voucher; and the functional interface component provides newly added login authentication, reimbursement receipt interface authentication and a synchronous reimbursement receipt state interface. According to the invention, full-process automation from reimbursement application receiving, information processing, data storage and state feedback is realized; the efficiency of the reimbursement process is improved, manual intervention is reduced, the error rate is reduced, and convenience is provided for information management.

Description

Technical Field

[0001] The present invention relates to the technical field of information processing terminals, and in particular to an integrated financial intelligent terminal acquiring machine. Background Art

[0002] With the development of information technology, reimbursement systems and acquiring machines have become indispensable tools in corporate operations. The existing technology includes a logistics-type financial reimbursement acquiring machine, which integrates scanning, identification, data transmission and other modules to achieve fast and accurate processing of reimbursement documents. The equipment is usually composed of a scanner, a processor, a display and a printer, and the workflow covers document scanning, information identification, data processing and reimbursement confirmation. However, in the existing technology, the reimbursement system and the acquiring machine operate independently and lack an effective data sharing and interaction mechanism. After the acquiring machine completes the delivery, the transaction data generated is not effectively connected with the reimbursement system. There are still problems such as cumbersome processes, unclear document rights and responsibilities, document rejection, and low reimbursement efficiency. It is impossible to achieve the intelligent goal of the entire process of fast automatic acquiring, scanning, review, return and archiving.

[0003] Prior art 1, application number: CN202310427650.3 discloses a financial reimbursement data processing system and method for an electric power company. The financial reimbursement data processing system is used for electric power reimbursement personnel to log in to the financial reimbursement data processing system through the company's intranet to perform financial online reimbursement. There is no need to reimburse the electric power financial personnel in person, which saves time for the electric power reimbursement personnel to perform financial reimbursement, greatly improving the efficiency of financial reimbursement. The electric power reimbursement personnel port is used for the electric power reimbursement personnel to log in to the financial reimbursement data processing system for financial reimbursement. The electric power financial personnel port is used for the electric power financial personnel to log in to the financial reimbursement data processing system to conduct an initial review of the financial reimbursement. The electric power leader port is used for the electric power leader to log in to the financial reimbursement data processing system to conduct a final review of the financial reimbursement. Although there is no need for a large amount of manual participation and the workload is small, the efficiency of financial reimbursement is greatly improved, which is conducive to rapid financial reimbursement. However, the connection relationship between the reimbursement system and the acquiring machine is not involved, resulting in a complex and cumbersome reimbursement process and low reimbursement efficiency.

[0004] Prior art 2, application number: CN202211535247.4 discloses a method and system for processing financial reimbursement forms, the method comprising: receiving the original form submitted by the reimbursement applicant, the original form containing the name of the reimbursement applicant, department information, and form ID; at the preset node of the form processing flow, writing the data in the original form into the financial intermediate node table corresponding to the node; after receiving the summary request submitted by the summary initiator, obtaining data from the financial intermediate node table to generate a financial summary table. Although it can improve the efficiency of reimbursement form summarization and meet the content requirements of financial summary reports for personnel at different levels. However, the reimbursement forms are still collected separately, resulting in unclear rights and responsibilities of documents and a high rejection rate of documents.

[0005] Prior art three, application number: CN 202210380038.0 discloses a method, device, equipment and medium for processing financial reimbursement. The financial reimbursement business system obtains the credit rating of the employee to be reimbursed corresponding to the identity of the employee to be reimbursed in the reimbursement request from the financial sharing center platform according to the reimbursement request; determines the financial reimbursement process corresponding to the credit rating of the employee to be reimbursed, and performs corresponding reimbursement processing under the corresponding financial reimbursement process according to the document information to be reimbursed in the reimbursement request; analyzes the reimbursement processing process of the employee to be reimbursed, obtains the indicators and corresponding scores of the employee to be reimbursed, and reports the indicators and corresponding scores of the employee to be reimbursed to the financial sharing center platform, so that the financial sharing center platform can update the corresponding credit rating of the employee to be reimbursed according to the indicators and corresponding scores of the employee to be reimbursed. Although it solves the problem that the determination of employee credit scores is not comprehensive, which may lead to uncontrollable reimbursement quality in the processing of reimbursement business. However, the reimbursement system is still separated from the document collecting machine, which provides limited help in standardizing the filling of reimbursement documents and the reimbursement process, and there is room for further improvement in improving reimbursement efficiency.

[0006] At present, the existing technologies 1, 2 and 3 have the problems of cumbersome reimbursement process operation, non-standard filling, high document rejection rate and low intelligence level of the whole process of document reimbursement. Therefore, the present invention provides an integrated financial intelligent terminal acquiring machine, which realizes the effective combination between the reimbursement system and the acquiring machine. Summary of the invention

[0007] In order to solve the above technical problems, the present invention provides an integrated financial intelligent terminal acquiring machine, comprising:

[0008] The autonomous document delivery component is responsible for receiving reimbursement documents, scanning and identifying the reimbursement documents, and extracting the first key information of the reimbursement documents; the first key information includes the amount, date and title;

[0009] The information processing component is responsible for extracting the second key information of the reimbursement document and processing the first key information and the second key information through the data processing engine. The processing includes automatic verification, classification and sorting. At the same time, the processed first key information and the second key information are electronically archived, and the corresponding approval form and accounting voucher are automatically printed for archiving;

[0010] The functional interface component is responsible for providing new login authentication, receiving reimbursement document interface authentication and synchronizing reimbursement document status interface; among them, the new login authentication includes Three Gorges Xingyun code scanning authentication, face recognition authentication and work card swiping authentication.

[0011] Optional, autonomous delivery component, including:

[0012] The document receiving module is responsible for receiving the completed reimbursement documents and placing them within the scanning area for temporary storage;

[0013] The image processing module is responsible for triggering the image capture process when the reimbursement document is placed in the scanning area and the visual sensor detects the existence of the reimbursement document; using the visual sensor to capture the image of the reimbursement document, performing preprocessing operations on the image, and obtaining the preprocessed image;

[0014] The result output module is responsible for identifying the size, direction, quality and first key information of the reimbursement document from the preprocessed image; at the same time, it provides a user interaction interface, and displays operation instructions and feedback information through a touch screen or display screen; among them, the preprocessing includes denoising and contrast enhancement.

[0015] Optional, result output module, including:

[0016] The image selection submodule is responsible for generating multiple candidate images for each automatic capture, and selecting an image from the multiple pre-processed candidate images as the main image for subsequent processing;

[0017] The physical characteristics submodule is responsible for measuring the width and height of the main image to determine the size of the reimbursement document; determining the orientation of its long and short sides to determine whether it is horizontal or vertical, and automatically adjusting the image direction according to the orientation; judging whether the main image meets the processing requirements according to the set quality standards; if it does, proceed to the next step; if it does not, prompt the user to re-place or re-scan the document;

[0018] The information extraction submodule is responsible for extracting the first key information from the main image that meets the processing requirements. Each time a key information is extracted, the result is immediately fed back to the user interface and the extraction result is displayed for user confirmation.

[0019] Optional, physical properties submodule, including:

[0020] The edge detection unit is responsible for converting the main image into a grayscale image after it is selected, binarizing the image using an adaptive Gaussian threshold, smoothing the image using a Gaussian filter, and performing edge detection using an edge detection algorithm based on the edge intensity gradient; identifying the outer border of the document in the image;

[0021] The size acquisition unit is responsible for applying polygonal approximation to the extracted contour, smoothing the contour and simplifying the edge, and detecting its shape features; calculating the minimum circumscribed rectangle of the contour, and performing geometry calculation by finding the extreme points of the contour to obtain the width and height of the image;

[0022] The orientation adjustment unit is responsible for determining the orientation of the file based on the ratio of width to height. If width > height, it is recorded as landscape; otherwise, it is recorded as portrait. Based on the orientation information, the image is automatically rotated 90 or 180 degrees, and the rotated image is resampled using bilinear interpolation or nearest neighbor interpolation algorithms.

[0023] Optional, edge detection unit, including:

[0024] The convolution operation subunit is responsible for edge detection of the main image after Gaussian smoothing. It uses two 3x3 convolution kernels to perform two convolution operations on the original image to obtain the gradient of the main image; calculate the gradient amplitude and direction of each pixel;

[0025] The gradient calculation subunit is responsible for comparing each pixel with its neighbors in the gradient direction according to the calculated gradient direction; if the magnitude of the pixel is greater than the magnitude of the neighboring pixels, the pixel is retained; otherwise, its intensity is set to zero;

[0026] The threshold comparison subunit is responsible for setting a high threshold and a low threshold, and judging the intensity of each pixel after non-maximum suppression; for pixels marked as strong edges, their connectivity is used to track the connected edges in the image, and any pixels connected to the strong edges are marked as edges to form a complete edge map.

[0027] Optional, information extraction submodule, including:

[0028] The feature merging unit is responsible for annotating the main images that meet the processing requirements, determining the location and format of the first key information; selecting a pre-trained model and fine-tuning it on the annotated main image containing the reimbursement document; constructing an image pyramid, generating multiple versions of input images from different scales, and extracting features from the input image at each scale; merging deep features from different scales; and converting the extracted features into fixed-length feature vectors through a fully connected layer;

[0029] The text acquisition unit is responsible for locating the key area using the bounding box and clarifying the relative position of the key area in the main image; using the OCR engine to perform character recognition on the extracted key area to obtain text information; and making rule judgments on the recognized text information to ensure the rules of the first key information;

[0030] The information verification unit is responsible for contextual analysis of the extracted text information. It determines that the amount is in a specific area and the invoice number is in the header by setting pattern matching rules. It uses the named entity recognition algorithm to analyze the text, and the extracted first key information has more semantic context.

[0031] Optional information processing components, including:

[0032] The content extraction module is responsible for extracting the second key information from the scanned image of the reimbursement document through the data processing engine OCR. The second key information includes departure time, arrival time, departure place, arrival place, passengers, seat class and ticket amount;

[0033] The engine classification module is responsible for automatically verifying, classifying and organizing the first key information and the second key information using an optical character recognition engine, an intelligent verification engine, a classification engine and a data sorting engine; at the same time, the user directly extracts the reimbursement document for processing. If the reimbursement document is rejected, the user can retrieve the document on his own, and replace it or submit the attachments according to the prompts, and then submit the reimbursement document again;

[0034] The archiving processing module is responsible for electronically archiving approved reimbursement documents, printing approval receipts, and archiving accounting vouchers.

