Business single photographing control method and device, equipment and medium

Abstract

The application relates to the technical field of image processing, and discloses a business single photographing control method, device, equipment and medium, which comprises the following steps: confirming a single certificate type and quantity according to business request data, and constructing a photographing sequence list; acquiring user photographing configuration, and judging whether an AI module is enabled or not. If the AI module is enabled, embedded AI components are called to photograph in sequence, and a picture sequence is generated. If the AI module is not enabled, it is judged whether the single certificate is a preset single certificate type: if yes, a Canvas dynamic screenshot component is called to guide photographing; otherwise, an original camera is called to photograph. All the photographing is strictly performed according to the photographing sequence list, and finally a picture sequence is generated, which can be applied to a financial technology, medical health and other business system platform, and can realize sequence control of business single photographing. The application improves the image quality of business single photographing.

Description

Technical Field

[0001] This invention relates to the field of image processing technology, and in particular to a business single-shot control method, apparatus, device, and medium. Background Technology

[0002] In the fintech sector, particularly in life insurance bancassurance, bank lending, and securities account opening, customers often need to upload multiple pages of documents (such as electronic application forms, disclaimers, identity documents, and signature pages) to complete the transaction. Existing document processing systems generally employ a "free upload + simple verification + manual check" model. This means users either take photos page by page using the system's native camera or select images from their album and upload them all at once. The system only performs basic checks on image format and size; the document order and image quality primarily rely on manual inspection. This model has obvious drawbacks: On the one hand, when users freely upload multiple pages of documents, the lack of a mandatory page number binding mechanism often results in disordered page order (such as uploading the signature page as the first page) or missing key pages, requiring manual reordering or resubmission by customers in subsequent quality control stages, which seriously affects business timeliness; on the other hand, key documents such as electronic insurance applications have fixed aspect ratio requirements (to facilitate quality control comparison and OCR recognition), but the existing system lacks effective shooting guidance. The images taken by users often have problems such as excessive margins, fingers obscuring the image, insufficient lighting, and perspective distortion, resulting in a low quality control pass rate. A lot of manpower is needed later for image cropping, correction, and resubmission, increasing operating costs and customer waiting time.

[0003] In the healthcare field, scenarios such as electronic medical record archiving, medical insurance reimbursement applications, and digitization of health records also involve the uploading and quality control of a large number of sequential documents. For example, after a patient's visit, they need to upload multiple pages of documents, including prescriptions, test request forms, inpatient expense lists, surgical informed consent forms, and medical insurance settlement statements. These documents usually have strict page order requirements (e.g., prescriptions first, then payment slips, then medical insurance reimbursement slips), and some key documents (such as the doctor's signature page and the hospital's stamped page) require clear, unobstructed images that meet archiving standards. Currently, medical institutions or third-party service platforms mostly use the common photo upload method, where users take photos one by one with their mobile phone cameras and then manually sort them or rely on manual annotation. The technical problems with this approach are as follows: First, the lack of an automated sequence control mechanism makes it easy for users to confuse the order of different documents, leading to rejection during medical insurance review or filing due to incorrect order. Second, the lack of targeted shooting guidance for key signature and stamp areas often results in problems such as reflection, blurring, and tilting in the images, failing to meet the requirements of medical data exchange standards (such as HL7 and FHIR) and the "Electronic Medical Record Application Management Specifications" for image authenticity and readability. Third, the existing system does not integrate embedded image preprocessing technology (such as edge detection, edge removal, and skew correction), resulting in inconsistent quality of uploaded raw images and significantly increasing the time cost of manual verification and processing in the background.

[0004] In summary, existing document processing solutions suffer from the following technical deficiencies in both the fintech and healthcare fields: First, lack of sequence control—when users freely upload multiple pages of documents, there is no mandatory page number binding or sequence constraint, which easily leads to misplaced or missing pages, requiring subsequent manual intervention for sorting; Second, insufficient shooting guidance—key documents (such as signature pages and pages with fixed aspect ratios) lack effective alignment and cropping mechanisms, making it difficult for users to independently capture compliant images, resulting in low quality control pass rates; Third, weak image preprocessing capabilities—mainstream solutions do not integrate real-time processing technologies such as embedded edge detection, edge removal, enhancement, and skew correction, requiring post-processing for issues such as edge retention, dirt, and distortion in the captured images. Summary of the Invention

[0005] This invention provides a business order shooting control method, device, computer equipment, and medium to solve the problems of lack of sequential control and low shooting quality in existing business order shooting control methods on the market.

[0006] Firstly, a business order shooting control method is provided, including: Confirm the document type and quantity of the business order to be uploaded based on the pre-acquired business request data; The document types are sorted according to preset rules, and a page-by-page shooting task is constructed for each document type according to the number of documents corresponding to each document type, generating a shooting order list; Retrieve the shooting configuration data stored by the user on the preset server; Determine whether the user has enabled the AI ​​module based on the shooting configuration data; If the user enables the AI ​​module, the embedded AI component preset in the target shooting device is invoked to shoot the business order according to the shooting order list and generate an image sequence. If the user has not enabled the AI ​​module, then determine whether the document type is a preset document type; If the document type is a preset document type, then the preset Canvas dynamic screenshot component in the target shooting device is invoked to shoot the business order according to the shooting order list, and an image sequence is generated; If the document type is not a preset document type, the native camera in the target shooting device is invoked to shoot the business order according to the shooting order list, and an image sequence is generated.

[0007] Secondly, a business order shooting control device is provided, including: Thirdly, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-described business single-shot control method.

[0008] Fourthly, a computer-readable storage medium is provided, which stores a computer program that, when executed by a processor, implements the steps of the above-described business single-shot control method.

