Certificate generation and delivery method and system for license service

Abstract

The application discloses a kind of certificate generation and delivery method and system for license service, belong to big data processing field, including the following steps: original photo and user identity mark are collected, and generate after processing certificate photo, original collection photo hash, after processing certificate photo hash, cutting rectangular coordinate and photo specification calculation value;After processing certificate photo is bound with user identity mark;Determine available cloud certificate photo;Certificate PDF and certificate PDF association data are generated;Collect and correct printed image;Double reference proofreading and photo visual quality feedback analysis are executed, and double reference proofreading result and photo visual quality feedback result are generated;Delivery instruction or rollback instruction is generated.The application establishes cloud photo source traceability and double reference proofreading mechanism, achieves the technical effect that PDF use photo specification is consistent, abnormal source can be distinguished and oriented rollback, solves the problem that photo generation deviation is inherited, amplified or wrongly attributed in the prior art.

Description

Technical Field

[0001] This invention relates to the field of big data processing technology, and in particular to a method and system for certificate generation and delivery for certificate-related business. Background Technology

[0002] With the continuous advancement of digital government affairs, digital education, and vocational skills certification, an increasing number of certificate processing procedures are shifting from traditional manual window processing, paper material submission, and offline centralized certificate production to online information collection, electronic data verification, automatic certificate generation, and self-service delivery. Applicants can submit their identity information and ID photos via mobile phone, self-service terminal, or business platform. The system automatically matches the certificate template based on the business type and generates the corresponding certificate file. In this process, ID photos typically undergo image processing such as cropping, background replacement, and brightness or contrast adjustment, and are then electronically assessed to ensure they meet the photo specifications for the corresponding certificate type.

[0003] In existing technologies, mobile terminals, self-service terminals, and dedicated printing equipment have been gradually incorporated into the processing of certificates and licenses. Existing mobile certificate processing systems collect identity information, upload document photos, call electronic certificate templates, and generate corresponding electronic certificate files via mobile phones, enabling online application, information submission, and result display for single certificate / license items. In some business scenarios, basic electronic certificate processing can be completed by combining functions such as identity verification, electronic signatures, and result downloads. Meanwhile, dedicated equipment for certificate / license output has been deployed in government service halls, campus service points, and training and assessment venues. Furthermore, some large self-service machines can integrate functions such as information entry, terminal operation, and printing output to complete on-site processing of specific types of certificates / licenses. In addition, existing government service centers typically operate AI electronic certificate processing, high-speed scanner data collection, and certificate printers independently, often using methods such as manually pasting photos, manually stamping, and printing certificates, with these three processes operating independently.

[0004] The above-mentioned technology has at least the following technical problems:

[0005] The certificate generation and delivery process typically begins by cropping, replacing the background, and adapting the specifications of the captured ID photo using image algorithms. The processed photo is then stored in the cloud. When issuing certificates later, the system directly calls the existing photo from the cloud, embeds it into the certificate template to generate a PDF (Portable Document Format) layout, and then prints it. After printing, the printed document is verified against the generated PDF or electronic version using methods such as QR code decoding, OCR (Optical Character Recognition) fields, and layout detection boxes. However, the front-end ID photo processing result becomes the benchmark for subsequent printing and verification. When the cloud photo deviates in head proportions, shoulder retention, background edges, or specification adaptation, this deviation is solidified in the certificate layout. If subsequent verification only checks whether the printed document matches the PDF, it's difficult to detect that the photo used in the PDF itself no longer meets the current certificate requirements. Further compliance verification makes it difficult to distinguish whether the anomaly originates from a deviation in cloud photo processing or a deviation in printing reproduction. Furthermore, even if the photo used in the PDF meets photo specifications electronically, the same printing device may still produce discrepancies during preheating and continuous printing. Slight fluctuations in output density and color grayscale can occur under different consumable conditions and in automatic calibration states. This can lead to issues such as low brightness or insufficient contrast in printed photos, especially for processed ID photos where facial brightness, contrast, or background edge color difference are near the acceptable threshold. Although the electronic image indicators meet the photo specifications, the lack of print reproduction margin can cause slight brightness loss or edge color changes after paper output, potentially causing the photo generation deviation to be inherited, amplified, or misattributed. This further makes it difficult for the system to determine whether the print abnormality originates from cloud-based photo processing deviations, insufficient acceptable margin in the electronic photo, or print reproduction fluctuations caused by the same printing device under different output states. Furthermore, since the three processes of manually pasting photos, manually stamping, and printing certificates are independent, the one-to-one correspondence between them is prone to errors during operation, leading to incorrect certificate-photo matching or process omissions. This further increases the error rate of certificate processing and reduces the system's automated processing efficiency. Summary of the Invention

[0006] To address the technical problems of inherited, amplified, or incorrectly attributed photo generation deviations in existing technologies, embodiments of the present invention provide a method and system for certificate generation and delivery in the field of certificate and document services. The technical solution is as follows:

[0007] On the one hand, a method for certificate generation and delivery for certificate-related business is provided, which includes:

[0008] The system receives the user's original facial image, user identification, current certificate type, and photo specification data; generates a processed certificate photo, original image hash, processed certificate photo hash, cropped rectangle coordinates, and calculated photo specification values; binds the processed certificate photo with the user identification to form a cloud-based certificate photo record; when processing a certificate, it retrieves the cloud-based certificate photo record and determines the available cloud-based certificate photo based on the photo specification data corresponding to the certificate type; embeds the available cloud-based certificate photo into the certificate template to generate a certificate PDF and associated data; collects and corrects the printed image to form a corrected printed image; performs a dual-benchmark verification and photo visual quality feedback analysis based on the certificate PDF, associated data, cloud-based certificate photo record, and corrected printed image, using photo source and photo specification as the first benchmark and print reproduction as the second benchmark, generating dual-benchmark verification results and photo visual quality feedback results; and generates a delivery instruction or rollback instruction based on the dual-benchmark verification results and photo visual quality feedback results.

[0009] On the other hand, a certificate generation and delivery system for certificate and license business is provided, the system including:

[0010] The ID photo specification generation module is used to receive the user's original face capture photo, user identity identifier, current ID photo type and photo specification data, and generate the processed ID photo, the hash of the original capture photo, the hash of the processed ID photo, the coordinates of the cropped rectangle and the calculated value of the photo specification.

[0011] The cloud-based photo archiving module is used to bind processed ID photos with user identity identifiers to form cloud-based ID photo records.

[0012] The photo matching module is used to retrieve cloud-based ID photo records when applying for a certificate, and determine the available cloud-based ID photos based on the photo specification data corresponding to the type of certificate to be processed.

[0013] The certificate template generation module is used to embed available cloud-based ID photos into the certificate template, generating a certificate PDF and associated data.

[0014] The print image correction module is used to acquire and correct the print image, forming a corrected print image.

[0015] The certificate photo verification module is used to perform dual-benchmark verification and photo visual quality feedback analysis based on the certificate PDF, certificate PDF associated data, cloud-based ID photo records, and corrected printed images. The verification is based on the photo source and photo specifications as the first benchmark and the print reproduction as the second benchmark. The module generates dual-benchmark verification results and photo visual quality feedback results.

[0016] The delivery rollback control module is used to generate delivery instructions or rollback instructions based on the dual-benchmark calibration results and the visual quality feedback results of the photos.

[0017] The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following:

[0018] 1. The certificate generation and delivery method for certificate-related business provided by this invention saves the original captured photo hash, the processed certificate photo hash, the cropping rectangle coordinates, and the photo specification calculation value when the photo is entered into the cloud database. Before issuing the certificate, it re-matches or regenerates usable certificate photos according to the current certificate type, so that the photos called in the certificate PDF have traceable source and verifiable standard basis, thereby realizing the consistency confirmation between the photo used in the PDF and the current certificate conditions, effectively solving the problem in the prior art that it is impossible to determine whether the photo called in the PDF still meets the current certificate conditions.