[0035] Optional, engine classification module, including:

[0036] The engine setting submodule is responsible for setting up each engine independently. The reimbursement documents to be processed are sent to the input queue for asynchronous processing. Each engine has an independent work queue for receiving tasks to be processed, and the tasks to be processed are distributed to the corresponding engines through the message queue service.

[0037] The queue service submodule is responsible for quickly converting the relevant information of the reimbursement document into a processing task when the user submits the reimbursement document, and sending it to the input queue; the message queue service takes the task from the input queue and distributes it to each work queue; each engine independently obtains and processes tasks from the queue by monitoring its work queue; at the same time, each engine can run concurrently, waiting for the next task, and after processing, write the result into the result storage or send it to the next link;

[0038] The result summary submodule is responsible for summarizing the processing results of all engines in the data sorting engine, and performing overall data sorting, storage and return.

[0039] Optionally, the engine classification module also contains:

[0040] The data extraction submodule is responsible for using the OCR engine to perform optical character recognition on the scanned reimbursement document image, converting the pixel data in the image into machine-readable text data, including image preprocessing, feature extraction, and character recognition; the output is a text data structure;

[0041] The preliminary verification submodule is responsible for performing a structured preliminary verification of the text data extracted by the OCR engine, including verifying whether the text format conforms to the preset pattern, checking whether the required fields exist, and identifying characters or strings that cannot be accurately recognized;

[0042] The logic verification submodule is responsible for performing advanced logic verification on the data that has passed the preliminary verification, and cross-validating through predefined rules and business logic;

[0043] The exception handling submodule is responsible for automatically marking all abnormal data found during the logic verification phase as objects that need attention, generating an error report that lists every non-compliance item found in detail; at the same time, relevant personnel will receive notifications about the abnormal situation so that they can intervene and handle it in time;

[0044] The feedback and correction submodule is responsible for presenting documents marked as abnormal in the user interface. Users can access the error report and view the specific error information. A user-friendly interface will be provided to assist users in revising and supplementing abnormal data. After the correction is completed, users can submit the corrected documents back to the system for re-verification.

[0045] The final confirmation submodule is responsible for marking the document data that meets the requirements as valid after all verification steps are completed; at this time, the reviewed and confirmed document data will be stored in the database and marked as ready to enter the subsequent approval process.

[0046] Optional, functional components, including:

[0047] The status synchronization module is responsible for synchronizing the status of reimbursement documents to the acquiring system when the reimbursement document is rejected, withdrawn by the user, submitted for approval by the user, or archived. For rejected or withdrawn reimbursement documents, users can return them through the self-service interface, while for archived documents, physical archiving will be performed. In addition, for reimbursement documents in the submission for approval status, users are not allowed to return them on their own.

[0048] The recognition and transformation module is responsible for upgrading and transforming the reimbursement documents. At the same time, the transformation will realize the multi-invoice recognition function to support the parallel processing of different types of invoices.

[0049] The data access control module is responsible for calling the authentication interface to obtain an access token based on the clientID and clientSecret obtained from the acquiring system; the token must be included in the HTTP request header in subsequent requests;

[0050] The image processing conversion module is responsible for calling the corresponding interface to send all image data to the reimbursement system for optical character recognition and intelligent review after the acquiring machine completes the scanning of all reimbursement documents; according to the status recognition of the reimbursement document, it will automatically perform the printing operation of the specified reimbursement document and ensure that each reimbursement document can be physically printed one by one.

[0051] The autonomous document submission component of the present invention receives and processes the reimbursement documents submitted by the user, performs scanning and recognition functions, extracts the information of the reimbursement documents in digital form; automatically extracts and identifies the first key information (such as amount, date and header). Significance: It improves the efficiency of the reimbursement process, making the user's submission process easier and faster; ensures the accurate extraction of key data and reduces the risk of human input errors; realizes digital storage of information, and lays a good foundation for subsequent data processing. The information processing component extracts the second key information of the reimbursement document, including possible project categories, expense types, etc.; verifies, classifies and organizes the first key information and the second key information through the data processing engine; completes electronic archiving, automatically stores the processed information, and prints approval receipts and accounting vouchers. Significance: Through automated data processing, the processing efficiency and accuracy of reimbursement documents are improved, and the need for manual operations is reduced; a systematic archiving and review process is provided, which helps enterprises to better manage their financial affairs and generate reports; ensures the tracking and integration of all reimbursement data, and facilitates compliance audits. The functional interface component provides a login authentication interface for user identity authentication to ensure that only authorized users can access the system; it handles interface authentication for reimbursement documents to ensure data security and access compliance; it provides an interface for synchronizing the status of reimbursement documents to provide real-time feedback on the review progress of reimbursement. Significance: Through secure identity authentication and interface authentication, it protects user privacy and data security and prevents unauthorized access; the real-time updated status of reimbursement documents improves the user experience, allowing users to understand the progress of reimbursement at any time and increase transparency; it is conducive to integration with other systems, improving the consistency and efficiency of the overall process, and achieving greater business value.

[0052] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description, claims, and drawings.

[0053] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0055] Figure 1 This is a block diagram of the integrated financial intelligent terminal acquiring machine in Embodiment 1 of the present invention;

[0056] Figure 2 This is a block diagram of the autonomous delivery component in Example 2 of the present invention;

[0057] Figure 3 This is a block diagram of a result output module in Embodiment 3 of the present invention;

[0058] Figure 4 This is a block diagram of the physical characteristics submodule in Example 4 of the present invention;

[0059] Figure 5 This is a block diagram of an edge detection unit in Embodiment 5 of the present invention;

[0060] Figure 6 This is a block diagram of the information extraction submodule in Example 6 of the present invention;

[0061] Figure 7 This is a block diagram of an information processing component in Embodiment 7 of the present invention;

[0062] Figure 8 The engine classification module frame in Example 8 of the present invention Figure 1 ;

[0063] Fig. 9 The engine classification module frame in Embodiment 9 of the present invention Figure 2 ;

[0064] Fig.10 The engine classification module frame in embodiment 10 of the present invention Figure 3 ;

[0065] Fig.11 This is a block diagram of the functional interface components in Example 11 of the present invention. DETAILED DESCRIPTION

[0066] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0067] The terms used in the embodiments of the present application are only for the purpose of describing specific embodiments, and are not intended to limit the embodiments of the present application. The singular forms of "a", "said" and "the" used in the embodiments of the present application and the appended claims are also intended to include plural forms, unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used herein refers to and includes any or all possible combinations of one or more associated listed items.

[0068] When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application. On the contrary, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the attached claims. In the description of the present application, it should be understood that the terms "first", "second", "third", etc. are only used to distinguish similar objects, and do not have to be used to describe a specific order or sequence, nor can they be understood as indicating or implying relative importance. For those of ordinary skill in the art, the specific meanings of the above terms in the present application can be understood according to the specific circumstances.

[0069] Example 1: Figure 1 As shown, an embodiment of the present invention provides an integrated financial intelligent terminal acquiring machine, comprising:

[0070] The autonomous document delivery component is responsible for receiving reimbursement documents, scanning and identifying the reimbursement documents, and extracting the first key information of the reimbursement documents; the first key information includes the amount, date, and title, etc.;

[0071] The information processing component is responsible for extracting the second key information of the reimbursement document and processing the first key information and the second key information through the data processing engine. The processing includes automatic verification, classification and sorting. At the same time, the processed first key information and the second key information are electronically archived, and the corresponding approval form and accounting voucher are automatically printed for archiving;

[0072] The functional interface component is responsible for providing new login authentication, receiving reimbursement document interface authentication and synchronizing reimbursement document status interface; among them, the new login authentication includes Three Gorges Xingyun code scanning authentication, face recognition authentication and work card swiping authentication.

[0073] The working principle and beneficial effects of the above technical solution are as follows: the autonomous delivery component of this embodiment receives the reimbursement document, scans and identifies the reimbursement document, and extracts the first key information of the reimbursement document; wherein the first key information includes the amount, date and title, etc.; the information processing component extracts the second key information of the reimbursement document, and processes the first key information and the second key information through the data processing engine, and the processing includes automatic verification, classification and sorting; at the same time, the processed first key information and the second key information are electronically archived, and the corresponding approval form and accounting voucher are automatically printed for archiving; the function interface component provides new login authentication, receiving reimbursement document interface authentication and synchronization reimbursement document status interface. The autonomous delivery component of the above solution receives and processes the reimbursement document submitted by the user, performs scanning and identification functions, extracts the information of the reimbursement document in digital form; automatically extracts and identifies the first key information (such as amount, date and title). Significance: Improves the efficiency of the reimbursement process, making the user's submission process simpler and faster; ensures the accurate extraction of key data and reduces the risk of human input errors; realizes the digital storage of information, and lays a good foundation for subsequent data processing. The information processing component extracts the second key information of the reimbursement document, including possible project categories, expense types, etc.; verifies, classifies and organizes the first key information and the second key information through the data processing engine; completes electronic archiving, automatically stores the processed information, and prints approval documents and accounting vouchers. Significance: Through automated data processing, the processing efficiency and accuracy of reimbursement documents are improved, and the need for manual operations is reduced; a systematic archiving and review process is provided, which helps enterprises to better manage their finances and generate reports; and all reimbursement data is tracked and integrated, providing convenience for compliance audits. The functional interface component provides a login authentication interface for user identity authentication to ensure that only authorized users can access the system; processes interface authentication for reimbursement documents to ensure data security and access compliance; and provides an interface for synchronizing the status of reimbursement documents to provide real-time feedback on the review progress of reimbursement. Significance: Through secure identity authentication and interface authentication, user privacy and data security are protected to prevent unauthorized access; the real-time updated status of reimbursement documents improves the user experience, enables users to understand the progress of reimbursement at any time, and increases transparency; it is conducive to integration with other systems, improving the consistency and efficiency of the overall process, and achieving greater commercial value.