[0009] In the above-mentioned business order shooting control method, device, computer equipment, and storage medium, the document type and quantity of the business order to be uploaded can be confirmed based on the pre-acquired business request data, providing an accurate data benchmark for subsequent sequence control. During the shooting process according to the shooting order list, the system verifies in real time whether the current shooting page is consistent with the expected page number. If the user skips a page, reshoots, or misses a shot, a prompt is given on the spot and submission is blocked until all pages are shot in sequence, thereby effectively reducing the risk of missing or incorrect documents due to human negligence. Then, the document types are sorted according to preset rules, and a page-by-page shooting task is constructed for each document type according to the quantity of documents corresponding to each document type, generating a shooting order list to ensure that the final image sequence is completely consistent with the business requirements, eliminating the need for manual sorting in the later stage and significantly improving the quality control pass rate and business processing efficiency. Next, the shooting configuration data stored by the user on the preset server is obtained, and it is determined whether the user has enabled the AI ​​module based on the shooting configuration data: if the user has enabled the AI ​​module, the preset embedded AI component in the target shooting device is called to shoot the business order according to the shooting order list and generate an image sequence, greatly reducing the need for manual sorting. This reduces processing costs, improves document image quality and subsequent OCR recognition accuracy. If the user has not enabled the AI ​​module, it further determines whether the document type is a preset document type. If the document type is a preset document type, it calls the preset Canvas dynamic screenshot component in the target shooting device to shoot the business document according to the shooting order list, generating an image sequence and improving image quality. If the document type is not a preset document type, it calls the native camera in the target shooting device to shoot the business document according to the shooting order list, generating an image sequence, thus realizing the order control of business document shooting and improving shooting quality. Attached Figure Description

[0010] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0011] Figure 1 This is a schematic diagram of an application environment for a business order shooting control method according to an embodiment of the present invention; Figure 2 This is a flowchart illustrating a business order shooting control method according to an embodiment of the present invention; Figure 3 This is a schematic diagram of a business order shooting control device in one embodiment of the present invention; Figure 4 This is a schematic diagram of the structure of a computer device according to an embodiment of the present invention; Figure 5 This is another structural schematic diagram of a computer device according to one embodiment of the present invention. Detailed Implementation

[0012] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0013] The business single-shot control method provided in this embodiment of the invention can be applied to, for example... Figure 1 In this application environment, the client communicates with the server via a network. The server can use the client to confirm the document type and quantity of the business order to be uploaded based on pre-acquired business request data. Then, it sorts the document types according to preset rules and constructs a page-by-page shooting task for each document type based on the quantity of documents corresponding to each document type, generating a shooting order list. Next, it obtains the shooting configuration data stored by the user on a preset server and determines whether the user has enabled the AI ​​module based on the shooting configuration data: if the user has enabled the AI ​​module, it calls the preset embedded AI component in the target shooting device to shoot the business order according to the shooting order list, generating an image sequence; if the user has not enabled the AI ​​module, it further determines whether the document type is a preset document type—if the document type is a preset document type, it calls the preset Canvas dynamic screenshot component in the target shooting device to shoot the business order according to the shooting order list, generating an image sequence; if the document type is not a preset document type, it calls the native camera in the target shooting device to shoot the business order according to the shooting order list, generating an image sequence. This realizes the sequential control of business order shooting and improves shooting quality. The client can be, but is not limited to, various personal computers, laptops, smartphones, tablets, and portable wearable devices. The server can be implemented using a standalone server or a server cluster consisting of multiple servers. The invention will now be described in detail through specific embodiments.

[0014] Please see Figure 2 As shown, Figure 2 A flowchart illustrating a business order shooting control method provided in an embodiment of the present invention includes the following steps: S1. Based on the pre-acquired business request data, confirm the document type and quantity of the business order to be uploaded.

[0015] In the fintech field, a business order is a structured electronic voucher used to record core information about financial transactions or business applications. It typically includes key fields such as customer identity information, product type, amount, terms, and signature, and exists as standardized document formats (such as PDF, OFD, and HTML pages) or key-value pairs in JSON / XML. For example, in life insurance bancassurance, the electronic application form serves as a business order. Its data includes structured data such as policy number, policyholder name, insured information, and insurance amount, as well as image pages that require quality control (such as signature pages and disclosure pages). Each page has a fixed page order and aspect ratio requirements to facilitate subsequent automatic comparison and compliance review.

[0016] In the healthcare field, the aforementioned business orders typically refer to document-type data related to patient diagnosis and treatment, cost settlement, or public health management. Common forms include prescriptions, test request forms, inpatient expense lists, medical insurance settlement statements, and nucleic acid test registration forms in electronic medical records. The data content is mostly semi-structured or unstructured: on the one hand, it contains key fields such as patient ID, consultation time, drug name, examination items, and cost amount; on the other hand, it includes image areas such as handwritten signatures, doctor's seals, and barcodes. These business orders need to meet the medical data exchange standards and legal requirements for the storage and quality control of electronic medical records. During the digitization process, the page order (e.g., prescription before payment) and compliant photography of key areas (e.g., the doctor's signature area must be clearly identifiable) are also emphasized.

[0017] In this embodiment of the invention, confirming the document type and quantity of the business order to be uploaded based on pre-acquired business request data includes: Obtain the service identifier from the service request data; Based on the business identifier, query business scenario information in a preset business database; Map the business scenario information to a business type; Load the document configuration rule table from the business database based on the business identifier; Based on the business type, query the standard page number corresponding to each document type in the document configuration rule table, and generate the number of documents.

[0018] In detail, obtaining the business identifier in the business request data involves parsing a field that can uniquely identify the current business order from the request parameters passed from the upstream business system, such as "policy application number", "insurance application number" or "policy maintenance request number". This field is usually encapsulated in a fixed node of the business request data in JSON or XML format, and its string value is directly read and extracted by the field name (such as "policyNo") as a key index for subsequent queries.

[0019] In detail, the step of querying business scenario information in a preset business database based on the business identifier involves using the extracted business identifier as a query condition to send a structured query statement (SQL) to the business database (such as the policy master table in a relational database like MySQL or Oracle) to retrieve the "business type" field (such as "new contract", "preservation", "claims") and the "product code" field (such as "ABC001") corresponding to the business order. If the business database is a distributed storage, the routing algorithm of the business identifier will locate the specific shard to execute the query.

[0020] In detail, mapping the business scenario information to a business type involves normalizing and combining the two original fields, "business type" and "product code," obtained from the query to generate a unique business type code (e.g., "new contract_ABC001"). This mapping process can be completed using the business type mapping table maintained by the configuration center. If the original field value is already defined in the mapping table, the corresponding standardized business type identifier is returned directly. If it is not defined, the business type is generated according to the default rules (e.g., "business type + product code" concatenation) for use when matching the document configuration rule table later.

[0021] In detail, the step of loading the document configuration rule table from the business database based on the business identifier involves dynamically loading a predefined "document configuration rule table" for the specific business scenario from the configuration database or configuration center according to the business line or organization information to which the business identifier belongs. This rule table uses the business type as the primary key and records a list of document types that need to be uploaded under each business type (such as electronic insurance application, disclaimer, etc.), the number of pages for each document, and a flag indicating whether it is a critical document. The loading method can employ a strategy of local caching plus remote retrieval to reduce database access pressure.