[0019] 2. This invention employs a dual-benchmark verification system, using both photo source and photo specification benchmarks, and a print reproduction benchmark, to differentiate between cloud-based photo processing deviations, PDF embedding anomalies, print discrepancies, and image capture anomalies. This enables precise rollback, reprocessing, or reprinting of certificate printing errors, effectively solving the problem in existing technologies where cloud-based photo processing deviations and print reproduction deviations are difficult to distinguish and prone to error attribution. Furthermore, this invention unifies the existing independently operating AI electronic certificate processing, high-speed scanner acquisition, and certificate printing processes in government service centers into a single system, effectively reducing the overall error rate and improving the automation level and processing efficiency of certificate delivery. Simultaneously, this system has no special hardware requirements; it requires no dedicated acquisition equipment and can run on ordinary devices, lowering the system deployment threshold. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying 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.

[0021] Figure 1 A flowchart illustrating a certificate generation and delivery method for certificate-related services provided in this application embodiment;

[0022] Figure 2 A diagram illustrating the dual-reference calibration mechanism provided in this application embodiment;

[0023] Figure 3 A schematic diagram of the structure of a certificate generation and delivery system for certificate-related services provided in an embodiment of this application;

[0024] Figure 4 A hierarchical rollback and anti-dead-loop closed-loop diagram provided for embodiments of this application;

[0025] Figure 5 This is an attribution diagram of visual quality anomalies provided in an embodiment of this application. Detailed Implementation

[0026] Embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of the present disclosure are shown in the drawings, it should be understood that embodiments of the present disclosure may be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of the present disclosure.

[0027] It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure. In the description of the embodiments of this disclosure, the term "comprising" and similar terms should be understood as open-ended inclusion, i.e., "including but not limited to". The term "based on" should be understood as "at least partially based on". The term "one embodiment" or "this embodiment" should be understood as "at least one embodiment". The terms "first", "second", etc., may refer to different or the same objects.

[0028] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0029] like Figure 1 The diagram shown is a flowchart of a certificate generation and delivery method for certificate-related services provided in this application embodiment. The method includes the following steps:

[0030] S1: Collect the user's original facial image and process the ID photo using algorithms. The system receives the user's original facial image, user identification, current ID type, and corresponding photo specifications. The original facial image is obtained by capturing the user's face using a high-speed scanner. During the capture process, the user should be reminded to pay attention to their appearance and meet ID photo specifications. In one specific embodiment, the acquisition unit integrates AI ID photo processing functionality, supporting fully automatic camera recognition and plug-and-play functionality for one-click ID photo creation. It also includes built-in standard ID photo sizes such as 1-inch, small 1-inch, 2-inch, and small 2-inch, and supports custom sizes. It features GPU-accelerated real-time preview and an intelligent viewfinder, providing a WYSIWYG experience. This data acquisition system has no special hardware requirements and can run on a standard device with a minimum of an i3 processor, 4GB of RAM, and an integrated graphics card. An i5 processor, 8GB of RAM, and a GTX 1050 or higher are preferred. The AI ​​image matting processing speed is 2 to 5 seconds per image. It supports Windows 7 / 10 / 11 64-bit systems, uses less than 500MB of RAM, and the installation includes approximately 2GB of model files. It can run completely offline. The system architecture uses the Qt 5.15.2 graphical interface framework, the OpenCV 4.12 image processing library, and the ONNX Runtime local offline inference engine. Its modular design separates camera control, image processing, and interface display. It features complete operation logs and exception handling mechanisms, including handling situations such as camera disconnection and model loading failure. User identity identifiers are used to distinguish different applicants for certificates. These identifiers include one or more of the following: ID card number, business system account, mobile phone number, WeChat real-name authorization identifier, Alipay real-name authorization identifier, facial recognition result identifier, and mailing address or residential address. When storing and retrieving data later, the system generates a unique internal user index code based on the above identity elements, rather than directly using the ID card number, mobile phone number, or third-party account plaintext as the photo retrieval key. User identity authentication can come from ID card real-name authentication, facial recognition, business system login status, WeChat real-name authorization, or other real-name identity elements recognized by the business party. The current certificate type and the corresponding photo specification data are provided by the relevant business authority.

[0031] The system first performs face detection on the user's original facial image and extracts the head vertex, chin point, minimum shoulder preservation boundary, and face region. The head vertex, chin point, and minimum shoulder preservation boundary can be obtained using existing facial keypoint detection, human contour detection, and foreground segmentation algorithms. Specifically, the head vertex is the uppermost boundary point of the person's head contour in the vertical coordinate of the image, the chin point is the lower boundary point of the face's chin contour in the vertical direction, and the minimum shoulder preservation boundary is the lower boundary of the person's shoulder region that needs to be preserved in the ID photo. If the number of detected faces is not 1, or if any of the head vertex, chin point, or minimum shoulder preservation boundary cannot be extracted, the system generates a re-acquisition instruction.

[0032] If all the above-mentioned points can be extracted, the system calculates the facial brightness, contrast, and sharpness values ​​of the user's original facial image. The facial brightness value is calculated based on the brightness of the facial region. For example, the pixels of the facial region can be converted into luminance components, which are weighted according to RGB color values. After removing the top 5% and bottom 5% of pixels, the system takes the average brightness of the remaining pixels as the facial brightness value. The facial contrast value is represented by the difference between the 90th and 10th percentile values ​​of the facial region's brightness distribution, reflecting the differences in the layers of facial features and contours. The facial sharpness value is preferably calculated using the Laplacian response variance of the grayscale image of the facial region. The lower the Laplacian response variance, the more blurred the edges, and the more likely the subject is to be out of focus or have unclear imaging. To ensure direct comparison of calculation results across different steps, the system first performs color space and bit depth unification on the images when calculating the brightness and contrast values ​​of the user's original face capture photo and the processed ID photo, and then normalizes the brightness components to 0 to 1. Specifically, the face brightness value is represented by a normalized average brightness within the range of 0 to 1, and the face contrast value is represented by a normalized brightness difference within the range of 0 to 1. The face sharpness value is calculated using the Laplacian response variance after unifying the image to 8-bit grayscale, and the variance value at this unified grayscale scale is used as the sharpness indicator. The head height ratio, top white space ratio, and shoulder retention ratio are all dimensionless ratios.

[0033] If the facial clarity value of the user's original face capture photo is lower than the facial clarity threshold corresponding to the current ID photo type, the system determines that the original face capture photo has a problem with out-of-focus or unclear image of the person, and generates a recapture command. If the facial brightness value or facial contrast value of the user's original face capture photo is lower than the facial brightness threshold corresponding to the current ID photo type, but can be adjusted to meet the requirements within the preset adjustment range through brightness compensation or contrast adjustment, the system performs controlled image quality processing. Controlled image quality processing is only used to adjust the brightness and contrast of the facial area and the overall photo, and does not change the geometric proportions of the person, the top of the head, the chin point, the lowest preserved boundary of the shoulders, the background color, nor does it beautify, distort, or replace the facial features. If the facial brightness value or facial contrast value of the user's original face capture photo is lower than the corresponding threshold and exceeds the preset adjustment range, the system generates a recapture command. The preset adjustment range refers to the maximum compensation range allowed for the normalized face brightness value and normalized face contrast value without changing the geometric proportions of the person, the top of the head, the chin point, the minimum preserved boundary of the shoulders, the background color, and the shape of the facial features. This maximum compensation range is pre-written into the photo specification configuration table, and it is preferable to set the upper limit of brightness compensation and the upper limit of contrast adjustment respectively.

[0034] The aforementioned face brightness threshold, face contrast threshold, and face clarity threshold need to be preset by technical personnel based on actual circumstances. If the relevant competent authority or certificate regulations have already provided corresponding requirements, those requirements should be directly adopted; otherwise, technical personnel should pre-calibrate them based on historical manually reviewed and approved photos and print samples. For example, a certain number of manually reviewed and approved ID photo samples can be selected, their face brightness, face contrast, and face clarity values ​​can be calculated, and the range of values ​​that can reliably pass the review can be written into the photo specification configuration table. For the same printing equipment, a certain print reproducibility margin can also be set for this threshold based on post-printing feedback.