[0074] In summary, the entire integrated financial intelligent terminal acquiring machine of this embodiment realizes the automation of the entire process from receiving reimbursement applications, information processing, to data storage and status feedback through the collaborative work of the autonomous delivery component, information processing component and functional interface component. It not only improves the efficiency of the reimbursement process, but also reduces manual intervention and error rates, while providing convenience in information management. It can help enterprises achieve higher accuracy, efficiency and compliance in financial management, and improve overall operational efficiency.

[0075] This embodiment builds a data sharing mechanism and utilizes the microservice architecture. Service governance ensures smooth collaboration and communication between services, including service registration, discovery, configuration, and routing. Service deployment pursues automation and independence, and containerization technology is the key. Service fault tolerance ensures stability through strategies such as timeouts, retries, and fuses. Using image file delivery, the image information is encoded and converted into data that can be stored, transmitted, and processed in the computer system into the DAMO database. The data transmission is encapsulated in JSON format, and the interface uses the OAuth 2.0 authentication mechanism for single sign-on control of Sanxia Xingyun. In addition, the interface supports the HTTPS protocol and uses SSL encryption technology to ensure the security of data transmission, which satisfies the data security control of the backend management system of the acquiring system and realizes the effective combination of the reimbursement system and the acquiring machine.

[0076] Example 2: Figure 2 As shown, based on Example 1, the autonomous delivery component provided by the embodiment of the present invention includes:

[0077] The document receiving module is responsible for receiving the completed reimbursement documents and placing them within the scanning area for temporary storage;

[0078] The image processing module is responsible for triggering the image capture process when the reimbursement document is placed in the scanning area and the visual sensor detects the existence of the reimbursement document; using the visual sensor to capture the image of the reimbursement document, performing preprocessing operations on the image, and obtaining the preprocessed image;

[0079] The result output module is responsible for identifying the size, direction, quality and first key information of the reimbursement document from the preprocessed image; at the same time, it provides a user interaction interface, and displays operation instructions and feedback information through a touch screen or display screen; among them, preprocessing includes denoising and contrast enhancement.

[0080] The working principle and beneficial effects of the above technical solution are as follows: the document receiving module of this embodiment receives the completed reimbursement document and places it within the scanning area to temporarily store the reimbursement document; the image processing module uses a visual sensor to capture the image of the reimbursement document, performs preprocessing operations on the image, and obtains the preprocessed image; the result output module identifies the size, direction, quality and first key information of the reimbursement document from the preprocessed image; and provides a user interaction interface to display operation instructions and feedback information through a touch screen or display screen; wherein the preprocessing includes denoising and contrast enhancement. The document receiving module of the above solution ensures the physical management of the reimbursement document before scanning, and prevents the misplacement and loss of the document through a standardized receiving process. Significance: Provides a clear physical interface so that users can easily operate when submitting reimbursement documents, reducing the risk of misplacing or losing documents; through the temporary storage function, improves the efficiency of subsequent image processing, ensures that the required documents have been properly processed and placed ready before scanning; provides a stable and reliable document source for the subsequent processing module, and improves the stability of the overall system. The image processing module uses a visual sensor to capture images of reimbursement documents to ensure that the image quality is high enough for subsequent processing; the captured images are preprocessed, including operations such as denoising and contrast enhancement to obtain clear and easy-to-recognize preprocessed images. Significance: Through image processing, the image quality is improved to provide accurate and clear input for the recognition module, thereby improving the accuracy of subsequent data extraction; denoising and contrast enhancement can eliminate interference caused by ambient light and print quality, ensuring the stability and consistency of recognition; the effect of preprocessing directly affects the quality of subsequent data extraction and analysis, and is crucial. The result output module extracts the size, direction, quality and first key information (such as amount, date, and heading) of the reimbursement document from the preprocessed image; provides a user interaction interface, and can display operation instructions and feedback information through a touch screen or display, including confirmation of the scan results or suggestions for rescanning. Significance: The ability to extract key data directly affects the degree of automation and accuracy of the reimbursement process, reduces manual intervention and improves efficiency; improves user experience through the user interaction interface, allowing users to intuitively see operation prompts and feedback, and enhances the controllability of operations; provides a feedback mechanism, when the recognition results are unsatisfactory, users can respond quickly to improve subsequent operations and ensure the ultimate accuracy of the data.

[0081] In summary, the document receiving module, image processing module and result output module in the autonomous document delivery component of this embodiment cooperate with each other to achieve efficient and accurate reimbursement document processing. The document receiving module ensures the standardization of physical management, the image processing module improves the clarity of the image, and the result output module is responsible for extracting key information and interacting with the user. This modular design not only improves the functionality and user experience of the system, but also lays a solid foundation for the automation of the entire reimbursement process.

[0082] Example 3: Figure 3 As shown, based on Example 2, the result output module provided in this embodiment of the present invention includes:

[0083] The image selection submodule is responsible for generating multiple candidate images for each automatic capture, and selecting an image from the multiple pre-processed candidate images as the main image for subsequent processing;

[0084] The physical characteristics submodule is responsible for measuring the width and height of the main image to determine the size of the reimbursement document; determining the orientation of its long and short sides to determine whether it is horizontal or vertical, and automatically adjusting the image direction according to the orientation; judging whether the main image meets the processing requirements according to the set quality standards; if it does, proceed to the next step; if it does not, prompt the user to re-place or re-scan the document;

[0085] The information extraction submodule is responsible for extracting the first key information from the main image that meets the processing requirements. Each time a key information is extracted, the result is immediately fed back to the user interface and the extraction result is displayed for user confirmation.

[0086] The working principle and beneficial effects of the above technical solution are as follows: the image selection submodule of this embodiment generates multiple candidate images each time it automatically captures, and selects an image from the multiple candidate images after preprocessing as the main image for subsequent processing; the physical property submodule measures the width and height of the main image to determine the size of the reimbursement document; determines the orientation of its long and short sides, determines whether it is horizontal or vertical, and automatically adjusts the image direction according to the orientation; determines whether the main image meets the processing requirements according to the set quality standards; if it meets, proceed to the next step; if it does not meet, prompt the user to re-place or re-scan the document; the information extraction submodule extracts the first key information from the main image that meets the processing requirements, and each time a key information is extracted, the result is immediately fed back to the user interface, and the extraction result is displayed for the user to confirm. The image selection submodule of the above scheme generates multiple candidate images for image preprocessing, selects the main image with the best effect for subsequent processing; and automatically compares the images using multiple evaluation criteria (such as clarity and contrast) to ensure that the selected images can contain valid information. Significance: Improves the accuracy and reliability of information extraction, and ensures that subsequent operations are based on high-quality images by filtering inappropriate candidate images; reduces the incidence of errors, improves the automation and intelligence of the system, and reduces the need for manual intervention; provides a better user experience, because the information finally confirmed by the user comes from multiple screened candidate images, rather than randomly generated images. The physical characteristics submodule measures the width and height of the reimbursement document in real time based on the selected main image, calculates and determines its size; determines the orientation of the image, identifies the long and short sides, and automatically adjusts the image direction to ensure consistency; evaluates the quality of the image and determines whether it meets the processing requirements based on pre-set standards (such as brightness, contrast, and clarity). Significance: Ensures the accuracy of subsequent information processing, avoids information extraction errors caused by incorrect image orientation through clear size and direction judgment; through the quality assessment mechanism, can promptly identify images that do not meet the processing standards, and feedback poor images to users to prevent the spread of erroneous data, effectively reducing the risk of reimbursement processing; improves the system's intelligent recognition ability, with a higher degree of automation, reduces the user's operation steps, and makes the user experience more humane. The information extraction submodule extracts the first key information from the main images that meet the processing requirements, such as the reimbursement amount, date, header and other specific information; the extraction results are fed back to the user interface in real time, so that users can immediately see the results of the system processing, and confirm and correct them. Significance: Achieve efficient data extraction, help users quickly obtain the required key information, and improve the overall speed of reimbursement processing; through instant feedback, enhance user participation and transparency of operations, users can confirm the extraction results in real time, and increase trust in the system; through the role of the information extraction module, timely embed data into the system process, making the subsequent reimbursement approval and financial management processes smoother and more efficient.

[0087] In summary, in the result output module of this embodiment, the image selection submodule, the physical characteristics submodule and the information extraction submodule respectively achieve the purpose of accurately extracting the reimbursement document information through their respective technical effects. Through the collaborative work of each submodule, the entire module not only improves the intelligence and efficiency of the system, but also achieves significant improvements in user experience, data accuracy and processing speed, and realizes the best practices of automated reimbursement processing.

[0088] Example 4: Figure 4 As shown, based on Example 3, the physical property submodule provided in this embodiment of the present invention includes:

[0089] The edge detection unit is responsible for converting the main image into a grayscale image after it is selected, binarizing the image using an adaptive Gaussian threshold, smoothing the image using a Gaussian filter, and performing edge detection using an edge detection algorithm based on the edge intensity gradient; identifying the outer border of the document in the image;

[0090] The size acquisition unit is responsible for applying polygonal approximation to the extracted contour, smoothing the contour and simplifying the edge, and detecting its shape features; calculating the minimum circumscribed rectangle of the contour, and performing geometry calculation by finding the extreme points of the contour to obtain the width and height of the image;

[0091] The orientation adjustment unit is responsible for determining the orientation of the file based on the ratio of width to height. If width > height, it is recorded as landscape; otherwise, it is recorded as portrait. Based on the orientation information, the image is automatically rotated 90 or 180 degrees, and the rotated image is resampled using bilinear interpolation or nearest neighbor interpolation algorithms.