[0022] In detail, the step of querying the standard page number corresponding to each document type in the document configuration rule table based on the business type to generate the document quantity involves using the mapped business type as the query key to filter out all matching record rows in the loaded document configuration rule table; for each row, the "document type" and "page number" fields are extracted to form a key-value pair list (e.g., "electronic insurance form": 3, "disclaimer": 6); finally, this list is output as a document configuration mapping table, where the "page number" is the document quantity corresponding to each document type, for use in the subsequent construction of the shooting order list.

[0023] In this embodiment of the invention, by confirming the document type and quantity of the business order to be uploaded based on the pre-acquired business request data, the tedious operation of manually selecting or memorizing the document list is avoided, the risk of missing or incorrect documents due to human negligence is reduced, and an accurate data benchmark is provided for subsequent sequence control.

[0024] S2. Sort the document types according to preset rules, and construct a page-by-page shooting task for each document type according to the number of documents corresponding to each document type, and generate a shooting order list.

[0025] In this embodiment of the invention, the step of sorting the document types according to preset rules, constructing a page-by-page shooting task for each document type based on the number of documents corresponding to each document type, and generating a shooting order list includes: The document types of each business order are sorted according to preset rules to form a sorted document type table. Add the number of documents for each document type to the sorted document type table to obtain the sorted document configuration mapping table. Based on the number of documents corresponding to each document type in the sorted document configuration mapping table, the number of shooting task entries corresponding to each document type is determined. Based on the current order of each document type and the corresponding number of shooting task entries in the sorted document configuration mapping table, a shooting order list is generated.

[0026] In detail, the step of sorting the document types of each business order according to preset rules to form a sorted document type table involves extracting all the "document type" fields from the confirmed document configuration mapping table to form a list of document types to be sorted; then reading the system's preset sorting rules—which are usually maintained in the form of configuration files or database tables—which stipulate that key documents (such as electronic insurance applications) must be listed first, and other documents are arranged in order of business importance or quality control requirements (such as disclaimers second, identity certificates last). If a document type is not explicitly defined in the rules, it is appended to the end of the list in the order of its original appearance; the sorting algorithm can use stable sorting (such as bubble sort or quick sort) to ensure that the relative order of documents with the same priority remains unchanged, and finally outputting a "document type table" sorted according to the rules.

[0027] In detail, adding the number of documents for each document type to the sorted document type table to obtain the sorted document configuration mapping table involves traversing the "sorted document type table" obtained in the previous step. For each document type in the table, the "page number" field (i.e., the number of documents) corresponding to that type is queried from the original document configuration mapping table, and the "page number" is appended as a new attribute column to the document type. For example, if the original mapping table contains "electronic insurance application: 3 pages", the page number information is still retained after sorting, forming a key-value pair structure for "electronic insurance application". All document types and their page numbers are organized together in the sorted order to generate a two-dimensional table structure "sorted document configuration mapping table". The rows of this table represent each document type, and the columns include "document type", "page number", and an optional "whether it is a critical document" flag.

[0028] In detail, the number of shooting task entries corresponding to each document type is determined based on the number of documents corresponding to each document type in the sorted document configuration mapping table. Based on the current order of each document type and the corresponding number of shooting task entries in the sorted document configuration mapping table, each document type is processed sequentially. For the document type and its page number N in the current row, N shooting task entries are generated in a loop N times. Each entry contains an incrementing global page number, the name of the document type, a critical document flag, and an initial state "to be shot". For example, if the electronic insurance application has 3 pages, then page numbers 1, 2, and 3 are generated. Then, the next document type, the disclaimer, has 6 pages, and the entries with page numbers 4 to 9 are generated, and so on. All entries are arranged in the order of generation, and finally a complete "shooting order list" is formed. The total number of entries in this list is equal to the sum of the page numbers of all documents. The subsequent shooting process will strictly follow the page number order of this list to execute the shooting tasks one by one.

[0029] In this embodiment of the invention, the document types are sorted according to preset rules, and a page-by-page shooting task is constructed for each document type according to the number of documents corresponding to each document type, generating a shooting order list, thereby improving the subsequent quality control pass rate and business processing efficiency.

[0030] S3. Obtain the shooting configuration data stored by the user on the preset server.

[0031] In this embodiment of the invention, obtaining the shooting configuration data stored by the user on a preset server includes: Obtain the user's identity identifier and generate a configuration query statement based on the identity identifier; A configuration query request is initiated to the server according to the configuration query statement, and the requested data is obtained; The shooting configuration data contained in the request data is extracted by parsing the request data.

[0032] In detail, obtaining the user's identity identifier means obtaining a unique credential for identifying the user from the current login session on the client, which is usually the user ID of the logged-in user (such as "user_123456"). This identity identifier can be stored in the client's memory variables, local storage, or secure sharing preferences, and can be read directly by calling the user information interface provided by the client.

[0033] In detail, generating a configuration query statement based on the identity identifier involves using the obtained user identity identifier as a query parameter and constructing a standard configuration query statement according to a preset server-side API interface specification. This query statement can be a RESTful style URL path parameter (such as " / api / v1 / config / user_123456"), a GET request with query parameters (such as " / api / v1 / config?userId=user_123456"), or a GraphQL query body or a binary request body based on Protobuf encoding. The specific format depends on the communication protocol agreed upon by the front-end and back-end of the system.

[0034] In detail, the step of initiating a configuration query request to the server based on the configuration query statement and obtaining the request data is achieved by using a network request module built into the client (such as HTTP Client, Axios, OkHttp, etc.) to encapsulate the configuration query statement generated in the previous step into a network request packet, setting appropriate request headers (such as Content-Type, Authorization, etc.), and sending the request to the preset server address. After receiving the request, the server queries the user's shooting configuration data (such as AI module enable flag, image quality, resolution, etc.) from the database or cache based on the identity identifier, and returns the result to the client in the form of a response body (usually in JSON format). The client receives this response, which is the original request data.

[0035] In detail, the step of extracting the shooting configuration data contained in the request data by parsing the request data involves parsing the response body returned by the server, first checking the response status code (such as HTTP 200 indicating success), and then extracting the configuration fields according to the agreed data format; for example, if the response body is {"code":0, "data":{"ai_module_enabled":true, "quality":"high"}}, then all fields under the data object are parsed as shooting configuration data. If parsing fails or the server returns an error code, the system's preset default configuration (such as AI module off, quality medium, etc.) is used as a fallback to ensure that subsequent processes can continue to be executed.