[0035] Subsequently, the system determines the coordinates of the cropping rectangle based on the photo specifications corresponding to the current certificate type, ensuring that the cropping rectangle simultaneously includes the head apex, chin point, and the lowest retained shoulder boundary. The cropping rectangle coordinates represent its position within the user's original face capture image, including the x-coordinate and y-coordinate of the top-left corner, the rectangle's width, and its height. Using these coordinates, the system can subsequently determine the regional correspondence between the processed certificate photo and the user's original face capture image, providing a locational basis for subsequent photo source verification, re-cropping, and certificate PDF data verification. The cropping rectangle is determined to meet the requirements for head height ratio, top margin ratio, and shoulder retention ratio. After determining the cropping rectangle, the system scales the image within it proportionally to the photo size required for the current certificate type and replaces the background with the required background color.

[0036] After generating the processed ID photo, the system maps the head vertex, chin point, and lowest shoulder retention boundaries onto the processed ID photo according to the coordinate transformation relationship of the cropping rectangle. It then calculates the head height ratio, top margin ratio, and shoulder retention ratio to determine whether the face size, top margin, and shoulder retention in the processed ID photo meet the layout requirements of the current ID photo type, avoiding issues such as an oversized head, insufficient top margin, cropped shoulders, or the person filling the photo area. Specifically, the head height ratio is the ratio of the vertical distance between the head vertex and chin point to the height of the processed ID photo. The top margin ratio is the ratio of the vertical distance between the top edge of the processed ID photo and the head vertex to the height of the processed ID photo. The shoulder retention ratio is the ratio of the vertical distance between the lowest shoulder retention boundary and the bottom edge of the processed ID photo to the height of the processed ID photo. Simultaneously, the system generates a background edge band of a preset width outside the foreground outline of the person. The background edge band is preferably located within 3 to 10 pixels outside the outline of the person. The specific width can be preset in the photo specification configuration table according to the photo resolution and certificate type. The system converts the pixels within the background edge band to the CIE Lab color space and calculates the color difference value between each pixel and the background color. It is preferred to use the CIEDE2000 color difference formula to calculate the color difference value between each pixel and the background color, and take the 95th percentile value of the color difference value within the background edge band as the background edge color difference value, so as to reduce the influence of a small number of isolated noise points on the judgment result, and at the same time, it can reflect whether there are residual, rough edges, gray edges or uneven colors in most background edge areas.

[0037] The system also calculates the facial brightness, contrast, and sharpness values ​​of the processed ID photo, using the same calculation rules as the corresponding indicators in the user's original facial image. The resulting photo specification calculation values ​​include head height percentage, top white space percentage, shoulder retention percentage, background edge color difference, facial brightness, contrast, and sharpness. If all the above photo specification calculation values ​​meet the photo specification data corresponding to the current ID photo type—that is, the head height percentage, top white space percentage, and shoulder retention percentage all fall within their corresponding ranges, the background edge color difference value is not greater than the corresponding background edge color difference threshold, and the facial brightness, contrast, and sharpness values ​​are not lower than their corresponding thresholds—then the system generates the processed ID photo. If the head height percentage, top white space percentage, or shoulder retention percentage does not meet the photo specification data—that is, any one of these percentages is lower than the lower limit or higher than the upper limit of its corresponding range—and the cropping rectangle still does not exceed the boundaries of the user's original facial image after adjustment, the system redetermines the cropping rectangle and recalculates the photo specification calculation values. If adjusting the cropping rectangle causes it to exceed the boundaries of the user's original face capture image, the system will generate a recapture command.

[0038] If the background edge color difference value does not meet the photo specification data, i.e., the background edge color difference value is greater than the background edge color difference threshold corresponding to the current certificate type, the system will re-execute the background replacement or edge blending process and recalculate the background edge color difference value. If the background edge color difference threshold still cannot be met after reprocessing up to the maximum number of times, the system will generate a re-collection instruction or a manual review instruction. The above-mentioned maximum number of times can be preset by technicians based on the certificate type, photo resolution, background replacement algorithm, and historical manual review samples. Preferably, the same background edge anomaly type will be automatically processed two or three times in the same certificate task.

[0039] After the above operations are completed, the system calculates the original captured photo hash and the processed ID photo hash for photos that meet the photo specification data and thresholds. The original captured photo hash is a unique image identifier generated based on the standardized pixel data of the user's original captured face photo. The processed ID photo hash is a unique image identifier generated based on the standardized pixel data of the processed ID photo after cropping, scaling, background replacement, and specification adaptation. When generating the hash, the system first performs orientation correction, color space unification, bit depth unification, and metadata removal on the image to be calculated. Then, it reads the image pixel data in a left-to-right, top-to-bottom order and assembles the image width, image height, color space identifier, bit depth identifier, and pixel data into hash data in a preset order. Subsequently, the SHA-256 algorithm is used to calculate the hash data to obtain the corresponding hash value. The original captured photo hash is used to verify whether the user's original captured face photo retrieved later is consistent with the original source when it was entered into the database. The processed ID photo hash is used to verify the consistency between the cloud ID photo record, the certificate PDF associated data, and the actual embedded photo in the PDF. When a photo meets the photo specification data, the system outputs the processed ID photo, the hash of the original captured photo, the hash of the processed ID photo, the coordinates of the cropped rectangle, the calculated value of the photo specification, the photo processing record, the user's identity identifier, and the current ID photo type. The photo processing record includes whether brightness compensation was performed, whether contrast adjustment was performed, whether the user's original face captured photo passed the sharpness judgment, whether the processed ID photo passed the photo specification judgment, the brightness compensation range, the contrast adjustment range, and the reason for processing.

[0040] S2: Store the processed ID photo and its source record in the cloud. The system receives the processed ID photo, the hash of the original captured photo, the hash of the processed ID photo, the coordinates of the cropped rectangle, the calculated value of the photo specifications, the photo processing record, the user's identity identifier, and the current ID photo type output by S1.

[0041] First, the system confirms the user's identity authentication result, which refers to the authentication status and authentication source record generated after the business system completes the authentication of the user's identity identifier. This includes whether the authentication was successful, the authentication method, the authentication time, and the authentication source identifier. Authentication methods include ID card real-name authentication, facial recognition, business system login status, WeChat real-name authorization, Alipay real-name authorization, or other real-name identity elements recognized by the business party. When the user's identity authentication result is successful, a unique user index code is generated. To avoid directly using ID card numbers, mobile phone numbers, or third-party account plaintext as the primary key for photo retrieval, the system generates an internal unique user index code based on the authenticated identity elements and binds this unique user index code to the processed ID photo. Next, the processed ID photo is bound and stored with the unique user index code, along with the original captured photo hash, the processed ID photo hash, the cropped rectangle coordinates, the photo specification calculation value, the photo processing record, the current ID photo type, the photo record status, and the photo version number. The cloud-based ID photo record includes currently valid records and historical version records. If the processed ID photo needs to be re-cropped or regenerated later, the system can also store the encrypted storage address of the user's original face photo and verify whether the original face photo of the called user is consistent with the source when it was entered into the database through the hash of the original photo.

[0042] When a user already has a processed ID photo of the same document type, the system compares the hash of the newly generated processed ID photo with the hash of the processed ID photo in the existing record. If they match, it means the content of the old and new processed ID photos is consistent, and the system retains the existing record. If they do not match, the system displays the old and new photos and their corresponding photo specification calculation values ​​to the user or the business processing terminal, allowing the user or the business processing terminal to confirm whether to retain the existing record or update to the new record. After the user confirms the update, the system sets the new photo as the current valid version and sets the original record as a historical version. If the business scenario requires automatic updates, the system can also automatically generate a new version after confirming that the new photo meets the photo specification data, but the historical record and the reason for the update are still retained.

[0043] Regarding the security of document retrieval, the system employs a three-layer retrieval mechanism. The first layer is retrieval after identity authentication; only after successful user authentication will the system allow retrieval of cloud-based document photo records using the user's unique index code. The second layer is a combined conditional retrieval; the system must simultaneously match the user's unique index code, the type of certificate to be processed, the photo record status, and the photo specification matching result to avoid erroneous retrieval due to users with the same name, old versions of photos, or obsolete photos. The photo specification matching result refers to the result formed by the system comparing the calculated photo specification value in the cloud-based document photo record with the photo specification data corresponding to the type of certificate to be processed, including four categories: successful match, unsuccessful match, need to be reprocessed according to the current certificate type, and inapplicable photo specification version. The system's parameters are categorized into three layers: A successful match indicates that the photo size, background color, head height percentage, top white space percentage, shoulder space percentage, background edge color difference, face brightness, face contrast, and face clarity all conform to the photo specifications corresponding to the certificate type being processed. A failed match indicates that at least one of the above calculated values ​​does not meet the photo specifications and needs to be reprocessed according to the current certificate type. A failed match also indicates that while there is a usable original photo source, the current processed ID photo is not suitable for the current certificate type, or the photo specification version is inapplicable. Furthermore, a failed match indicates that the photo specification configuration version corresponding to the cloud-based ID photo record is earlier than the currently valid version, requiring re-verification or reprocessing. The third layer is pre-call verification. Even if a candidate photo is retrieved, the system still checks whether the photo is the currently valid version, whether the processed ID photo hash matches the cloud record, whether the calculated photo specification value meets the current certificate requirements, and whether there is a photo version that is being updated but not yet confirmed. Through these operations, the system can reduce erroneous retrieval, duplicate retrieval, and certificate mismatches.