[0092] The working principle and beneficial effects of the above technical solution are as follows: when the main image is selected, the edge monitoring unit of this embodiment converts it into a grayscale image, and uses an adaptive Gaussian threshold to binarize the image; uses Gaussian filtering to smooth the image, and uses a more efficient edge detection algorithm to perform edge detection based on the edge intensity gradient; identifies the outer frame of the document in the image; the size acquisition unit applies polygonal approximation to the extracted contour, smoothes the contour and simplifies the edge, and detects its shape features; calculates the minimum circumscribed rectangle of the contour, and calculates the width and height of the image by finding the extreme points of the contour; the orientation adjustment unit determines the orientation of the document based on the ratio of width to height; if the width>height, it is recorded as horizontal; otherwise it is recorded as vertical; based on the orientation information, the image is automatically rotated 90 degrees or 180 degrees, and the rotated image is resampled by bilinear interpolation or nearest neighbor interpolation algorithm. The edge monitoring unit of the above scheme converts the main image into a grayscale image, reduces the data complexity, and lays the foundation for subsequent processing; based on the local pixel brightness characteristics, the image under different lighting conditions can accurately extract the edge and enhance the feature contrast; effectively remove the noise in the image while maintaining the clarity of the edge, improving the accuracy of subsequent edge detection; applying advanced edge detection algorithms, improves the accuracy of identifying the outer border of the document. Significance: Ensure that clear and accurate contour information is obtained in the subsequent processing stage, and improve the reliability of the entire processing process; through adaptive processing and optimization algorithms, the system can adapt to various environmental conditions and achieve stronger versatility, which is crucial for the lighting changes that may occur in actual use scenarios; reduce the risk of misidentification and ensure that the extracted edge information accurately reflects the true shape of the document. The size acquisition unit simplifies the extracted contour, reduces unnecessary complexity, and maintains the characteristics of the shape, making edge processing more efficient; through geometric calculations, find the extreme points of the contour, simplify the complex contour into a rectangle, and calculate its width and height. Significance: Ensure that the size of the reimbursement document can be accurately identified, thereby providing the necessary data support for subsequent processing steps; polygonal approximation reduces computing pressure, improves processing speed, and improves overall efficiency while ensuring data accuracy; through effective geometric calculations, accurately determine the actual size of the image, providing a more solid foundation for information extraction and subsequent data processing. The orientation adjustment unit automatically determines the orientation (horizontal or vertical) of the document by analyzing the ratio of width and height, and makes judgments based on geometric features; if incorrect orientation is detected, the system can automatically perform a 90-degree or 180-degree rotation, and resample through bilinear interpolation or nearest neighbor interpolation algorithms to restore image quality.Significance: Automated orientation recognition and adjustment minimizes user intervention, making system operation smoother and more efficient, and improving user experience; ensuring that subsequent information extraction and data processing are based on a consistent format, avoiding information errors and processing delays caused by image orientation problems; and ensuring that the rotated image will not be significantly distorted through high-quality interpolation technology, maintaining the clarity and accuracy of the extracted information.

[0093] In summary, in the physical property submodule of this embodiment, the synergistic effect of the edge monitoring unit, the size acquisition unit and the orientation adjustment unit enables the system to extract clear and accurate size and orientation information from the image. This not only enhances the efficiency and accuracy of data processing, but also improves the overall user experience, while playing an important role in intelligent and automated processing. The entire system can efficiently process reimbursement documents, ensuring accurate information extraction and subsequent smooth data processing.

[0094] Example 5: Figure 5 As shown, on the basis of embodiment 4, the edge detection unit provided in the embodiment of the present invention includes:

[0095] The convolution operation subunit is responsible for edge detection of the main image after Gaussian smoothing. It uses two 3x3 convolution kernels to perform two convolution operations on the original image to obtain the gradient of the main image; calculate the gradient amplitude and direction of each pixel;

[0096] The gradient calculation subunit is responsible for comparing each pixel with its neighbors in the gradient direction according to the calculated gradient direction; if the magnitude of the pixel is greater than the magnitude of the neighboring pixels, the pixel is retained; otherwise, its intensity is set to zero;

[0097] The threshold comparison subunit is responsible for setting a high threshold and a low threshold, and judging the intensity of each pixel after non-maximum suppression; for pixels marked as strong edges, their connectivity is used to track the connected edges in the image, and any pixels connected to the strong edges are marked as edges to form a complete edge map;

[0098] If the intensity > high threshold, it is marked as a strong edge;

[0099] If the intensity < low threshold, mark it as non-edge;

[0100] If the intensity is between the low and high thresholds, the edge is determined based on connectivity, and pixels connected to strong edges are marked as edges.

[0101] The working principle and beneficial effects of the above technical solution are as follows: the convolution operation subunit of this embodiment performs edge detection on the main image after Gaussian smoothing, uses two 3x3 convolution kernels, performs two convolution operations on the original image, and obtains the gradient of the main image; calculates the gradient amplitude and direction of each pixel; the gradient calculation subunit compares each pixel with its neighborhood in the gradient direction according to the calculated gradient direction; if the amplitude of the pixel is greater than the amplitude of the neighboring pixel, the pixel is retained; otherwise, its intensity is set to zero; the threshold comparison subunit sets a high threshold and a low threshold, and judges the intensity of each pixel after non-maximum suppression; for pixels marked as strong edges, use their connectivity to track the connected edges in the image, and any pixels connected to the strong edges are marked as edges to form a complete edge map; if the intensity is greater than the high threshold, it is marked as a strong edge; if the intensity is less than the low threshold, it is marked as a non-edge; if the intensity is between the low threshold and the high threshold, it is determined whether it is an edge based on the connectivity, and the pixels connected to the strong edge are marked as edges. The convolution operation subunit of the above scheme applies two 3x3 convolution kernels to the Gaussian smoothed image, and the convolution operation can extract the edge information of the image. Specifically, the convolution kernel focuses on the gradients in the horizontal and vertical directions respectively; after the convolution, the horizontal and vertical gradients obtained are used to calculate the gradient amplitude and direction of each pixel. Significance: Converting image information into gradient information allows subsequent processing to focus on the parts of the image that change strongly (i.e., edges), providing important basic data for subsequent non-maximum suppression and edge connection; by calculating the gradient of each pixel, the intensity and direction of different features in the image can be effectively obtained, thereby providing an accurate basis for subsequent edge detection. The gradient calculation subunit compares the amplitude of each pixel with the amplitude of its neighboring pixels (adjacent pixels along the gradient direction) according to the calculated gradient direction. If the current pixel is a local maximum, the intensity of the pixel is retained; otherwise, its intensity is set to zero; through non-maximum suppression, the edge width is reduced, ensuring that only pixels that truly represent the edge are retained, and the edge performance is refined. Significance: By eliminating non-edge pixels, the accuracy and precision of edge detection are significantly improved, and false positives are reduced; only local maxima are retained, which helps to form clear and detailed edge lines; the contours of each edge in the final edge map are accurate and clear, providing a high-quality data foundation for subsequent processing links (such as thresholding and edge connectivity analysis). The threshold comparison subunit sets high and low thresholds to classify pixels and distinguish between strong edges, non-edges, and weak edges (pixels in between); by analyzing the connectivity between strong edges and weak edges, weak edge pixels connected to strong edges are marked as edges, thus forming a complete edge connection.Significance: It ensures that in the final output edge image, only pixels connected to strong edges are identified as edges, which can effectively eliminate isolated noise points and improve the integrity and accuracy of the image; by utilizing the connectivity method, local edges (weak edges) are allowed to form effective connections with strong edges, thereby ensuring the continuity of important features during image segmentation or subsequent recognition, greatly enhancing the edge detection quality.

[0102] In summary, the various subunits of the edge detection unit of this embodiment work together to achieve effective conversion from image to edge information. The convolution operation subunit extracts edge features and calculates gradients through edge detection; the gradient calculation subunit refines the edge to ensure that the true edge pixels are retained; and the threshold comparison subunit further enhances the edge continuity and overall quality of the image. The edge detection process not only improves the accuracy and clarity of edge recognition, but also lays a solid foundation for subsequent image processing and information extraction.

[0103] Example 6: Figure 6 As shown, based on Example 3, the information extraction submodule provided by the embodiment of the present invention includes:

[0104] The feature merging unit is responsible for annotating the main images that meet the processing requirements, determining the location and format of the first key information; selecting a pre-trained model and fine-tuning it on the annotated main image containing the reimbursement document; constructing an image pyramid, generating multiple versions of input images from different scales, and extracting features from the input image at each scale; merging deep features from different scales; and converting the extracted features into fixed-length feature vectors through a fully connected layer;

[0105] The text acquisition unit is responsible for locating the key area using the bounding box and clarifying the relative position of the key area in the main image; using the OCR engine to perform character recognition on the extracted key area to obtain text information; and making rule judgments on the recognized text information to ensure the rules of the first key information;

[0106] Date format: For example, check whether it is in the format of YYYY-MM-DD or MM / DD / YYYY.

[0107] Amount format: Checks whether the amount contains a currency symbol (such as $ or ￥) and appropriate decimal places (such as 123.45);

[0108] Invoice number format: Checks the specific structure of the invoice number (such as the length of the combined letters and numbers);

[0109] The information verification unit is responsible for contextual analysis of the extracted text information. It determines that the amount is in a specific area and the invoice number is in the header by setting pattern matching rules. It uses the named entity recognition algorithm to analyze the text, and the extracted first key information has more semantic context.