[0036] In this embodiment of the invention, by acquiring the shooting configuration data stored by the user on a preset server, the configuration is stored in the cloud and synchronized with multiple devices, thereby improving the user experience.

[0037] S4. Determine whether the user has enabled the AI ​​module based on the shooting configuration data.

[0038] If the user enables the AI ​​module, then S5 is executed: by calling the preset embedded AI component in the target shooting device, the business order is photographed according to the shooting order list, and an image sequence is generated.

[0039] In this embodiment of the invention, the step of capturing the business order according to the shooting order list by calling a preset embedded AI component in the target shooting device to generate an image sequence includes: The original images of each business order are obtained according to the shooting order to obtain the original image sequence; The embedded AI component is used to perform edge detection on each image in the original image sequence to generate edge detection results; Based on the edge detection results, each image in the original image sequence is processed to remove edges, resulting in an edge-removed image sequence. Each image in the edge-removed image sequence is subjected to brightness enhancement and contrast enhancement processing to obtain an enhanced image sequence; The embedded AI component is used to perform distortion correction processing on each image in the enhanced image sequence to obtain a corrected image sequence.

[0040] In detail, the process of acquiring the original images of each business order according to the shooting order to obtain the original image sequence involves the embedded AI component calling the camera of the target shooting device to capture the image of each business order in real time during the process of the user shooting page by page according to the shooting order list. For the current page number to be shot, the device will take a clear preview frame or a high-definition frame when the user presses the shutter, store the frame image as raw data in bitmap or YUV format in the memory buffer, and then process the next page in sequence. After all pages have been shot, the original images are organized into an "original image sequence" according to the page number order.

[0041] In detail, the step of using the embedded AI component to perform edge detection on each image in the original image sequence and generate an edge detection result involves sequentially inputting each frame of the original image sequence into the edge detection model in the embedded AI component; the model performs grayscale conversion and Gaussian filtering on the image, calculates the gradient magnitude and direction of each pixel, and finally outputs a binary edge detection result image (white pixels represent possible document boundaries, and black pixels represent the background) through non-maximum suppression and double threshold connection. This result includes obvious edge information such as the outer contour of the document and internal page numbers and signature lines.

[0042] In detail, the process of removing edges from each image in the original image sequence based on the edge detection results to obtain a sequence of edge-removed images involves using a contour finding algorithm (such as OpenCV's findContours) to locate the outermost quadrilateral contour of the document based on the edge detection result image generated in the previous step. Then, the minimum bounding rectangle of the contour is calculated or the coordinates of the four vertices are directly obtained. Perspective transformation is then used to stretch the region within the contour in the original image into a standard rectangular image, thereby automatically cropping out irrelevant areas such as the background, fingers, and margins, retaining only the document itself. This process is performed independently on each image in the original image sequence to generate a set of "edge-removed image sequences" with redundant boundaries removed.

[0043] In detail, the process of performing brightness enhancement and contrast enhancement on each image in the edge-removed image sequence to obtain the enhanced image sequence involves converting the edge-removed images from the RGB color space to the HSV or LAB color space, applying an adaptive histogram equalization (CLAHE) algorithm to the brightness channel (L or V) to limit the contrast amplification and suppress noise, then moderately stretching the saturation channel, and finally converting back to the RGB space. For images with low light, gamma correction can also be used to enhance shadow details. This process can be executed frame-by-frame by an image enhancement operator library integrated in the embedded AI component (such as OpenCV Mobile or a lightweight version of Halcon), outputting an "enhanced image sequence" with clearer visual effects and sharper text.

[0044] In detail, the process of using the embedded AI component to perform distortion correction processing on each image in the enhanced image sequence to obtain a corrected image sequence involves sequentially detecting and correcting trapezoidal distortion and barrel / pincushion distortion caused by lens distortion or shooting angle in the enhanced images. First, the images are dedistorted by mapping using the camera calibration parameters (distortion coefficients pre-stored on the device) of the embedded AI component. Then, perspective correction is performed using the homography matrix between the vertices of the document and the vertices of the ideal rectangle in the edge detection results, so that the border of the business form appears as a regular rectangle in the final image and the aspect ratio meets the quality control requirements (such as the fixed aspect ratio of an electronic insurance policy). After each frame is corrected, a standardized business form image is output.

[0045] In this embodiment of the invention, the business order is photographed according to the shooting order list by calling the embedded AI component preset in the target shooting device, generating an image sequence, reducing manual processing costs, and improving the quality of document images and subsequent recognition accuracy.

[0046] If the user has not enabled the AI ​​module, then execute S6 to determine whether the document type is a preset document type.

[0047] In this embodiment of the invention, the preset document type refers to "electronic insurance application".

[0048] If the document type is a preset document type, then execute S7: by calling the preset Canvas dynamic screenshot component in the target shooting device to shoot the business order according to the shooting order list, and generate an image sequence.

[0049] In this embodiment of the invention, the step of capturing the business order according to the shooting order list by calling the preset Canvas dynamic screenshot component in the target shooting device to generate an image sequence includes: Obtain a screenshot image captured by the target imaging device; Extract the edge position information of the business order from the captured screenshot image; Obtain the reference frame position information of the preset reference frame in the captured screenshot image; Calculate the positional deviation between the edge of the business order and the reference frame based on the edge position information and the reference frame position information; Determine whether the positional deviation is greater than a preset deviation threshold; If the position deviation is greater than the deviation threshold, then a position adjustment prompt is generated based on the position deviation and displayed to the user based on the position deviation, and then the process returns to the step of obtaining the screenshot image captured by the target shooting device. If the position deviation is less than or equal to the deviation threshold, then the business order image in the captured screenshot image is extracted based on the edge position information; After the images of each business order are extracted, the image sequence is generated.

[0050] In detail, acquiring the screenshot image from the target shooting device is achieved by using the Canvas dynamic screenshot component to call the real-time video stream of the target shooting device (such as the rear camera of a mobile phone or tablet), and then... <video>The current frame image in an element is drawn to a hidden or visible element. <canvas>On the element, the pixel data of the frame is obtained through canvas.toDataURL() or canvas.getContext('2d').getImageData() methods to generate a static screenshot image; the resolution of this screenshot image is usually consistent with the video stream and can be used for subsequent edge detection and alignment judgment.

[0051] In detail, obtaining the reference frame position information of the preset reference frame in the captured screenshot image involves reading pre-set reference frame data from the configuration parameters of the Canvas component. This reference frame is a fixed semi-transparent rectangle drawn on the video screen, and its position, width, height, and aspect ratio are all predefined by the system (e.g., set according to the fixed aspect ratio of the key electronic insurance policy). The coordinates of the upper left and lower right corners of the reference frame are stored in the component's state variables and can be directly retrieved as the reference frame position information. This information is determined during component initialization and does not change with the user's shooting angle.