[0044] S3: When a user applies for a certificate, the system receives the user's identity identifier, the type of certificate to be processed, the photo specification configuration table, and the cloud-based ID photo record database. In this step, the user's identity identifier is used to query the cloud-based ID photo record database, which is the source of the retrieved data.

[0045] The system retrieves candidate photos corresponding to the user's unique index code from the cloud-based ID photo record database. Then, based on the type of certificate to be processed, the system reads the corresponding photo specification data from the photo specification configuration table. In the photo specification configuration table, each certificate type corresponds to a set of photo specification data, with an effective time and version number set, so that the configuration can be updated without modifying the entire process when certificate photo requirements change. Specifically, the system uses the certificate type as an index and photo size, background color, head height percentage range, top white space percentage range, shoulder retention percentage range, background edge color difference threshold, face brightness threshold, face contrast threshold, and face clarity threshold as specification items, and writes corresponding values, units, judgment methods, and source identifiers for each specification item. The judgment methods include range judgment, upper limit judgment, and lower limit judgment; the source identifier is used to distinguish whether the specification item originates from the competent business department, certificate template rules, historical qualified sample calibration, or print feedback correction. When processing a certificate, the system first searches for the corresponding specification record based on the type of certificate to be processed. Then, it compares the calculated photo specification values ​​of the candidate photos item by item according to the judgment method of each specification item to determine whether the candidate photo meets the photo requirements corresponding to the current certificate type. That is, whether it meets the photo size, background color, head height ratio range, top white space ratio range, shoulder retention ratio range, background edge color difference threshold, face brightness threshold, face contrast threshold, and face clarity threshold obtained by mapping the photo specification configuration table for the type of certificate to be processed.

[0046] The system compares the candidate photo's certificate type, photo size, background color, head height percentage, top white space percentage, shoulder retention percentage, background edge color difference, face brightness, face contrast, and face clarity with the photo specifications corresponding to the certificate type to be processed. If all the above data of the candidate photo meets the photo specifications corresponding to the certificate type to be processed, the candidate photo is selected as the usable cloud-based ID photo. If there is no candidate photo that meets the current photo specifications, but there is a record of the original user face capture photo source for the same user, and the system can call the original user face capture photo through this source record, the system first verifies whether the original user face capture photo being called is consistent with the original source in the database using the original capture photo hash. After the verification is consistent, the system returns S1 and regenerates the processed ID photo according to the certificate type to be processed. If there is no candidate photo that meets the current photo specifications, and the existing original user face capture photo source is unavailable, or the existing source cannot regenerate an ID photo that meets the current photo specifications, the system generates a re-capture instruction.

[0047] S4: The system receives available cloud-based ID photos, cloud-based ID photo records, certificate template data, and certificate business field data. The certificate business field data is generated by the certificate business processing system and includes the certificate number, certificate holder's name, identity information, certificate type, issuance date, validity period, business review result, issuing authority information, and QR code content.

[0048] The system writes the certificate business field data into the certificate template data and determines the photo area coordinates, photo area size, and template version according to the certificate template configuration table. Before generating the certificate PDF, a compatibility check is performed on the available cloud-based ID photos and the certificate template photo area. This compatibility check includes verifying the existence of the photo area coordinates, whether the photo area size matches the template version, whether the photo aspect ratio can be fully embedded in the photo area without cropping or stretching, whether the photo resolution meets the template output requirements, whether the photo display boundaries fall within the photo area range, whether the business fields can be fully written into their corresponding field areas, and whether the QR code area overlaps with the photo area or business field area. The template output requirements refer to the minimum requirements for photo display quality when generating the PDF and printing, including the output pixel size of the photo within the template photo area, the printing resolution, the physical size of the photo area, and the minimum image resolution allowed when generating the PDF. If the number of effective pixels in the processed ID photo embedded in the template photo area is lower than the template output requirements, the printed photo may appear blurry or lack detail, therefore, direct generation of the certificate PDF is not allowed. The photo display boundary refers to the rectangular boundary actually occupied in the certificate page coordinate system after the cloud-based ID photo is embedded into the certificate template in a proportional manner. The photo display boundary falling within the photo area means that the left, right, top, and bottom boundaries of this rectangle are all within the preset photo area of ​​the certificate template and do not intrude into the areas of the name, certificate number, QR code, or other business fields.

[0049] If the available cloud-based ID photo cannot be embedded into the certificate template photo area without cropping or stretching, or if the business fields cannot be completely written into the corresponding area, the system will not generate a certificate PDF. Instead, it will generate a photo re-matching instruction to return to S3. Upon receiving this instruction, S3 will re-search for candidate photos. If no candidate photos are available but the original user face photo is available, S3 will generate a reprocessing instruction to return to S1. If the photo and template compatibility check passes, the system will embed the available cloud-based ID photo into the certificate template photo area in a proportional manner. Subsequently, the system will generate a certificate PDF and write the processed ID photo hash, the original photo hash, the cropping rectangle coordinates, the photo specification calculation value, and the photo processing record into the certificate PDF's associated data. Thus, the certificate PDF not only contains the content displayed on the page but also carries the photo source and photo specification verification basis.

[0050] S5: Acquire the printed image and determine if it is verifiable. The system receives the certificate PDF, print-complete marker, and certificate template data. The DZ100 color laser certificate printer receives and prints the PDF file, and generates a print-complete marker after the certificate printing is complete. After the print-complete marker is generated, the Xinliangtian document scanner acquires the printed image. The system locates the four corners of the certificate page or the fixed edge line of the template in the printed image and calculates the sharpness value of the printed image. The sharpness value of the printed image is preferably calculated from the fixed edge line of the page, the text area, and the QR code area. Specifically, the system first extracts the area containing the fixed edge line of the page, the area containing the text field, and the area containing the QR code from the printed image, converts these areas to grayscale images, and then performs Laplacian edge response calculation on the grayscale images to obtain the edge response value of each area. Subsequently, the variance of the edge response values ​​of the above areas is calculated, and the lowest value of the variance values ​​of the three areas is taken as the sharpness value of the printed image. If the image is clear, the grayscale changes of page edges, text strokes, and QR code edges are more obvious, resulting in a higher Laplace response variance. If the image is blurry, these edge changes are weakened, leading to a lower Laplace response variance. If the four corners of the page or the fixed edges of the template cannot be located, it indicates that the captured image has problems such as occlusion, excessive tilt, or page incompleteness. If the sharpness value is less than the print image sharpness threshold, it indicates that the captured image is blurry and unsuitable for subsequent automatic verification. The print image sharpness threshold is pre-calibrated by the system during deployment based on the print image capture device, certificate template, and historical qualified capture samples. For example, the system can capture a batch of manually confirmed clear, complete, and unobstructed print images, calculate the sharpness values ​​of their fixed page edges, text areas, and QR code areas, and write the lower limit of sharpness that can stably pass verification into the print quality threshold table.

[0051] If the four corners of the page or the fixed edges of the template cannot be located, or if the sharpness value is less than the sharpness threshold of the printed image, the system generates a command to re-acquire the printed image. If the four corners of the page or the fixed edges of the template can be located, and the sharpness value is greater than or equal to the sharpness threshold of the printed image, the system performs perspective correction on the printed image based on the four corners of the page or the fixed edges of the template to obtain a corrected printed image.