[0110] The working principle and beneficial effects of the above technical solution are as follows: the feature merging unit of this embodiment annotates the main image that meets the processing requirements, determines the position and format of the first key information; selects a pre-trained model and performs fine-tuning on the annotated main image containing the reimbursement document; constructs an image pyramid, generates multiple versions of input images from different scales, and extracts features from the input image of each scale; merges the deep features from different scales; converts the extracted features into a feature vector of fixed length through a fully connected layer; the text acquisition unit uses a bounding box to locate the key area and clarifies the relative position of the key area in the main image; uses an OCR engine to perform text extraction on the extracted key area Character recognition to obtain text information; rule judgment is performed on the recognized text information to ensure the rules of the first key information; date format: for example, check whether it is in the format of YYYY-MM-DD or MM / DD / YYYY; amount format: check whether the amount contains currency symbols (such as $ or ￥) and appropriate decimal places (such as 123.45); invoice number format: check the specific structure of the invoice number (such as the length of the combined letters and numbers); the information verification unit performs context analysis on the extracted text information, and determines that the amount is in a specific area and the invoice number is in the head by setting pattern matching rules; use the named entity recognition algorithm to analyze the text, and the extracted first key information has more semantic context. The feature merging unit of the above scheme annotates the main image and identifies the location and format of key information; selects a pre-trained model for fine-tuning to adapt to specific types of reimbursement documents; constructs an image pyramid to generate input images of different scales to ensure that it can handle texts of different sizes and inclinations; merges features extracted at different scales to form a comprehensive feature vector; through annotation, the location of key information (such as invoice number, amount and date) in the image can be clearly determined to improve the accuracy of subsequent extraction; fine-tunes the pre-trained model to adapt it to specific tasks and improve the effect and accuracy of feature extraction; constructs an image pyramid to obtain multi-scale features, enhances the model's robustness to images, and can effectively recognize texts of different sizes and directions; merges multi-level features into a unified feature vector to improve the expression ability of the recognition model. Significance: It lays a good foundation for subsequent text extraction, realizes efficient and accurate information extraction on specific types of reimbursement documents, and plays a great role in promoting the instillation of automation and intelligent business processes. The text acquisition unit can accurately locate key information through bounding box analysis, making information extraction more targeted; using OCR technology, the text in the image can be quickly converted into machine-readable text format to improve extraction efficiency; the extracted text is checked for rules to ensure that it meets certain format standards (such as date, amount, etc.). Significance: It ensures the accuracy and effectiveness of information extraction, thereby reducing the burden of data cleaning and processing in the later stage, and provides reliable data input in the automated reimbursement process.Through context analysis, the information verification unit can more intelligently understand the relationship between information and improve the accuracy of extraction; by verifying the extracted information with the expected format and position relationship, it can effectively eliminate erroneous information and context inconsistencies; through NER, it can extract more meaningful context data, enrich reimbursement information, and improve the integrity of information. Significance: Further improve the accuracy and intelligence of extracted information, provide higher quality basic data for subsequent data processing and decision-making, and enhance the transparency and credibility of the automated reimbursement system.

[0111] The method for obtaining the pre-trained model in this embodiment is trained on a large-scale data set (such as ImageNet) and has good feature extraction capabilities. The following are the steps to obtain the pre-trained model: select a variety of commonly used deep learning architectures such as ResNet, EfficientNet, VGG, MobileNet, etc.; use the standard library in the popular deep learning framework (such as PyTorch or TensorFlow) to download the pre-trained model directly from the model library; further train on the labeled data set containing reimbursement documents. The last few layers of the model (such as the fully connected layer) need to be modified to adapt to specific tasks, such as outputting different categories. Training steps: prepare a data loader to input training data into the model; set the loss function and optimizer; start the training process and update the model parameters through backpropagation and optimization algorithms.

[0112] The steps to construct an image pyramid are as follows: suppose there is an input image, first know its original size (width and height); scale (usually between 0.5 and 1.0); for each scaling ratio, use the image scaling function to generate the corresponding image; for each pyramid level image of different sizes are extracted to generate the corresponding feature representation; each scale feature will be input into the subsequent network layer.

[0113] The merging of multi-scale features can enhance the comprehensive feature representation capability of the model. In a fine-tuned deep learning model, a feature vector is generated for each scale of image input. Features can be obtained through the output results of the last few convolutional layers; at the same time, feature extraction intermediate layers can be designed in the model to obtain feature outputs; concatenation: concatenate the feature vectors from each scale by column (or row) to form a larger feature vector; finally, input the merged feature vector into the fully connected layer to generate a fixed-length output feature vector, which provides input for subsequent classification or regression tasks;

[0114] In summary, through the collaboration of the three units of feature merging, text acquisition and information verification, the information extraction submodule of this embodiment can achieve highly accurate, fast and intelligent extraction of key information from reimbursement documents with complex structures, reduce manual intervention and improve the automation of the process. It reduces labor costs, reduces error rates, and improves data quality, providing timely and reliable data support for the financial management and business flow of enterprises or organizations.

[0115] Example 7: Figure 7 As shown, based on Example 1, the information processing component provided by the embodiment of the present invention includes:

[0116] The content extraction module is responsible for extracting the second key information from the scanned image of the reimbursement document through the data processing engine OCR. The second key information includes departure time, arrival time, departure place, arrival place, passengers, seat class and ticket amount, etc.;

[0117] The engine classification module is responsible for automatically verifying, classifying and organizing the first key information and the second key information using an optical character recognition engine, an intelligent verification engine, a classification engine and a data sorting engine; at the same time, the user directly extracts the reimbursement document for processing. If the reimbursement document is rejected, the user can retrieve the document on his own, and replace it or submit the attachments according to the prompts, and then submit the reimbursement document again;

[0118] The archiving processing module is responsible for electronically archiving approved reimbursement documents, printing approval receipts, and archiving accounting vouchers.

[0119] The working principle and beneficial effects of the above technical solution are as follows: the content extraction module of this embodiment extracts the second key information from the scanned image of the reimbursement document through the data processing engine OCR, and the second key information includes departure time, arrival time, departure place, arrival place, passengers, seat class and ticket amount, etc.; the engine classification module adopts optical character recognition engine, intelligent verification engine, classification engine and data sorting engine to automatically verify, classify and sort the first key information and the second key information; at the same time, the user directly extracts the reimbursement document for processing. If the reimbursement document is rejected, the user takes the order on his own, and replaces or supplements the attachments according to the prompts, and then submits the reimbursement document again; the filing processing module electronically archives the approved reimbursement documents, prints the approval face sheet, and archives the accounting vouchers. The content extraction module of the above solution can quickly identify and extract key information from the scanned image through OCR technology, reducing the time of manual input and review; combined with OCR and pre-trained model for fine-tuning, it can handle various fonts, formats and layouts, thereby improving recognition accuracy; it can handle a large number of reimbursement documents and improve overall approval and management efficiency. Significance: It significantly reduces the manual workload and speeds up information processing, allowing enterprises or organizations to manage expense reimbursement more efficiently; because it can automatically process more complex expense reimbursement information, it reduces the error rate and improves the quality and reliability of data. The engine classification module uses an intelligent verification engine to perform rule and logic verification on the extracted information to ensure the validity and consistency of the information; the classification engine classifies the information according to predetermined standards to facilitate subsequent processing and archiving; it provides users with the function of directly extracting and processing reimbursement documents, and dynamically optimizes through a feedback mechanism (such as rejection reason prompts). Users can make changes or submit attachments based on system feedback. Significance: It improves the flexibility of information processing and can quickly respond to different types of reimbursement documents; it enhances the transparency and user experience of the reimbursement process, allowing users to understand the processing status in a timely manner and improve overall satisfaction. The archiving processing module stores the reimbursement documents that have been approved in compliance in the electronic system, making information query and management more convenient and avoiding the loss and loss of paper documents; providing paper support for approved reimbursement documents and vouchers to facilitate subsequent financial audits and inspections; systematizing all approved reimbursement information to further improve the traceability and manageability of data. Significance: Through electronic means, it enhances the security and efficiency of data management, reduces storage costs and space requirements, and improves the speed of information retrieval; provides support for the compliance and audit of financial reimbursement, and reduces the risks caused by manual management.

[0120] In this embodiment, the OCR (optical character recognition) engine is responsible for converting the text information in the scanned or photographed document image into editable and searchable digital text; it supports the recognition of multiple fonts, layouts and formats to ensure the accurate extraction of information. Intelligent verification engine: automatically verifies the data extracted by the OCR engine to verify its accuracy, completeness and compliance; it may include numerical range verification (such as whether the amount is within a reasonable range), logical verification (such as whether the date is valid, whether the context is consistent), etc. Classification engine: automatically classifies documents according to preset rules or machine learning algorithms; the classification basis may include document type (such as travel expenses, office supplies expenses), department, nature of expense, etc. Data sorting engine: formats and standardizes the verified and classified data to ensure data consistency and readability; it may include removing redundant information, correcting erroneous data, and generating standard reports.

[0121] In summary, the entire information processing component of this embodiment integrates multiple functional modules to form an automated and intelligent reimbursement document processing system. It helps to reduce processing time and labor costs; improve the transparency of the reimbursement process and the self-service capabilities of users; enhance information accuracy and reduce errors, improve data quality; improve the convenience and security of data management through electronic archiving, and promote internal management efficiency of the enterprise; through these automated and intelligent processing methods, organizations can achieve more efficient financial management, improve operational efficiency, and ensure compliance and audit requirements.

[0122] Example 8: Figure 8 As shown, based on Example 7, the engine classification module provided by the embodiment of the present invention includes:

[0123] The engine setting submodule is responsible for setting up each engine independently. The reimbursement documents to be processed are sent to the input queue for asynchronous processing. Each engine has an independent work queue for receiving tasks to be processed, and the tasks to be processed are distributed to the corresponding engines through the message queue service.

[0124] The queue service submodule is responsible for quickly converting the relevant information of the reimbursement document into a processing task when the user submits the reimbursement document, and sending it to the input queue; the message queue service takes the task from the input queue and distributes it to each work queue; each engine independently obtains and processes tasks from the queue by monitoring its work queue; at the same time, each engine can run concurrently, waiting for the next task, and after processing, write the result into the result storage or send it to the next link;

[0125] The result summary submodule is responsible for summarizing the processing results of all engines in the data sorting engine, and performing overall data sorting, storage and return.