[0052] In detail, the step of calculating the positional deviation between the edge of the business order and the reference frame based on the edge position information and the reference frame position information involves comparing the coordinates of the four vertices of the rectangular outline of the business order with the coordinates of the four vertices of the reference frame, and calculating the offset in the horizontal direction (e.g., the pixel difference between the left edge and the left edge of the reference frame, and the pixel difference between the right edge and the right edge of the reference frame) and the offset in the vertical direction (the difference between the top edge and the bottom edge). The deviation can be the maximum value, the average value, or the independent deviation vector of each side. The specific calculation formula is: horizontal deviation = x-coordinate of the left edge of the business order - x-coordinate of the left edge of the reference frame, vertical deviation = y-coordinate of the top edge of the business order - y-coordinate of the top edge of the reference frame, and finally outputting a positional deviation data structure containing the direction and magnitude of the deviation.

[0053] In detail, determining whether the positional deviation is greater than a preset deviation threshold involves comparing the calculated positional deviation (e.g., a deviation of 15 pixels on the left edge and a deviation of -8 pixels on the top edge) with preset deviation thresholds (e.g., ±10 pixels allowed in the horizontal direction and ±10 pixels allowed in the vertical direction) item by item. If the absolute value of the deviation in any direction exceeds the corresponding threshold, it is determined to be "greater than the threshold". Conversely, if the deviations in all directions are within the threshold range, it is determined to be "less than or equal to the threshold". This threshold can be dynamically configured according to the strictness of the actual quality control requirements, and is usually set to 1% to 3% of the total image width.

[0054] In detail, the step of generating a position adjustment prompt based on the position deviation if the position deviation exceeds the deviation threshold, and then displaying it to the user based on the position deviation before returning to the step of obtaining the screenshot of the target shooting device, involves generating user-understandable prompt text based on the direction and magnitude of the deviation when the deviation exceeds the threshold. For example, "Please move the document to the left by about 10 pixels" or "Please move the document up and zoom in slightly." The prompt can be displayed to the user through text bubbles or speech synthesis in the Canvas component interface. After the user adjusts the position of the document or the angle of the phone according to the prompt, the system automatically or manually refreshes the video stream, obtains a new screenshot image, and repeats the above detection process until the deviation meets the requirements.

[0055] In detail, if the positional deviation is less than or equal to the deviation threshold, then extracting the business order image from the captured screenshot based on the edge position information involves performing a perspective transformation on the current captured screenshot image using the business order edge position information (coordinates of the four vertices) detected in the previous step when the deviation meets the requirements: mapping the irregular quadrilateral region of the business order to a standard rectangle with a fixed aspect ratio through a homography matrix, while cropping the background area outside the reference frame and leaving the edges, resulting in an image containing only the main body of the business order and with perspective distortion corrected; this extraction process is directly implemented using Canvas's drawImage and transformation matrix, outputting a business order image that perfectly matches the size of the reference frame.

[0056] In detail, after the images of each business order are extracted, the generation of the image sequence involves repeating the entire process from obtaining the screenshot to extracting the image for each page number (each page of the business order) in the shooting order list. Each time a page of business order images is successfully extracted, it is stored in a temporary array in page number order, and the status of the corresponding entry in the shooting order list is updated to "shot". When all pages have been shot and the images have been extracted, the image array sorted in ascending page number order is output as the final image sequence for subsequent quality control and archiving storage.

[0057] In this embodiment of the invention, extracting the edge position information of the business order in the captured screenshot image includes: Convert the captured screenshot image into a grayscale image; The grayscale image is subjected to Gaussian blur denoising processing to obtain a denoised image; Calculate the gradient magnitude of the denoised image to obtain a gradient magnitude image; Pixels in the gradient magnitude image with magnitudes higher than a threshold are marked as strong edges, generating an edge-marked image; Extract the largest rectangular contour from the edge-marked image to obtain the edge position information.

[0058] In detail, the step of converting the captured screenshot image into a grayscale image involves using the Canvas component. The captured screenshot image is typically an RGB three-channel color image. To reduce computation and highlight brightness changes, a weighted average method is used to convert the R, G, and B components of each pixel into single-channel grayscale values ​​according to human eye sensitivity (e.g., 0.299×R + 0.587×G + 0.114×B), generating a grayscale image with the same width and height as the original image. This step can be efficiently completed in the Canvas's JavaScript environment by iterating through the ImageData pixel array or calling the cvtColor function in OpenCV.js.

[0059] In detail, the Gaussian blur denoising process performed on the grayscale image to obtain the denoised image involves convolving the grayscale image with a two-dimensional Gaussian kernel (e.g., 5×5 in size, standard deviation σ=1.4). The new value of each pixel is obtained by weighted averaging of its neighboring pixels, with the weights following a Gaussian distribution. Gaussian blur can effectively smooth out minor brightness variations caused by camera sensor noise and uneven lighting, while preserving the main structure of the business single edge. In actual mobile implementation, the Canvas filter property or WebAssembly-optimized convolution library can be used to accelerate the calculation.

[0060] In detail, the step of calculating the gradient magnitude of the denoised image to obtain a gradient magnitude image involves using the Sobel operator to calculate the first derivative of the denoised image in the horizontal direction (Gx) and the vertical direction (Gy) respectively. For each pixel, the gradient magnitude = sqrt(Gx² + Gy²) and the gradient direction = arctan(Gy / Gx). After the calculation, a gradient magnitude image with the same size as the original image is obtained. The larger the pixel value, the more drastic the brightness change at that point, and the more likely it is to be an edge point. This step can be implemented using OpenCV's Sobel function or by writing a custom convolution loop.

[0061] In detail, the step of marking pixels with amplitude values ​​higher than the amplitude threshold in the gradient amplitude image as strong edges and generating an edge-marked image adopts the dual-threshold strategy in the Canny algorithm: First, a high threshold is set (e.g., 30% of the maximum gradient amplitude). All pixels in the gradient amplitude image are traversed, and pixels with amplitude values ​​higher than the high threshold are directly marked as strong edges (e.g., assigned a value of 255). Pixels with amplitude values ​​between the low and high thresholds are marked as weak edges (assigned a value of 128), and pixels below the low threshold are suppressed as background (0). Then, through the edge connection algorithm, weak edges connected to strong edges are also promoted to strong edges, and finally a binary edge-marked image containing only continuous and closed edges is generated, where white pixels are reliable business single-boundary candidate points.