[0052] S6: Performs dual-benchmark verification on certificate PDFs, cloud-based ID photo records, and printed copies, and integrates visual quality verification and anomaly attribution for printed photos. For example... Figure 2The diagram shown illustrates the dual-reference calibration mechanism provided in this application embodiment. The system receives the certificate PDF, certificate PDF associated data, corrected printed image, cloud-based ID photo records, certificate template data, photo specification data corresponding to the certificate type to be processed, photo quality threshold data, print quality threshold data, photo processing records, and historical adjustment records. The print quality threshold data can be stored in the form of a print quality threshold table, including print reproducibility thresholds, allowable brightness reduction, allowable contrast reduction, and allowable increase in background edge color difference. The print reproducibility threshold is used to determine whether there is a positional or dimensional deviation between the printed photo area and the PDF photo area; the allowable brightness reduction is used to determine whether the reduction in facial brightness in the printed photo area compared to the PDF photo area exceeds the allowable range; the allowable contrast reduction is used to determine whether the reduction in facial contrast in the printed photo area compared to the PDF photo area exceeds the allowable range; and the allowable increase in background edge color difference is used to determine whether the increase in background edge color difference in the printed photo area compared to the PDF photo area exceeds the allowable range. The photo quality threshold data is used to determine whether the embedded photo in the certificate PDF or the printed photo area itself meets the photo quality requirements, including at least facial brightness thresholds, facial contrast thresholds, and background edge color difference thresholds. The aforementioned thresholds were pre-calibrated by technicians using the same printing equipment, the same certificate template, and historical samples of qualified prints. Additionally, it should be noted that the photo quality threshold data can reference the face brightness threshold, face contrast threshold, background edge color difference threshold, and face sharpness threshold from the photo specification data. Alternatively, stricter internal calibration thresholds can be set without relaxing the photo specification data.

[0053] First, the system performs photo source and photo specification benchmark verification. The system reads the processed ID photo hash from the certificate PDF's associated data and also reads the processed ID photo hash from the cloud ID photo record. Simultaneously, the system reads the embedded photo from the certificate PDF and processes it according to the same image normalization rules as the processed ID photo in S1, including color space unification, size confirmation, bit depth unification, and pixel arrangement, and calculates the PDF embedded photo hash. Then, the system compares the PDF embedded photo hash, the processed ID photo hash from the certificate PDF's associated data, and the processed ID photo hash from the cloud ID photo record.

[0054] If all three (photo source, image source, and data source) match, the system generates a photo source consistency conclusion. If not all three match, the system generates a photo source inconsistency conclusion and records the inconsistency type. If the hash of the photo embedded in the PDF matches the hash of the processed ID photo in the certificate PDF association data, but they do not match the hash of the processed ID photo in the cloud ID photo record, the photo used in the PDF is determined to be inconsistent with the currently valid record in the cloud. If the hash of the processed ID photo in the certificate PDF association data matches the hash of the processed ID photo in the cloud ID photo record, but the hash of the photo embedded in the PDF does not match, the actual photo embedded in the PDF is determined to be abnormal. If the hash of the photo embedded in the PDF matches the hash of the processed ID photo in the cloud ID photo record, but the hash of the processed ID photo in the certificate PDF association data does not match, the PDF association data writing is determined to be abnormal. If all three are inconsistent, the photo retrieval, PDF binding, or data record is determined to be seriously inconsistent. Even if only two of the three match, it still constitutes a photo source inconsistency and cannot be used as a basis for delivery.

[0055] Assuming the photo source is consistent, the system reads the cropping rectangle coordinates, calculated photo specifications, and photo specification configuration version number from the certificate PDF's associated data, and compares the calculated photo specifications with the currently valid photo specification data for the certificate type to be processed. If any of the calculated photo specifications—head height percentage, top white space percentage, shoulder retention percentage, background edge color difference, face brightness, face contrast, or face clarity—does not meet the photo specification data corresponding to the certificate type to be processed, the system generates a conclusion that the PDF photo specifications are inconsistent. If all specifications meet the requirements, the system generates a conclusion that the PDF photo specifications are consistent.

[0056] Secondly, the system performs a print layout reproduction benchmark check. The system converts the certificate PDF into a PDF benchmark image at the same resolution as the corrected print image, and determines the photo area rectangle in the PDF benchmark image based on the certificate template data. The PDF benchmark image refers to the electronic layout image obtained by rasterizing the certificate PDF at a preset resolution, used as the standard reference image for print image verification. By placing the PDF benchmark image and the corrected print image at the same resolution and in the same page coordinate system, the system can compare whether the print faithfully reproduces the PDF layout. Specifically, during the conversion, the system reads the pixel width and pixel height of the corrected print image and calculates the target rendering resolution according to the physical page size of the certificate template; then, it rasterizes the certificate PDF page into a bitmap image at this target rendering resolution, ensuring that the generated PDF benchmark image and the corrected print image have the same page pixel size. If the corrected print image has already been adjusted to the preset page size of the template, the system directly rasterizes the certificate PDF according to that preset page size. The system locates the photo area rectangle in the corrected printed image by recording the top-left corner x-coordinate, top-left corner y-coordinate, width, and height of both the PDF and printed photo area rectangles. It then compares the PDF reference image and the corrected printed image in the same coordinate system to determine if there are any horizontal, vertical, width, or height differences. Horizontal offset is the difference between the top-left corner x-coordinates of the two photo areas; vertical offset is the difference between the top-left corner y-coordinates; width difference is the difference between the widths of the two photo areas; and height difference is the difference between the heights of the two photo areas. If the horizontal offset is not greater than a horizontal offset threshold, the vertical offset is not greater than a vertical offset threshold, the width difference is not greater than a width difference threshold, and the height difference is not greater than a height difference threshold, the system generates a "printed layout reproduction qualified" conclusion. If any of these values ​​exceeds their respective print reproduction thresholds, the system generates a "printed layout reproduction deviation" conclusion. The print reproducibility threshold is preset by technicians, but preferably pre-calibrated by the system based on certificate template size, printing equipment accuracy, and historical qualified print samples. During the current task comparison, it is then converted into corresponding pixel thresholds based on the target rendering resolution, including horizontal offset thresholds, vertical offset thresholds, width difference thresholds, and height difference thresholds. For example, the system can collect a certain number of manually verified qualified print images, calculate the horizontal offset, vertical offset, width difference, and height difference of their photo area relative to the PDF photo area, and write the offset range that can reliably pass manual review into the print quality threshold table.

[0057] Next, the system performs visual quality feedback analysis on the printed photos. The system extracts the face region from the photo area of ​​the PDF reference image and the photo area of ​​the corrected printed image, respectively, and calculates the face brightness and contrast values ​​for both the PDF and printed photo areas. The face brightness values ​​in the PDF and printed photo areas correspond to the face brightness values ​​in S1, and are calculated using the representative average of the face region's brightness components. Similarly, the face contrast values ​​in the PDF and printed photo areas correspond to the face contrast values ​​in S1, and are calculated using the difference between the higher and lower representative brightness values ​​within the face region. Face sharpness has already been used in S1 as a criterion for judging whether the user's original face image is out of focus; therefore, the system will not return to S1 for reprocessing due to face sharpness issues during the post-printing proofreading stage. The system also extracts the background edge band from the outside of the figure outline in the PDF photo area and the printed photo area, and calculates the background edge color difference value according to the calculation rules of the background edge color difference value in S1, which is used to determine whether the gray edge, jagged edge, afterimage or uneven background of the figure is magnified after printing.

[0058] If the same visual quality anomaly exists in both the PDF photo area and the printed photo area, the system generates a conclusion indicating a visual quality anomaly in the PDF photo. This conclusion records that the anomaly is not unique to the printed photo but already exists in the PDF photo and continues to be reproduced during the printing process. The conclusion includes the anomaly type, location of the anomaly, corresponding index values ​​for the PDF photo area and the printed photo area, corresponding photo quality thresholds, and suggested fallback paths. If the visual quality anomaly is insufficient brightness or contrast, the system generates a controlled image quality adjustment suggestion to return to S1. If the PDF photo area meets the photo quality threshold while the printed photo area does not, the system further determines whether the anomaly is due to a printing output anomaly or a printed photo acquisition anomaly. When the page borders, text areas, and QR code areas in the corrected printed photo simultaneously exhibit blurriness, darkening, or exposure anomalies, it indicates that the anomaly may be caused by the quality of the printed photo image, and the system generates a suggestion to re-acquire the printed photo image. In this context, "print output anomaly" indicates that while the overall captured image is usable for proofreading, the facial brightness, facial contrast, or background edge color difference values ​​in the printed image area exceed the permissible range of the print quality threshold data compared to the PDF image area. Specifically, if the facial brightness or contrast value in the printed image area is below the facial brightness threshold, or the background edge color difference value is above the background edge color difference threshold, the system determines it as an abnormal print output. "Print acquisition anomaly" indicates that the paper print itself may not have quality issues, but the currently captured image is unreliable for proofreading due to blurriness, darkness, overexposure, obstruction, or unclear page boundaries.