[0126] The working principle and beneficial effects of the above technical solution are as follows: the engine setting submodule of this embodiment sets up each engine independently, and the reimbursement documents to be processed are sent to the input queue for asynchronous processing; each engine has an independent work queue for receiving tasks to be processed, and distributes the tasks to be processed to the corresponding engine through the message queue service; when the user of the queue service submodule submits a reimbursement document, the relevant information of the reimbursement document is quickly converted into a processing task and sent to the input queue; the message queue service takes the task out of the input queue and distributes the task to each work queue; each engine independently obtains and processes tasks from the queue by listening to its work queue; at the same time, each engine can run concurrently, waiting for the next task, and after processing, the result is written into the result storage or sent to the next link; the result summary submodule summarizes the processing results of all engines in the data sorting engine, and performs overall data sorting, storage and return. Each engine in the engine setting submodule of the above scheme can be independently set according to specific needs, such as adjusting the recognition parameters of OCR, the rules of intelligent verification, etc.; providing an interface or API so that users can quickly adapt to new processing requirements (such as reimbursement documents in different formats); the reimbursement documents to be processed will be sent to the input queue to ensure that the distribution and processing of tasks can be carried out asynchronously, thereby improving overall efficiency. Significance: Improve the modularity and flexibility of the system, so that each engine can make adjustments to different business needs, thereby improving processing efficiency; reduce the system's dependence on a single engine, increase the fault tolerance of processing, and if a problem occurs in one engine, other engines can continue to work normally. The queue service submodule converts reimbursement documents into processing tasks and uses queues (such as message queues) to achieve asynchronous processing to avoid long waiting times; ensure that multiple engines can concurrently process multiple tasks at the same time, thereby improving the overall system throughput; and effectively handle the distribution, retry and failure management of tasks through message queue services. Significance: Significantly improve the response speed of the system, so that users will not feel delays when submitting reimbursement forms, which improves user experience; reduce system load, better manage task submission during peak hours, and reduce the pressure of any single module on the entire system through decoupling. The result summary submodule integrates the processing results of different engines to ensure the consistency and integrity of the data; after the data processing is completed, the results are returned to the user or downstream system, providing necessary feedback information (such as processing results, error information, etc.); effectively manage and store the processing results of all engines for subsequent analysis and auditing. Significance: It improves the transparency of data processing, allowing users to clearly understand the processing results, thereby enhancing trust; it provides basic data support for subsequent data analysis, report generation and decision-making, and improves the overall management efficiency of financial reimbursement.

[0127] In summary, the engine classification module of this embodiment embodies the following significance by dividing the reimbursement document processing flow into several independent but mutually cooperative sub-modules: modular design allows the system to be expanded or integrated with new functions more easily in the future; through asynchronous processing and independent engine settings, even if some components have problems, the overall system can still operate normally; through rapid task conversion and effective feedback mechanism, users will experience a smoother operating experience during use; the parallel processing capabilities of each engine enable the system to efficiently utilize computing resources and improve overall processing capabilities; the result summary sub-module ensures the orderly management of all processing results, facilitating data retrieval and subsequent analysis. The entire engine classification module not only improves the processing efficiency and accuracy of reimbursement documents, but also lays a technical foundation for future business expansion and technology upgrades.

[0128] Example 9: Fig. 9 As shown, based on Example 7, the engine classification module provided by the embodiment of the present invention further includes:

[0129] The data extraction submodule is responsible for using the OCR engine to perform optical character recognition on the scanned reimbursement document image and convert the pixel data in the image into machine-readable text data, including image preprocessing (such as denoising and binarization), feature extraction (recognizing the shape and outline of characters), and character recognition (using the trained model to predict the character category); the output is a text data structure, which contains all the extracted recognizable characters in the document and their corresponding position and format information;

[0130] The preliminary verification submodule is responsible for performing a structured preliminary verification of the text data extracted by the OCR engine, including verifying whether the text format conforms to the preset pattern (such as the date format "YYYY-MM-DD" and the amount format "#,###.00"), checking whether the required fields exist (such as the reimbursement person, reimbursement amount, etc.), and identifying characters or strings that cannot be accurately recognized;

[0131] The logic verification submodule is responsible for performing advanced logic verification on data that has passed the preliminary verification, and cross-validating through predefined rules and business logic; it includes:

[0132] Verify whether the reimbursement amount exceeds the set budget limit, such as comparing the reimbursement amount with the corresponding data field in the budget database;

[0133] Verify that the reimbursement date is within the set range, for example by comparing the current date with the report date to confirm its rationality;

[0134] Verify the matching between the reimbursement person and his / her department, for example, by comparing the reimbursement person information with the department field in the employee information database for consistency check; the goal of logical verification is to ensure the business rationality of the relationship between the data and the content;

[0135] The exception handling submodule is responsible for automatically marking all abnormal data found in the logic verification phase as objects that need attention, and generating an error report that lists in detail every non-compliant item found, including the specific field name, original data, error type, and recommended correction direction; at the same time, relevant personnel will receive notifications about abnormal conditions so that they can intervene in time;

[0136] The feedback and correction submodule is responsible for presenting documents marked as abnormal in the user interface. Users can access the error report and view the specific error information. A user-friendly interface will be provided to assist users in revising and supplementing abnormal data, including text boxes for entering missing fields, drop-down menus for selecting the correct category, and prompt information to guide users to make compliance processing. After the correction is completed, the user can submit the corrected document back to the system for re-verification.

[0137] The final confirmation submodule is responsible for marking the document data that meets the requirements as valid after all verification steps are completed. At this time, the document data that has been reviewed and confirmed will be stored in the database and marked as ready to enter the subsequent approval process. It includes updating status, archiving data, and generating necessary transaction logs for auditing.

[0138] The working principle and beneficial effects of the above technical solution are as follows: the data extraction submodule of this embodiment uses the OCR engine to perform optical character recognition on the scanned reimbursement document image, and converts the pixel data in the image into machine-readable text data, including image preprocessing (such as denoising, binarization), feature extraction (recognizing the shape and outline of characters), and character recognition (using the trained model to predict the character category); the output is a text data structure, which contains all the extracted recognizable characters in the document and their corresponding position and format information; the preliminary verification submodule performs a structured preliminary verification on the text data extracted by the OCR engine, including verifying whether the text format conforms to the standard. The module complies with the preset mode (such as the date format "YYYY-MM-DD" and the amount format "#,###.00"), checks whether the required fields exist (such as the person making the reimbursement, the reimbursement amount, etc.), and identifies the characters or strings that cannot be accurately identified; the logic verification submodule performs advanced logic verification on the data that has passed the preliminary verification, and cross-validates through predefined rules and business logic; including: verifying whether the reimbursement amount exceeds the set budget limit, such as comparing the reimbursement amount with the corresponding data field in the budget database; verifying whether the reimbursement date is within the set range, such as by comparing the current date and the report date to confirm its rationality; verifying the reimbursement person and his department For example, by comparing the reimbursement person information with the department fields in the employee information database for consistency check; the goal of logical verification is to ensure the business rationality of the relationship between data and content; all abnormal data found by the exception handling submodule during the logical verification stage will be automatically marked as objects that need attention by the system, and an error report will be generated, which lists in detail each non-compliant item found, including the specific field name, original data, error type and recommended correction direction; at the same time, relevant personnel will receive notifications about abnormal conditions so that they can intervene in time; the documents marked as abnormal in the feedback and correction submodule will be presented in the user interface, and users can The user can access the error report and view the specific error information; a user-friendly interface will be provided to assist the user in revising and supplementing the abnormal data, including text boxes for entering missing fields, drop-down menus for selecting the correct category, and prompt information to guide the user to make compliance processing; after the correction is completed, the user can submit the corrected document back to the system for re-verification; after all the verification steps of the final confirmation submodule are completed, the document data that meets the requirements will be marked as valid; at this time, the document data that has been reviewed and confirmed will be stored in the database and marked as ready to enter the subsequent approval process; including updating status, archiving data and generating necessary transaction logs for auditing. The data extraction submodule of the above scheme uses optical character recognition (OCR) technology to accurately convert pixel data in the image into processable text data; apply denoising and binarization operations to improve recognition accuracy and reduce errors and noise interference; use the trained model to recognize the shape and outline of characters to ensure the high accuracy of the extraction results.Significance achieved: Improve the automation and accuracy of data extraction, reduce manual intervention, and improve processing efficiency; lay a reliable data foundation for subsequent verification and logical processing. The preliminary verification submodule quickly detects the format legitimacy of text data (such as date format, amount format) through regular expressions or pattern matching; automatically checks the existence of required fields to ensure that key data is not missed; accurately identifies characters that cannot be accurately converted in OCR extraction, and cleans information before entering the subsequent processing link. Significance achieved: Through efficient preliminary verification, ensure the quality of data entering the subsequent steps and reduce potential errors in subsequent processing; improve the overall stability and reliability of the entire system and prevent erroneous data from flowing into the subsequent processing process. The logical verification submodule cross-validates the logical relationship of the data set according to the predefined business logic rules; checks the logical consistency of the data by comparing the customer database, budget database and employee information; ensures that the reimbursement amount, date, and personnel information are in line with the rationality of business operations. Significance achieved: Ensure the logical rationality between data, reduce the burden of manual intervention and review, and improve the reliability of data processing; through logical review, risk control can be achieved, unreasonable reimbursement requests can be prevented, and the financial compliance of the enterprise can be improved. The exception handling submodule system will automatically mark the exceptions found in the logic check and generate a detailed error report; relevant personnel can receive notifications of abnormal status in time so that they can take quick action to solve the feedback errors. Significance achieved: The response time to abnormal situations has been significantly improved, so that problems can be quickly identified and handled, which helps to maintain the smoothness of business processes; improve the transparency and traceability of reimbursement review, and ensure that all operations are documented for future audits. The feedback and correction submodule provides users with a friendly interface, which can easily view and correct marked abnormal data; through text boxes, drop-down menus and prompts, help users to effectively complete the correction. Significance achieved: Promote user participation, improve user satisfaction with reimbursement document correction and system acceptance; improve data accuracy through feedback and correction links, and ensure that the final data meets business requirements. The final confirmation submodule confirms the validity of all document data that have passed all checks and stores them in the database; records the status changes in each exception handling and data confirmation process, which is helpful for auditing and problem tracking. Significance achieved: Ensure the integrity and accuracy of data so that no data errors will occur in the subsequent approval process; improve approval efficiency and ensure that each document can be processed in a timely manner through clear process control.

[0139] In summary, through the collaborative work of the above submodules, the engine classification module in this embodiment can realize efficient, accurate and automatic processing of reimbursement documents. Each submodule not only improves the processing speed and quality, but also increases the stability and controllability of the system, provides the financial department with an efficient reimbursement processing tool, and promotes the standardization and informatization of corporate financial management.

[0140] Example 10: Fig.10 As shown, based on Example 7, the engine classification module provided by the embodiment of the present invention further includes:

[0141] The information parsing submodule is responsible for structuring the first key information and the key information of the extracted reimbursement document; performing semantic analysis on the identified text information and extracting specific key information using natural language processing technology; including:

[0142] Text segmentation: divide the overall text information into segments according to delimiters or format rules, and identify each field;

[0143] Entity recognition: Perform named entity recognition (NER) on the text through the trained natural language processing model to determine the semantic category of each field, such as extracting "department" as the entity category and identifying the corresponding department name from the text;

[0144] Data standardization: Standardize the format of the extracted text information, including unifying the case of strings, removing extra spaces, converting to a preset encoding format, etc.