[0062] In detail, the step of extracting the largest rectangular contour from the edge-marked image to obtain the edge position information involves using a contour search algorithm (such as OpenCV's findContours in RETR_EXTERNAL mode) to extract all closed contours from the edge-marked image; calculating the minimum bounding rectangle or fitted quadrilateral for each contour, and filtering based on conditions such as area, aspect ratio, and rectangularity—typically, business orders occupy a large area in the image and are close to rectangular in shape; selecting the contour with the largest area as the business order contour, and outputting the coordinates of its four vertices (top left, top right, bottom right, bottom left) in a uniform order as edge position information; if no rectangular contour that meets the conditions is found, a null value is returned and a reshoot guidance is triggered.

[0063] In this embodiment of the invention, the business order is captured by calling the preset Canvas dynamic screenshot component in the target shooting device according to the shooting order list, generating an image sequence, which avoids interference from background, fingers, etc., and improves image quality.

[0064] If the document type is not a preset document type, then execute S8: by calling the native camera in the target shooting device, shoot the business order according to the shooting order list, and generate an image sequence.

[0065] In this embodiment of the invention, after the business order is photographed by calling the native camera in the target shooting device according to the shooting order list and generating an image sequence, the image sequence is sent to the server.

[0066] In the fintech field, this solution can be applied to scenarios requiring the uploading of multiple pages of documents and strict quality control, such as life insurance bancassurance business, bank account opening, and loan applications. The system automatically generates a sequential list of shooting tasks based on the business type, and through a tiered shooting strategy, ensures that the aspect ratio, clarity, and page order of key documents (such as electronic insurance applications and signature pages) are fully compliant. It automatically eliminates background interference and perspective distortion, thereby improving the document quality control pass rate, reducing manual review costs, and ensuring the compliant and efficient operation of financial business processes.

[0067] In the healthcare field, this solution can be applied to scenarios requiring the sequential collection of medical documents, such as electronic medical record archiving, medical insurance reimbursement applications, and health record digitization. The system automatically matches prescriptions, test request forms, expense lists, informed consent forms, etc., to be photographed based on business rules. A sequential control mechanism forces photographing in page order. For key documents (such as signature pages and stamped pages), Canvas reference frames or AI-assisted correction are used to ensure image quality meets medical data standards (such as HL7 and FHIR) and legal archiving requirements. This effectively avoids reimbursement rejections or medical disputes due to confusing documents or unclear images, improving the efficiency and quality of medical information digitization.

[0068] As can be seen, the above solution proposes a business document shooting control method to solve problems such as disordered order, poor image quality, and non-compliance of key documents when manually uploading documents in financial businesses (such as life insurance and bancassurance). The solution first automatically confirms the type of document to be uploaded and the standard number of pages for each document (e.g., 3 pages for an electronic application form, 6 pages for a disclaimer) based on the business request data. Then, it constructs a page-by-page shooting order list according to the quality control requirements, binding a unique identifier to each page number. Next, it obtains the shooting configuration data stored by the user on the server side and determines whether the AI ​​module is enabled. If AI is enabled, the embedded AI component is called to perform real-time edge detection, edge removal, image enhancement, and perspective correction, directly outputting compliant images. If AI is not enabled, it further determines whether the current document is a preset key document (such as an electronic application form). For key documents, the Canvas dynamic screenshot component is activated, overlaying a reference frame to guide the user to align and perform hard cropping, generating a standardized image with a fixed aspect ratio. For ordinary documents, the system's native camera is directly used for shooting. All shooting strictly follows the shooting order list, ensuring that the images are arranged in page order, ultimately resulting in a complete image sequence. This solution eliminates missing pages and out-of-order documents through a sequence control mechanism and balances compliance and user experience through a tiered shooting strategy. It significantly improves the pass rate of document quality control, reduces labor costs, and is suitable for fields such as finance and healthcare that require rigorous processing of sequential documents.

[0069] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0070] In one embodiment, a business order shooting control device is provided, which corresponds one-to-one with the business order shooting control method described in the above embodiments. For example... Figure 3 As shown, the single-shot control device includes a data acquisition module 101, a list construction module 102, a configuration acquisition module 103, a configuration judgment module 104, and a tiered shooting module 105. Detailed descriptions of each functional module are as follows: The data acquisition module 101 is used to confirm the document type and quantity of the business order to be uploaded based on the pre-acquired business request data. The list construction module 102 is used to sort the document types according to preset rules, and construct a page-by-page shooting task for each document type according to the number of documents corresponding to each document type, and generate a shooting order list. The configuration acquisition module 103 is used to acquire the shooting configuration data stored by the user on the preset server. The configuration judgment module 104 is used to determine whether the user has enabled the AI ​​module based on the shooting configuration data. The tiered shooting module 105, if the user enables the AI ​​module, calls the preset embedded AI component in the target shooting device to shoot the business order according to the shooting order list and generate an image sequence. If the user does not enable the AI ​​module, it determines whether the document type is a preset document type. If the document type is a preset document type, it calls the preset Canvas dynamic screenshot component in the target shooting device to shoot the business order according to the shooting order list and generate an image sequence. If the document type is not a preset document type, it calls the native camera in the target shooting device to shoot the business order according to the shooting order list and generate an image sequence.

[0071] In one embodiment, when the data acquisition module 101 performs the step of confirming the document type and quantity of the business order to be uploaded based on the pre-acquired business request data, it is specifically used for: Obtain the service identifier from the service request data; Based on the business identifier, query business scenario information in a preset business database; Map the business scenario information to a business type; Load the document configuration rule table from the business database based on the business identifier; Based on the business type, query the standard page number corresponding to each document type in the document configuration rule table, and generate the number of documents.

[0072] In one embodiment, the list construction module 102, when performing the steps of sorting the document types according to preset rules and constructing a page-by-page shooting task for each document type based on the number of documents corresponding to each document type to generate a shooting order list, is specifically used for: The document types of each business order are sorted according to preset rules to form a sorted document type table. Add the number of documents for each document type to the sorted document type table to obtain the sorted document configuration mapping table. Based on the number of documents corresponding to each document type in the sorted document configuration mapping table, the number of shooting task entries corresponding to each document type is determined. Based on the current order of each document type and the corresponding number of shooting task entries in the sorted document configuration mapping table, a shooting order list is generated.