[0059] Finally, the system outputs the dual-reference proofreading results and the photo visual quality feedback results. Among them, the dual-reference proofreading results include the photo source conclusion, the PDF photo usage specification conclusion, and the print layout reproduction conclusion; the photo visual quality feedback results include the PDF photo visual quality conclusion, the printed photo visual quality conclusion, the abnormal type, the inconsistent type, and the recommended fallback path.

[0060] S7: Execute delivery or hierarchical fallback according to the dual-reference proofreading results and the photo visual quality feedback results, and perform parameter feedback adjustment according to the results. The system receives the dual-reference proofreading results, the photo visual quality feedback results, the cloud ID photo records, the certificate PDF, the photo specification data corresponding to the certificate type to be processed, the photo processing records, and the historical adjustment records. Combining the judgment in S6, different operations are performed on different results, such as Figure 4 As shown, it is the hierarchical fallback and anti-infinite loop closed-loop diagram provided by the embodiment of the present application, and the specific execution is as follows.

[0061] If the photo source conclusion is that the photo sources are consistent, the PDF photo usage specification conclusion is that the PDF photo usage specifications are consistent, the print layout reproduction conclusion is that the print layout reproduction is qualified, and the photo visual quality feedback result is qualified, then the system generates a delivery instruction, puts the printed copy into the 24-hour self-service collection cabinet, records that the certificate has been delivered, and synchronously sends the collection QR code information to the user's mobile phone. Among them, the photo visual quality feedback result being qualified means that both the PDF photo area and the printed photo area meet the requirements of the photo quality threshold data, and there is no visual quality attenuation in the printed photo area exceeding the allowable range of the print quality threshold data compared to the PDF photo area. Specifically, the face brightness value in the PDF photo area is not lower than the face brightness threshold, the face contrast value is not lower than the face contrast threshold, and the background edge color difference value is not greater than the background edge color difference threshold; the face brightness value in the printed photo area is not lower than the face brightness threshold, the face contrast value is not lower than the face contrast threshold, and the background edge color difference value is not greater than the background edge color difference threshold; and the face brightness decrease amount in the printed photo area compared to the PDF photo area is not greater than the allowable brightness decrease amount, the face contrast decrease amount is not greater than the allowable contrast decrease amount, and the background edge color difference increase amount is not greater than the allowable background edge color difference increase amount. Only when all the above conditions are met simultaneously, the photo visual quality feedback result is recorded as qualified. If the photo source conclusion is that the photo sources are inconsistent, the system does not deliver the printed copy and performs hierarchical fallback according to the inconsistent type. If the inconsistent type is abnormal PDF actual embedded photo or abnormal PDF associated data writing, the system returns to S4 to regenerate the certificate PDF and the certificate PDF associated data. If the inconsistent type is that the PDF used photo is inconsistent with the current valid record in the cloud, the system returns to S3 to re-match the cloud ID photo, and after S3 outputs the available cloud ID photo again, it enters S4 to generate a new certificate PDF.

[0062] If the photo source is consistent, but the PDF photo specification conclusion indicates a mismatch, the system will not deliver the printed copy. Instead, it will return to S3 to rematch cloud-based candidate photos that match the photo specification data corresponding to the certificate type to be processed. If S3 confirms that there is no cloud-based candidate photo that meets the current photo specification data, but there is an available source of the user's original face capture photo, then S3 will generate a reprocessing instruction to return to S1. If the source of the user's original face capture photo is unavailable, or the user's original face capture photo cannot be regenerated into a processed ID photo that meets the current photo specification data, then the system will generate a recapture instruction.

[0063] like Figure 5 The diagram shown is a visual quality anomaly attribution diagram provided in this application embodiment. If the photo sources are consistent and the PDF photo specifications are consistent, but the photo visual quality feedback results show that both the PDF photo area and the printed photo area have brightness values ​​lower than the face brightness threshold, or contrast values ​​lower than the face contrast threshold, and the number of adjustments for the same anomaly type in the photo processing record has not reached the upper limit of the adjustment count, then the system generates a photo processing parameter adjustment instruction to return to S1. This photo processing parameter adjustment instruction includes the anomaly type, adjustment direction, allowed adjustment range, number of adjustments already made, and upper limit of the adjustment count. After receiving this instruction, S1 performs controlled image quality processing only without changing the geometric proportions of the person, the background color, and the calculated results of the ID photo specifications.

[0064] If the number of automatic adjustments for the same visual quality anomaly type reaches the adjustment limit, or if the image specifications and quality thresholds still cannot be met simultaneously after reprocessing by S1, the system will stop automatic adjustment and generate a re-acquisition command or a manual review command. The adjustment limit is preset by the system deployment personnel, preferably set to a maximum of two or three automatic adjustments for the same anomaly type in the same certificate task, to avoid entering an infinite loop where the anomaly cannot be adjusted and the system repeatedly rolls back for processing.

[0065] If the photo source and PDF photo specifications are consistent, but the print layout reproduction result is a deviation, the system directly generates a reprint instruction and returns to S5 to re-acquire the printout image. If the photo source and PDF photo specifications are consistent, and the visual quality of the PDF photo area is acceptable, but the visual quality of the printout photo area is unacceptable, the system performs a rollback based on the photo visual quality feedback. When the corrected printout image is blurry, too dark, abnormally exposed, or has unclear page edges, the system generates a re-acquire the printout image instruction and returns to S5. When the PDF reference image is normal, the corrected printout image is generally calibrated, but the printout photo area has decreased brightness, decreased contrast, or increased background edge color difference compared to the PDF photo area, the system generates a reprint instruction and returns to S5 to re-acquire the printout image.

[0066] Furthermore, based on the visual quality feedback of the photos, the system makes restrictive corrections to the internal processing thresholds already existing in the photo specification configuration table. It should be noted that even when using the same printing equipment, a processed ID photo that meets the photo specification data electronically may still exhibit issues such as low brightness, insufficient contrast, or abnormal background edges in the printed version. This is because the output density and color grayscale of the same printing equipment fluctuate slightly in preheating, continuous printing, consumable status, and automatic calibration states. Additionally, for processed ID photos where face brightness, face contrast, or background edge color difference are near the acceptable threshold, although their electronic image indicators meet the photo specification data, the lack of reserved printing reproduction margin may result in the printed version falling outside the acceptable range. Therefore, the system makes restrictive corrections to the face brightness threshold, face contrast threshold, and background edge color difference threshold in S1, ensuring that S1 reserves printing reproduction margin when generating processed ID photos. Among them, the stricter correction of the face brightness threshold and face contrast threshold refers to raising the lower limit of face brightness and face contrast used when the system allows passage; the stricter correction of the background edge color difference threshold refers to lowering the upper limit of background edge color difference used when the system allows passage.

[0067] When the photo source is consistent, the PDF photo specification is consistent, the printed layout reproduction is satisfactory, and the overall printed image meets the proofreading requirements, but the photo area of ​​the printed document shows insufficient brightness, insufficient contrast, or abnormal background edges in at least three of the last five valid print feedbacks relative to the photo area of ​​the PDF, the system records this abnormality as print visual attenuation feedback. Valid print feedback requires consistent photo source, consistent PDF photo specification, satisfactory printed layout reproduction, a clear overall printed image, and the printing device not reporting ink shortage, toner shortage, paper jam, calibration failure, or other equipment malfunctions.