[0145] The rule matching submodule is responsible for classifying the parsed information based on predefined classification logic and business rules, including:

[0146] Rule set establishment: pre-configure classification rule sets, including different expense types, corresponding reimbursement processes and conditions;

[0147] Condition evaluation: For each document information, the analysis results are compared with the rule set members one by one, and its category is determined by conditional judgment (such as "if the expense type is 'travel expense', then enter the travel reimbursement process");

[0148] Dynamic allocation: Responsively allocate document information that meets the specified conditions to the corresponding reimbursement process or category, and record the details of each decision for subsequent tracking and auditing;

[0149] The image storage submodule is responsible for storing the classified reimbursement document images in the preset storage location, including the following operations:

[0150] File organization: Organize image files by category and date, and create a corresponding directory structure to ensure efficient and convenient access;

[0151] Data encryption and backup: During the storage process, ensure the security of document image files, including encrypting the stored data and backing up them regularly to prevent data loss or leakage;

[0152] Metadata attachment: When storing image files, relevant metadata (such as classification information, storage time, processing status, etc.) are attached to provide necessary information support for subsequent query and review.

[0153] The working principle and beneficial effects of the above technical solution are as follows: the information parsing submodule of this embodiment structures the first key information and the key information of the extracted reimbursement document; performs semantic analysis on the identified text information, and extracts specific key information using natural language processing technology; the rule matching submodule classifies the parsed information based on the predefined classification logic and business rules; the image storage submodule stores the classified reimbursement document image in a preset storage location. The information parsing submodule of the above scheme accurately divides the text information in the document into various fields to ensure that the information is structured and convenient for subsequent processing and analysis; uses the trained natural language processing model to accurately identify important entities in the text (such as departments, expense types), ensuring that key data is extracted and identified; by unifying the format and cleaning redundant data, the consistency and readability of the information are improved, ensuring that errors will not be caused by inconsistent formats during subsequent processing. The significance achieved: It provides basic support for the overall information processing, facilitates the implementation of subsequent classification and rule matching through structured data; reduces the risk of errors in manual operations, improves the overall efficiency and accuracy of information processing, and improves data quality, laying a good foundation for subsequent processes. The rule matching submodule system configures classification rules in advance to ensure that the classification process meets actual business needs, making the classification highly stable and flexible; by comparing the parsed information and rule sets item by item, the system can flexibly judge and accurately classify each reimbursement request, reducing the complexity of human judgment; automatically assigning the document information that meets the conditions to the specified reimbursement process or category, and recording relevant decisions to ensure the traceability of classification. Significance achieved: Improve the automation level of reimbursement processing, reduce manual intervention, and improve work efficiency; ensure that each reimbursement document can be circulated according to actual needs, improve the compliance and transparency of operations, and support subsequent audit operations to enhance the overall security of data management. The image storage submodule organizes image files by category and date to create an effective file storage structure to ensure the efficiency and convenience of information retrieval; implement security control on data during the storage process to prevent unauthorized access and data loss, and ensure the long-term security of data through backup; attach key information (such as classification, timestamp, processing status, etc.) to each image file to provide necessary context information for subsequent query and review. Significance achieved: By strengthening data security and organization, effective management of reimbursement documents can be achieved, thereby reducing potential data risks; through systematic and structured image storage, rapid acquisition and review can be supported, work efficiency can be improved, and audit and compliance requirements can be better met.

[0154] In summary, through the synergy of the above submodules, the engine classification module of this embodiment aims to improve the automation and intelligence level of reimbursement document processing while ensuring the accuracy and security of information. This not only improves work efficiency and reduces labor costs, but also helps to achieve the company's requirements for financial management compliance and transparency. In summary, the realization of the above has important practical significance for improving internal management efficiency, supporting decision-making, and reducing corporate economic risks.

[0155] Example 11: Fig.11 As shown, based on Example 1, the functional components provided by this embodiment of the present invention include:

[0156] The status synchronization module is responsible for synchronizing the status of reimbursement documents to the acquiring system when the reimbursement document is rejected, withdrawn by the user, submitted for approval by the user, or archived. For rejected or withdrawn reimbursement documents, users can return them through the self-service interface, while for archived documents, physical archiving will be performed. In addition, for reimbursement documents in the "Submit for Approval" status, users are not allowed to return them on their own.

[0157] The recognition and transformation module is responsible for upgrading and transforming the reimbursement documents. At the same time, the transformation will realize the multi-invoice recognition function to support the parallel processing of different types of invoices.

[0158] Reimbursement documents include advances, repayments, business trip applications, vacation applications, transportation records for relocated personnel, transportation records for cadre exchanges, travel expenses, transportation expenses for cadre exchanges, compensatory leave, business trip and compensatory leave, travel expenses for visiting relatives, conference and consulting fee budget, conference and consulting fee settlement, business entertainment fee budget, business entertainment fee settlement, expense reimbursement, convalescent expenses, miscellaneous assets, special fee settlement, union fee budget, union fee settlement, retired employee fees and hardship allowance;

[0159] The data access control module is responsible for calling the authentication interface to obtain an access token based on the clientID and clientSecret obtained from the acquiring system. The token must be included in the HTTP request header in subsequent requests to ensure the authority to call other business interfaces. The token is valid for 60 minutes and needs to be obtained again after timeout.

[0160] The image processing conversion module is responsible for calling the corresponding interface to send all image data to the reimbursement system for optical character recognition and intelligent review after the receiving machine completes the scanning of all reimbursement documents. According to the status recognition of the reimbursement document, the printing operation of the specified reimbursement document will be automatically performed, and each reimbursement document can be physically printed one by one.

[0161] The working principle and beneficial effects of the above technical solution are as follows: the status synchronization module of this embodiment synchronizes the status of the reimbursement document to the acquiring system when the reimbursement document is rejected, withdrawn by the user, submitted for approval by the user, and the document is archived; for rejected or withdrawn reimbursement documents, the user can return the document through the self-service interface, and for archived documents, the physical archiving operation will be performed; in addition, reimbursement documents in the "submit for approval" state will not allow the user to perform independent return operations; the identification and transformation module will perform the upgrade and transformation of the reimbursement document, and at the same time, the transformation will realize the multi-invoice recognition function to support the parallel processing of different types of invoices; the reimbursement documents include advance payments, repayments, business trip applications, vacation applications, transportation registrations for relocated personnel, transportation registrations for cadre exchanges, travel expenses, transportation expenses for cadre exchanges, compensatory leave, and business trip part-time jobs. Compensatory leave, travel expenses for visiting relatives, budget for conference and consulting fees, settlement of conference and consulting fees, budget for business entertainment fees, settlement of business entertainment fees, expense reimbursement, convalescent expenses, miscellaneous assets, settlement of special fees, budget for union fees, settlement of union fees, retired employee fees and hardship allowances; the data access control module calls the authentication interface to obtain the access token based on the clientID and clientSecret obtained from the acquiring system; the token must be included in the HTTP request header in subsequent requests to ensure the authority to call other business interfaces; the token is valid for 60 minutes and needs to be obtained again after the timeout; the image processing conversion module will call the corresponding interface to send all image data to the reimbursement system for optical character recognition and intelligent review after the acquiring machine completes the scanning of all reimbursement documents. According to the status recognition of the reimbursement document, the printing operation of the specified reimbursement document will be automatically performed, and it is ensured that each reimbursement document can be physically printed one by one. The status synchronization module of the above solution realizes the instant synchronization of the reimbursement document status between different systems, ensuring that the acquiring system always has the latest document status information; allowing users to return rejected or withdrawn documents through the self-service interface, improving user experience and system convenience; clearly distinguishing the operation permissions in different states, for example, users are not allowed to return documents in the approval status, reducing the risk of misuse of operations. Significance achieved: It ensures the consistency and accuracy of data, helps prevent approval and process errors caused by inconsistent document status information; improves the transparency of the reimbursement process, allows users to understand the document status in real time, and enhances users' trust in the system; optimizes the security of the overall process through fault-tolerant mechanisms and permission control, and reduces potential risks. The identification and transformation module has systematically upgraded and transformed as many as 23 types of reimbursement documents, enhancing the system's support for various types of documents; the transformation realizes the parallel processing capability of multiple bills and improves the efficiency of bill processing. Significance achieved: Improve the flexibility and adaptability of the reimbursement system, and be able to better respond to changes and expansions in business needs; optimize the efficiency of user operations, achieve efficient management of different types of bills, and reduce operation time and error rate.The data access control module obtains the token through clientID and clientSecret to ensure that the system has the necessary permission control when interacting with data; in each subsequent interface call, the token is included in the HTTP request header to verify the requester's permission. Significance achieved: Strengthens system security to ensure that only verified requests can access sensitive information and perform important operations; significantly reduces the risk of unauthorized access and data leakage, and provides strong support for data protection. After the image processing conversion module completes the scan, the system can quickly and accurately send the image data to the reimbursement system for subsequent processing; realizes optical character recognition and intelligent review processes, automatically extracts document information and conducts review, and improves the automation and efficiency of operations; the system can automatically trigger the printing operation based on the document status recognition to ensure that each reimbursement document is physically printed. Significance achieved: Improves the efficiency and accuracy of image processing, reduces the manual review process, and improves the speed of overall business processing; through automated management and printing mechanisms, reduces the workload of operators, and optimizes resource allocation and workflow.

[0162] In summary, through the coordinated operation of various modules in this embodiment, the entire functional component system can achieve efficient and automated reimbursement document processing. Each module not only improves the functionality and reliability of the system, but also greatly enhances the user experience and system security. The integration of the design and implementation of this embodiment not only helps to improve work efficiency, but also provides more powerful support for enterprises in financial management and compliance.

[0163] Embodiment 12: Based on Embodiments 1 to 11, the data interaction process between the display and reimbursement system of the integrated financial intelligent terminal acquiring machine provided in the embodiment of the present invention and the acquiring machine includes:

[0164] When you click on the travel expense report in the reimbursement system, a pop-up window will appear asking "Is this a pure business travel document?"