[0073] In one embodiment, the configuration acquisition module 103, when performing the acquisition of shooting configuration data stored by the user on a preset server, is specifically used for: Obtain the user's identity identifier and generate a configuration query statement based on the identity identifier; A configuration query request is initiated to the server according to the configuration query statement, and the requested data is obtained; The shooting configuration data contained in the request data is extracted by parsing the request data.

[0074] In one embodiment, the tiered shooting module 105, when executing the step of shooting the business order according to the shooting order list by calling the embedded AI component preset in the target shooting device to generate an image sequence, is specifically used for: The original images of each business order are obtained according to the shooting order to obtain the original image sequence; The embedded AI component is used to perform edge detection on each image in the original image sequence to generate edge detection results; Based on the edge detection results, each image in the original image sequence is processed to remove edges, resulting in an edge-removed image sequence. Each image in the edge-removed image sequence is subjected to brightness enhancement and contrast enhancement processing to obtain an enhanced image sequence; The embedded AI component is used to perform distortion correction processing on each image in the enhanced image sequence to obtain a corrected image sequence.

[0075] In one embodiment, the tiered shooting module 105, when executing the step of shooting the business order according to the shooting order list by calling the preset Canvas dynamic screenshot component in the target shooting device and generating an image sequence, is specifically used for: Obtain a screenshot image captured by the target imaging device; Extract the edge position information of the business order from the captured screenshot image; Obtain the reference frame position information of the preset reference frame in the captured screenshot image; Calculate the positional deviation between the edge of the business order and the reference frame based on the edge position information and the reference frame position information; Determine whether the positional deviation is greater than a preset deviation threshold; If the position deviation is greater than the deviation threshold, then a position adjustment prompt is generated based on the position deviation and displayed to the user based on the position deviation, and then the process returns to the step of obtaining the screenshot image captured by the target shooting device. If the position deviation is less than or equal to the deviation threshold, then the business order image in the captured screenshot image is extracted based on the edge position information; After the images of each business order are extracted, the image sequence is generated.

[0076] In one embodiment, the graded shooting module 105, when performing the extraction of the edge position information of the business order in the captured screenshot image, is specifically used for: Convert the captured screenshot image into a grayscale image; The grayscale image is subjected to Gaussian blur denoising processing to obtain a denoised image; Calculate the gradient magnitude of the denoised image to obtain a gradient magnitude image; Pixels in the gradient magnitude image with magnitudes higher than a threshold are marked as strong edges, generating an edge-marked image; Extract the largest rectangular contour from the edge-marked image to obtain the edge position information.

[0077] This invention provides a business document shooting control device to solve problems such as disordered order, poor image quality, and non-compliance of key documents when manually uploading documents in financial businesses (such as life insurance and bancassurance). The solution first automatically confirms the type of document to be uploaded and the standard number of pages for each document (e.g., 3 pages for an electronic insurance application, 6 pages for a disclaimer) based on the business request data. Then, it constructs a page-by-page shooting order list according to quality control requirements, binding a unique identifier to each page number. Next, it obtains the shooting configuration data stored by the user on the server and determines whether the AI ​​module is enabled. If AI is enabled, it calls the embedded AI component to perform real-time edge detection, edge removal, image enhancement, and perspective correction, directly outputting compliant images. If AI is not enabled, it further determines whether the current document is a preset key document (such as an electronic insurance application). For key documents, it activates the Canvas dynamic screenshot component, overlays a reference frame to guide the user to align and hard-crop, generating a standardized image with a fixed aspect ratio. For ordinary documents, it directly calls the system's native camera for shooting. All shooting strictly follows the shooting order list, ensuring that the images are arranged in page order, ultimately resulting in a complete image sequence. This solution eliminates missing pages and out-of-order issues through a sequential control mechanism, and balances compliance and user experience through a tiered shooting strategy, significantly improving the pass rate of document quality control and reducing labor costs. It is suitable for fields such as finance and healthcare that require rigorous processing of sequential documents. Specific limitations regarding the business order shooting control device can be found in the limitations of the business order shooting control method described above, and will not be repeated here. Each module in the aforementioned business order shooting control device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the corresponding operations of each module.

[0078] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 4 As shown, the computer device includes a processor, memory, network interface, and database connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile and / or volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The network interface is used to communicate with external clients via a network connection. When the computer program is executed by the processor, it implements the functions or steps of a single-shot control method on the server side.

[0079] In one embodiment, a computer device is provided, which may be a client, and its internal structure diagram may be as follows: Figure 5 As shown, the computer device includes a processor, memory, network interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The network interface is used to communicate with an external server via a network connection. When the computer program is executed by the processor, it implements the functions or steps on the client side of a business single-shot control method.

[0080] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to perform the following steps: Confirm the document type and quantity of the business order to be uploaded based on the pre-acquired business request data; The document types are sorted according to preset rules, and a page-by-page shooting task is constructed for each document type according to the number of documents corresponding to each document type, generating a shooting order list; Retrieve the shooting configuration data stored by the user on the preset server; Determine whether the user has enabled the AI ​​module based on the shooting configuration data; If the user enables the AI ​​module, the embedded AI component preset in the target shooting device is invoked to shoot the business order according to the shooting order list and generate an image sequence. If the user has not enabled the AI ​​module, then determine whether the document type is a preset document type; If the document type is a preset document type, then the preset Canvas dynamic screenshot component in the target shooting device is invoked to shoot the business order according to the shooting order list, and an image sequence is generated; If the document type is not a preset document type, the native camera in the target shooting device is invoked to shoot the business order according to the shooting order list, and an image sequence is generated.

[0081] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor: Confirm the document type and quantity of the business order to be uploaded based on the pre-acquired business request data; The document types are sorted according to preset rules, and a page-by-page shooting task is constructed for each document type according to the number of documents corresponding to each document type, generating a shooting order list; Retrieve the shooting configuration data stored by the user on the preset server; Determine whether the user has enabled the AI ​​module based on the shooting configuration data; If the user enables the AI ​​module, the embedded AI component preset in the target shooting device is invoked to shoot the business order according to the shooting order list and generate an image sequence. If the user has not enabled the AI ​​module, then determine whether the document type is a preset document type; If the document type is a preset document type, then the preset Canvas dynamic screenshot component in the target shooting device is invoked to shoot the business order according to the shooting order list, and an image sequence is generated; If the document type is not a preset document type, the native camera in the target shooting device is invoked to shoot the business order according to the shooting order list, and an image sequence is generated.

[0082] It should be noted that the functions or steps that can be implemented by the computer-readable storage medium or computer device described above can be referred to the relevant descriptions on the server side and client side in the foregoing method embodiments. To avoid repetition, they will not be described one by one here.