[0068] When valid print feedback shows that the face brightness value in the printed photo area is repeatedly lower than the face brightness value in the PDF photo area, and the difference between the two is greater than or equal to the allowable brightness reduction in the print quality threshold table, the system will adjust the face brightness threshold used for S1 processing in the photo specification configuration table to a stricter direction. That is, without changing the minimum qualification requirements stipulated by the competent authority, the system will raise the lower limit of face brightness used for internal release, so that the ID photo after subsequent processing will no longer be close to the minimum brightness boundary. Preferably, "repeatedly" means that at least three of the last five valid print feedbacks are lower than the corresponding value in the PDF photo area. The correction amount is determined based on the median of the brightness difference in the above-mentioned at least three abnormal feedbacks, but the correction amount for a single instance shall not exceed the maximum correction step size of the preset brightness threshold in the photo specification configuration table.

[0069] When valid print feedback shows that the face contrast value in the printed photo area is lower than the face contrast value in the PDF photo area multiple times, and the difference between the two is greater than or equal to the allowable contrast reduction in the print quality threshold table, the system will adjust the face contrast threshold used for S1 processing in the photo specification configuration table to a more stringent direction, so that the ID photo after subsequent processing has a contrast margin. Preferably, "multiple times" means that at least three of the last five valid print feedbacks are lower than the corresponding value in the PDF photo area. The correction amount is determined based on the median of the contrast difference in the above-mentioned at least three abnormal feedbacks, but the correction amount for a single instance shall not exceed the maximum correction step size of the contrast threshold preset in the photo specification configuration table.

[0070] When valid print feedback shows that the background edge color difference value of the printed photo area is higher than that of the PDF photo area multiple times, and the difference between the two is greater than or equal to the allowable increase in background edge color difference in the print quality threshold table, the system will adjust the background edge color difference threshold used for S1 processing in the photo specification configuration table to a more stringent direction. Preferably, "multiple times" means that at least three of the last five valid print feedbacks are higher than the corresponding values ​​of the PDF photo area. The adjustment amount is determined based on the median of the background edge color difference increments in the above-mentioned at least three abnormal feedbacks, but the adjustment amount for a single instance shall not exceed the maximum adjustment step size of the preset background edge color difference threshold in the photo specification configuration table. If the triggering conditions of brightness abnormality, contrast abnormality, and background edge abnormality are met simultaneously in the same batch of valid print feedback, the system will perform restricted adjustments on the corresponding thresholds respectively. The thresholds are independent of each other and no priority relationship is set.

[0071] The aforementioned modifications must not relax the photo specifications corresponding to the current certificate type, and must not change the geometric proportions of the person, the top of the head, the chin point, the minimum reserved boundary of the shoulders, the background color, or the shape of the facial features. In other words, the print feedback parameters are only to allow S1 a margin for print reproduction when generating the processed ID photo, not to modify the official photo specifications or lower the photo acceptance standard. If the same anomaly does not recur, or the overall printed image is unclear, or the printing device is in an abnormal state, the system will not update the print feedback parameters, but will only perform a re-capture of the printed image, reprinting, or manual review.

[0072] like Figure 3 The diagram shown is a structural schematic of a certificate generation and delivery system for certificate-related business provided in this application embodiment, including a certificate photo specification generation module, a cloud photo archiving module, a photo usage condition matching module, a certificate layout generation module, a printed image correction module, a certificate usage verification module, and a delivery rollback control module.

[0073] The ID photo specification generation module is used to receive the user's original face capture photo, user identity identifier, current ID photo type and photo specification data, and generate the processed ID photo, the hash of the original capture photo, the hash of the processed ID photo, the coordinates of the cropped rectangle and the calculated value of the photo specification.

[0074] The cloud-based photo archiving module is used to bind processed ID photos with user identity identifiers to form cloud-based ID photo records.

[0075] The photo matching module is used to retrieve cloud-based ID photo records when applying for a certificate, and determine the available cloud-based ID photos based on the photo specification data corresponding to the type of certificate to be processed.

[0076] The certificate template generation module is used to embed available cloud-based ID photos into the certificate template, generating a certificate PDF and associated data.

[0077] The print image correction module is used to acquire and correct the print image, forming a corrected print image.

[0078] The certificate photo verification module is used to perform dual-benchmark verification and photo visual quality feedback analysis based on the certificate PDF, certificate PDF associated data, cloud-based ID photo records, and corrected printed images. The verification is based on the photo source and photo specifications as the first benchmark and the print reproduction as the second benchmark. The module generates dual-benchmark verification results and photo visual quality feedback results.

[0079] The delivery rollback control module is used to generate delivery instructions or rollback instructions based on the dual-benchmark calibration results and the photo visual quality feedback results.

[0080] Through the above description of the implementation methods, those skilled in the art can clearly understand that, for the sake of convenience and brevity, only the division of the above functional modules is used as an example. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the above functions can be divided into different functional modules to complete all or part of the functions described above.

[0081] In the embodiments provided in this application, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the embodiments described above are merely illustrative; for instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces, or indirect coupling or communication connection between devices or units, and may be electrical, mechanical, or other forms.

[0082] The units described as separate components may or may not be physically separate. A component shown as a unit can be one or more physical units, located in one place or distributed in multiple different locations. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0083] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0084] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a readable storage medium. Based on this understanding, the technical solutions of the embodiments of this application, or the parts that contribute to the solution, or all or part of the technical solution, can be embodied in the form of a software product. This software product is stored in a storage medium and includes several instructions to cause a device (which may be a microcontroller, chip, etc.) or processor to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0085] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for certificate generation and delivery in certificate-related business, characterized in that, Includes the following steps: S1: Receive the user's original face capture photo, user identity identifier, current certificate type and photo specification data, and generate the processed certificate photo, original capture photo hash, processed certificate photo hash, cropping rectangle coordinates and photo specification calculation value; S2: Bind the processed ID photo with the user's identity identifier to form a cloud-based ID photo record; S3: When processing a certificate, retrieve the cloud-based ID photo records and determine the available cloud-based ID photos based on the photo specification data corresponding to the type of certificate to be processed; S4: Embed the available cloud-based ID photo into the certificate template to generate a certificate PDF and associated data between the certificate PDF; S5: Acquire and correct the printed image to generate a corrected printed image; S6: Based on the certificate PDF, the certificate PDF associated data, the cloud ID photo record and the corrected printed image, perform dual-benchmark verification and photo visual quality feedback analysis with photo source and photo specifications as the first benchmark and print reproduction as the second benchmark, and generate dual-benchmark verification results and photo visual quality feedback results. S7: Generate a delivery instruction or rollback instruction based on the dual-benchmark calibration results and the photo visual quality feedback results; The dual-benchmark calibration results include: The photo source conclusion is generated by comparing the embedded photo hash in the PDF, the processed ID photo hash in the certificate PDF associated data, and the processed ID photo hash in the cloud ID photo record. When the photo sources are consistent, the calculated value of the photo specifications in the certificate PDF association data is compared with the photo specifications data corresponding to the certificate type to be processed, and a PDF photo specifications conclusion is generated. The certificate PDF is converted into a PDF reference image. The PDF photo area rectangle and the printout photo area rectangle are located in the PDF reference image and the corrected printout image, respectively. The horizontal coordinate of the upper left corner, the vertical coordinate of the upper left corner, the width and the height of the PDF photo area rectangle, and the horizontal coordinate of the upper left corner, the vertical coordinate of the upper left corner, the width and the height of the printout photo area rectangle are recorded. The absolute difference between the horizontal coordinate of the top left corner of the printed photo area rectangle and the horizontal coordinate of the top left corner of the PDF photo area rectangle is used as the horizontal offset; the absolute difference between the vertical coordinate of the top left corner of the printed photo area rectangle and the vertical coordinate of the top left corner of the PDF photo area rectangle is used as the vertical offset; the absolute difference between the width of the printed photo area rectangle and the width of the PDF photo area rectangle is used as the width difference; and the absolute difference between the height of the printed photo area rectangle and the height of the PDF photo area rectangle is used as the height difference. Based on the comparison results of the horizontal offset and the horizontal offset threshold, the vertical offset and the vertical offset threshold, the width difference and the width difference threshold, and the height difference and the height difference threshold, a conclusion on the reproduction of the printed version is generated.