[0165] (1) Select 'Yes': the user will not be able to upload receipts, and the receipt can only be reimbursed as a pure business travel receipt;

[0166] (2) Select 'No': The user will receive a QR code with the order number sent by the acquiring machine application on the Xingyun app. The user can use this QR code to place an order on the acquiring machine.

[0167] Using decoding technology, this pattern is converted back to the original information, and the Xingyun app will receive a QR code, which can be used to place an order on the receiving machine;

[0168] For documents that have been approved and electronically archived in the reimbursement system, the intelligent document collector can automatically print the corresponding approval receipt and accounting voucher for archiving, completing the closed loop of the entire process.

[0169] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. An integrated financial intelligent terminal acquiring machine, characterized in that: Include: The autonomous document delivery component is responsible for receiving reimbursement documents, scanning and identifying the reimbursement documents, and extracting the first key information of the reimbursement documents; the first key information includes the amount, date and title; The information processing component is responsible for extracting the second key information of the reimbursement document and processing the first key information and the second key information through the data processing engine. The processing includes automatic verification, classification and sorting. At the same time, the processed first key information and the second key information are electronically archived, and the corresponding approval form and accounting voucher are automatically printed for archiving; The functional interface component is responsible for providing new login authentication, receiving reimbursement document interface authentication and synchronizing reimbursement document status interface; among them, the new login authentication includes Three Gorges Xingyun code scanning authentication, face recognition authentication and work card swiping authentication.

2. The integrated financial intelligent terminal acquiring machine according to claim 1, characterized in that: Autonomous delivery components, including: The document receiving module is responsible for receiving the completed reimbursement documents and placing them within the scanning area for temporary storage; The image processing module is responsible for triggering the image capture process when the reimbursement document is placed in the scanning area and the visual sensor detects the existence of the reimbursement document; Use a visual sensor to capture an image of the reimbursement document, perform a preprocessing operation on the image, and obtain a preprocessed image; The result output module is responsible for identifying the size, direction, quality and first key information of the reimbursement document from the preprocessed image; at the same time, it provides a user interaction interface, and displays operation instructions and feedback information through a touch screen or display screen; among them, the preprocessing includes denoising and contrast enhancement.

3. The integrated financial intelligent terminal acquiring machine as claimed in claim 2, characterized in that: The result output module includes: The image selection submodule is responsible for generating multiple candidate images for each automatic capture, and selecting an image from the multiple pre-processed candidate images as the main image for subsequent processing; The physical characteristics submodule is responsible for measuring the width and height of the main image to determine the size of the reimbursement document; determining the orientation of its long and short sides to determine whether it is horizontal or vertical, and automatically adjusting the image direction according to the orientation; judging whether the main image meets the processing requirements according to the set quality standards; if it does, proceed to the next step; if it does not, prompt the user to re-place or re-scan the document; The information extraction submodule is responsible for extracting the first key information from the main image that meets the processing requirements. Each time a key information is extracted, the result is immediately fed back to the user interface and the extraction result is displayed for user confirmation.

4. The integrated financial intelligent terminal acquiring machine as claimed in claim 3, characterized in that: Physical properties submodule, including: The edge detection unit is responsible for converting the main image into a grayscale image after it is selected, binarizing the image using an adaptive Gaussian threshold, smoothing the image using a Gaussian filter, and performing edge detection using an edge detection algorithm based on the edge intensity gradient; identifying the outer border of the document in the image; The size acquisition unit is responsible for applying polygonal approximation to the extracted contour, smoothing the contour and simplifying the edges, and detecting its shape features; Calculate the minimum circumscribed rectangle of the contour, find the extreme points of the contour and perform geometry calculation to obtain the width and height of the image; The orientation adjustment unit is responsible for determining the orientation of the file based on the ratio of width to height. If width > height, it is recorded as landscape; otherwise, it is recorded as portrait. Based on the orientation information, the image is automatically rotated 90 or 180 degrees, and the rotated image is resampled using bilinear interpolation or nearest neighbor interpolation algorithms.

5. The integrated financial intelligent terminal acquiring machine as claimed in claim 4, characterized in that: Edge detection unit, including: The convolution operation subunit is responsible for edge detection of the main image after Gaussian smoothing. It uses two 3x3 convolution kernels to perform two convolution operations on the original image to obtain the gradient of the main image; calculate the gradient amplitude and direction of each pixel; The gradient calculation subunit is responsible for comparing each pixel with its neighbors in the gradient direction according to the calculated gradient direction; If the magnitude of a pixel is greater than the magnitude of its neighboring pixels, the pixel is retained; Otherwise, set its intensity to zero; The threshold comparison subunit is responsible for setting a high threshold and a low threshold, and judging the intensity of each pixel after non-maximum suppression; for pixels marked as strong edges, their connectivity is used to track the connected edges in the image, and any pixels connected to the strong edges are marked as edges to form a complete edge map.

6. The integrated financial intelligent terminal acquiring machine as claimed in claim 3, characterized in that: Information extraction submodule, including: The feature merging unit is responsible for annotating the main images that meet the processing requirements, determining the location and format of the first key information; selecting a pre-trained model and fine-tuning it on the annotated main image containing the reimbursement document; constructing an image pyramid, generating multiple versions of input images from different scales, and extracting features from the input image at each scale; merging deep features from different scales; and converting the extracted features into fixed-length feature vectors through a fully connected layer; The text acquisition unit is responsible for locating the key area using the bounding box and clarifying the relative position of the key area in the main image; using the OCR engine to perform character recognition on the extracted key area to obtain text information; and making rule judgments on the recognized text information to ensure the rules of the first key information; The information verification unit is responsible for contextual analysis of the extracted text information. It determines that the amount is in a specific area and the invoice number is in the header by setting pattern matching rules. It uses the named entity recognition algorithm to analyze the text, and the extracted first key information has more semantic context.

7. The integrated financial intelligent terminal acquiring machine according to claim 1, characterized in that: Information processing components, including: The content extraction module is responsible for extracting the second key information from the scanned image of the reimbursement document through the data processing engine OCR. The second key information includes departure time, arrival time, departure place, arrival place, passengers, seat class and ticket amount; The engine classification module is responsible for automatically verifying, classifying and organizing the first key information and the second key information using an optical character recognition engine, an intelligent verification engine, a classification engine and a data sorting engine; at the same time, the user directly extracts the reimbursement document for processing. If the reimbursement document is rejected, the user can retrieve the document on his own, and replace it or submit the attachments according to the prompts, and then submit the reimbursement document again; The archiving processing module is responsible for electronically archiving approved reimbursement documents, printing approval receipts, and archiving accounting vouchers.

8. The integrated financial intelligent terminal acquiring machine as claimed in claim 7, characterized in that: Engine classification module, including: The engine setting submodule is responsible for setting up each engine independently. The reimbursement documents to be processed are sent to the input queue for asynchronous processing. Each engine has an independent work queue for receiving tasks to be processed, and the tasks to be processed are distributed to the corresponding engines through the message queue service. The queue service submodule is responsible for quickly converting the relevant information of the reimbursement document into a processing task when the user submits the reimbursement document, and sending it to the input queue; the message queue service takes the task from the input queue and distributes it to each work queue; each engine independently obtains and processes tasks from the queue by monitoring its work queue; at the same time, each engine can run concurrently, waiting for the next task, and after processing, write the result into the result storage or send it to the next link; The result summary submodule is responsible for summarizing the processing results of all engines in the data sorting engine, and performing overall data sorting, storage and return.

9. The integrated financial intelligent terminal acquiring machine as claimed in claim 7, characterized in that: The engine classification module also includes: The data extraction submodule is responsible for using the OCR engine to perform optical character recognition on the scanned reimbursement document image, converting the pixel data in the image into machine-readable text data, including image preprocessing, feature extraction, and character recognition; the output is a text data structure; The preliminary verification submodule is responsible for performing a structured preliminary verification of the text data extracted by the OCR engine, including verifying whether the text format conforms to the preset pattern, checking whether the required fields exist, and identifying characters or strings that cannot be accurately recognized; The logic verification submodule is responsible for performing advanced logic verification on the data that has passed the preliminary verification, and cross-validating through predefined rules and business logic; The exception handling submodule is responsible for automatically marking all abnormal data found during the logic verification phase as objects that need attention, generating an error report that lists every non-compliance item found in detail; at the same time, relevant personnel will receive notifications about the abnormal situation so that they can intervene and handle it in time; The feedback and correction submodule is responsible for presenting documents marked as abnormal in the user interface. Users can access the error report and view the specific error information. A user-friendly interface will be provided to assist users in revising and supplementing abnormal data. After the correction is completed, the user can submit the corrected document back to the system for re-verification; The final confirmation submodule is responsible for marking the document data that meets the requirements as valid after all verification steps are completed; At this point, the document data that has been reviewed and confirmed will be stored in the database and marked as ready to enter the subsequent approval process.

10. The integrated financial intelligent terminal acquiring machine according to claim 1, characterized in that: Functional components, including: The status synchronization module is responsible for synchronizing the status of reimbursement documents to the acquiring system when the reimbursement document is rejected, withdrawn by the user, submitted for approval by the user, or archived. For rejected or withdrawn reimbursement documents, users can return them through the self-service interface, while for archived documents, physical archiving will be performed. In addition, for reimbursement documents in the submission for approval status, users are not allowed to return them on their own. The recognition and transformation module is responsible for upgrading and transforming the reimbursement documents. At the same time, the transformation will realize the multi-invoice recognition function to support the parallel processing of different types of invoices. The data access control module is responsible for calling the authentication interface to obtain an access token based on the clientID and clientSecret obtained from the acquiring system; the token must be included in the HTTP request header in subsequent requests; The image processing conversion module is responsible for calling the corresponding interface to send all image data to the reimbursement system for optical character recognition and intelligent review after the acquiring machine completes the scanning of all reimbursement documents; according to the status recognition of the reimbursement document, it will automatically perform the printing operation of the specified reimbursement document and ensure that each reimbursement document can be physically printed one by one.

Images