[0083] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0084] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0085] It should be noted that in the technical solutions disclosed in this invention, the acquisition of user information (personal image data (e.g., facial videos or pictures, facial feature videos or pictures, etc.) and personal privacy information (e.g., name, ID number, occupation, address, etc.)) is all completed with the user's knowledge and consent, and the acquisition of the relevant user information is legal and compliant.

[0086] Finally, it should be noted that any AI models, software tools, or components not belonging to this company appearing in the embodiments of this application are merely illustrative examples and do not represent actual use. All user personal information involved in the embodiments of this application is authorized (with knowledge and consent) by the relevant parties or fully authorized by all parties, and the executing entity can obtain it through various legal and compliant means. The collection, storage, use, processing, transmission, provision, and disclosure of the information, data, and signals involved all comply with relevant laws and regulations and do not violate public order and good morals. The above-described embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.< / canvas> < / video>

Claims

1. A business order shooting control method, characterized in that, include: Confirm the document type and quantity of the business order to be uploaded based on the pre-acquired business request data; The document types are sorted according to preset rules, and a page-by-page shooting task is constructed for each document type according to the number of documents corresponding to each document type, generating a shooting order list; Retrieve the shooting configuration data stored by the user on the preset server; Determine whether the user has enabled the AI ​​module based on the shooting configuration data; If the user enables the AI ​​module, the embedded AI component preset in the target shooting device is invoked to shoot the business order according to the shooting order list and generate an image sequence. If the user has not enabled the AI ​​module, then determine whether the document type is a preset document type; If the document type is a preset document type, then the preset Canvas dynamic screenshot component in the target shooting device is invoked to shoot the business order according to the shooting order list, and an image sequence is generated; If the document type is not a preset document type, the native camera in the target shooting device is invoked to shoot the business order according to the shooting order list, and an image sequence is generated.

2. The business order shooting control method as described in claim 1, characterized in that, The step of confirming the document type and quantity of the business order to be uploaded based on the pre-acquired business request data includes: Obtain the service identifier from the service request data; Based on the business identifier, query business scenario information in a preset business database; Map the business scenario information to a business type; Load the document configuration rule table from the business database based on the business identifier; Based on the business type, query the standard page number corresponding to each document type in the document configuration rule table, and generate the number of documents.

3. The business order shooting control method as described in claim 1, characterized in that, The step involves sorting the document types according to preset rules, constructing a page-by-page shooting task for each document type based on the number of documents corresponding to each document type, and generating a shooting order list, including: The document types of each business order are sorted according to preset rules to form a sorted document type table. Add the number of documents for each document type to the sorted document type table to obtain the sorted document configuration mapping table. Based on the number of documents corresponding to each document type in the sorted document configuration mapping table, the number of shooting task entries corresponding to each document type is determined. Based on the current order of each document type and the corresponding number of shooting task entries in the sorted document configuration mapping table, a shooting order list is generated.

4. The business order shooting control method as described in claim 1, characterized in that, The step of obtaining the shooting configuration data stored by the user on a preset server includes: Obtain the user's identity identifier and generate a configuration query statement based on the identity identifier; A configuration query request is initiated to the server according to the configuration query statement, and the requested data is obtained; The shooting configuration data contained in the request data is extracted by parsing the request data.

5. The business order shooting control method as described in claim 1, characterized in that, The step of capturing the business order according to the shooting order list by calling a preset embedded AI component in the target shooting device to generate an image sequence includes: The original images of each business order are obtained according to the shooting order to obtain the original image sequence; The embedded AI component is used to perform edge detection on each image in the original image sequence to generate edge detection results; Based on the edge detection results, each image in the original image sequence is processed to remove edges, resulting in an edge-removed image sequence. Each image in the edge-removed image sequence is subjected to brightness enhancement and contrast enhancement processing to obtain an enhanced image sequence; The embedded AI component is used to perform distortion correction processing on each image in the enhanced image sequence to obtain a corrected image sequence.

6. The business order shooting control method as described in claim 1, characterized in that, The step of capturing the business order by calling the preset Canvas dynamic screenshot component in the target shooting device according to the shooting order list and generating an image sequence includes: Obtain a screenshot image captured by the target imaging device; Extract the edge position information of the business order from the captured screenshot image; Obtain the reference frame position information of the preset reference frame in the captured screenshot image; Calculate the positional deviation between the edge of the business order and the reference frame based on the edge position information and the reference frame position information; Determine whether the positional deviation is greater than a preset deviation threshold; If the position deviation is greater than the deviation threshold, then a position adjustment prompt is generated based on the position deviation and displayed to the user based on the position deviation, and then the process returns to the step of obtaining the screenshot image captured by the target shooting device. If the position deviation is less than or equal to the deviation threshold, then the business order image in the captured screenshot image is extracted based on the edge position information; After the images of each business order are extracted, the image sequence is generated.

7. The business order shooting control method as described in claim 6, characterized in that, The step of extracting the edge position information of the business order in the captured screenshot image includes: Convert the captured screenshot image into a grayscale image; The grayscale image is subjected to Gaussian blur denoising processing to obtain a denoised image; Calculate the gradient magnitude of the denoised image to obtain a gradient magnitude image; Pixels in the gradient magnitude image with magnitudes higher than a threshold are marked as strong edges, generating an edge-marked image; Extract the largest rectangular contour from the edge-marked image to obtain the edge position information.

8. A business order shooting control device, characterized in that, include: The data acquisition module is used to determine the document type and quantity of the business order to be uploaded based on the pre-acquired business request data. The list building module is used to sort the document types according to preset rules, and build a page-by-page shooting task for each document type according to the number of documents corresponding to each document type, and generate a shooting order list. The configuration acquisition module is used to acquire the shooting configuration data stored by the user on the preset server. A configuration judgment module is used to determine whether the user has enabled the AI ​​module based on the shooting configuration data; The tiered shooting module, if the user enables the AI ​​module, calls the preset embedded AI component in the target shooting device to shoot the business order according to the shooting order list and generate an image sequence. If the user does not enable the AI ​​module, it determines whether the document type is a preset document type. If the document type is a preset document type, it calls the preset Canvas dynamic screenshot component in the target shooting device to shoot the business order according to the shooting order list and generate an image sequence. If the document type is not a preset document type, it calls the native camera in the target shooting device to shoot the business order according to the shooting order list and generate an image sequence.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the business single-shot control method as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the business single-shot control method as described in any one of claims 1 to 7.

Images