2. The certificate generation and delivery method for certificate-related services as described in claim 1, characterized in that, The process of receiving the user's original facial image, user identification, current certificate type, and photo specifications includes: The user's original face image is subjected to person detection, and the top of the head, the chin point, the lowest preserved boundary of the shoulders, and the face region are extracted; The cropping rectangle is determined based on the photo specification data, such that the cropping rectangle includes the head vertex, the chin point, and the lowest preserved boundary of the shoulder. The image within the cropping rectangle is scaled to the photo size corresponding to the current certificate type, and the background is replaced according to the background color in the photo specification data; Based on the cropping rectangle, the head vertex, the chin point, and the lowest retained shoulder boundary are mapped to the processed ID photo, and the head height ratio, top white space ratio, and shoulder retention ratio are calculated; A background edge band is generated outside the figure outline. The difference between the pixel color within the background edge band and the background color is calculated to obtain the background edge color difference value. The photo specification data includes at least the photo size, background color, head height percentage range, top white space percentage range, shoulder retention percentage range, background edge color difference threshold, face brightness threshold, face contrast threshold, and face clarity threshold. The photo specification calculation values ​​include at least the head height percentage, top white space percentage, shoulder retention percentage, background edge color difference value, face brightness value, face contrast value, and face clarity value.

3. The certificate generation and delivery method for certificate-related services as described in claim 1, characterized in that: The process of binding the processed ID photo with the user's identity identifier includes: A unique user index code is generated based on the user's identity identifier; the processed ID photo, the hash of the original captured photo, the hash of the processed ID photo, the coordinates of the cropped rectangle, the calculated value of the photo specification, the current ID photo type, the photo record status, and the photo version number are associated and stored with the unique user index code; when the same user already has a processed ID photo of the same ID photo type, the existing record is retained or a new cloud ID photo record is generated based on the hashes of the old and new processed ID photos; The original captured photo hash is generated based on the normalized pixel data of the original captured photo, and the processed ID photo hash is generated based on the normalized pixel data of the processed ID photo. The normalized pixel data is obtained by performing orientation correction, color space unification, bit depth unification, metadata removal, and pixel order arrangement on the image to be calculated.

4. The certificate generation and delivery method for certificate-related services as described in claim 1, characterized in that: The process of determining available cloud-based ID photos based on the photo specifications corresponding to the type of certificate to be processed includes: Read the corresponding photo specification data according to the type of certificate to be processed; Based on the user's identity identifier, the processed ID photo is retrieved from the cloud ID photo record, and the retrieved processed ID photo is used as a candidate photo; Compare the calculated photo specification values ​​of the candidate photos with the photo specification data corresponding to the type of certificate to be processed item by item; When the candidate photo does not meet the photo specification data corresponding to the certificate type to be processed, and there is an original photo source that can be called, the original photo source is verified based on the hash of the original photo, and S1 is returned to regenerate the processed ID photo. The photo specification data is provided by the photo specification configuration table. The photo specification configuration table uses certificate type as the primary index and photo size, background color, head height ratio range, top white space ratio range, shoulder retention ratio range, background edge color difference threshold, face brightness threshold, face contrast threshold, and face clarity threshold as specification items. For each specification item, a value, judgment method, source identifier, effective time, and photo specification configuration version number are set.

5. The certificate generation and delivery method for certificate-related services as described in claim 1, characterized in that: The generation of the certificate PDF and the associated certificate PDF data includes: The photo area coordinates, photo area size, and template version are determined based on the certificate template data, and the embedding and adaptation status between the available cloud ID photo and the certificate template is checked. When the available cloud ID photo meets the embedding adaptation state, the available cloud ID photo is embedded into the photo area of ​​the certificate template in an equal proportion, and the cloud ID photo record version number and photo specification configuration version number are written into the certificate PDF association data.

6. The certificate generation and delivery method for certificate-related services as described in claim 1, characterized in that: The process of acquiring and correcting the printed image also includes: After the printing is completed and the markings are generated, obtain the image of the printed part and locate the four corners of the page or the fixed edge of the template in the image of the printed part. The image sharpness value of the printed part is calculated based on the fixed edge line of the page, the text area, and the QR code area. When the four corners of the page or the fixed edge line of the template can be located and the image sharpness value of the printed part meets the image sharpness threshold, perspective correction is performed on the printed part image to obtain the corrected printed part image.

7. The certificate generation and delivery method for certificate-related services as described in claim 1, characterized in that: The visual quality feedback analysis of the photographs includes: Extract the face region and background edge band from the PDF photo area and the printed photo area respectively; Calculate the brightness value of the face in the PDF photo area, the contrast value of the face in the PDF photo area, the color difference value of the background edge in the PDF photo area, the brightness value of the face in the printed photo area, the contrast value of the face in the printed photo area, and the color difference value of the background edge in the printed photo area; Based on the comparison results of the above values ​​with the photo quality threshold data and the print quality threshold data, the visual quality conclusions for PDF photos, the visual quality conclusions for printed photos, the anomaly types, and the suggested rollback paths are generated. The print quality threshold data includes the allowable decrease in brightness, the allowable decrease in contrast, and the allowable increase in background edge color difference. Printing visual attenuation feedback is generated when the decrease in the brightness value of the face in the printed photo area relative to the brightness value of the face in the PDF photo area reaches the allowable decrease in brightness, or the decrease in the contrast value of the face in the printed photo area relative to the contrast value of the face in the PDF photo area reaches the allowable decrease in contrast, or the increase in the color difference value of the background edge in the printed photo area relative to the color difference value of the background edge in the PDF photo area reaches the allowable increase in the color difference of the background edge.

8. The certificate generation and delivery method for certificate-related services as described in claim 4, characterized in that: The generation of delivery instructions or rollback instructions includes: When the conclusion that the photo source is consistent, the conclusion that the PDF photo usage specification is consistent, the conclusion that the printed version reproduction is qualified, and the photo visual quality feedback result is qualified, a delivery instruction is generated. When the source of the photo is determined to be inconsistent, generate a regenerate certificate PDF instruction or a rematch instruction based on the type of inconsistency. When the PDF is found to be inconsistent with the specification, a rematch instruction is generated. When the printout reproduction result is a deviation, a reprint instruction is generated; When the printed image is abnormal, a command to re-acquire the printed image is generated; Furthermore, when the visual attenuation feedback of the same type of printing reaches a preset number of valid printing feedbacks under the same certificate type, template version, and printing device conditions, the internal processing threshold in the photo specification configuration table is subject to restricted correction. The internal processing thresholds include at least the face brightness threshold, the face contrast threshold, and the background edge color difference threshold. The restricted correction does not change the preset photo specification data corresponding to the current certificate type, nor does it change the geometric proportions of the person, the top of the head, the chin point, the lowest preserved boundary of the shoulders, the background color, and the shape of the facial features.

9. A certificate generation and delivery system for certificate and license services, employing the certificate generation and delivery method for certificate and license services as described in any one of claims 1-8, characterized in that, include: The module includes: ID photo standard generation module, cloud photo archiving module, photo usage condition matching module, certificate layout generation module, printed image correction module, certificate usage verification module, and delivery rollback control module. The ID photo specification generation module is used to receive the user's original face capture photo, user identity identifier, current ID photo type and photo specification data, and generate the processed ID photo, the hash of the original capture photo, the hash of the processed ID photo, the coordinates of the cropping rectangle and the calculated value of the photo specification. The cloud-based photo archiving module is used to bind the processed ID photo with the user's identity identifier to form a cloud-based ID photo record; The photo matching module is used to retrieve the cloud-based ID photo records when processing a certificate, and determine the available cloud-based ID photos based on the photo specification data corresponding to the type of certificate to be processed. The certificate layout generation module is used to embed the available cloud-based ID photo into the certificate template to generate a certificate PDF and certificate PDF associated data; The print image correction module is used to acquire and correct the print image to form a corrected print image; The certificate photo verification module is used to perform dual-benchmark verification and photo visual quality feedback analysis based on the certificate PDF, the certificate PDF associated data, the cloud ID photo record, and the corrected printed image, with photo source and photo specifications as the first benchmark and print reproduction as the second benchmark, and generate dual-benchmark verification results and photo visual quality feedback results. The delivery rollback control module is used to generate a delivery instruction or rollback instruction based on the dual-benchmark calibration results and the photo visual quality feedback results.

Images