An automatic compliance verification method, system, device, and medium for warehousing scenarios.

Abstract

This application provides a method, system, equipment, and medium for automatic compliance verification in warehousing scenarios, belonging to the field of intelligent warehouse management technology. The method includes: collecting environmental photos and close-up photos of invoices related to the warehousing operation; performing image recognition on the environmental photos to determine whether they simultaneously contain three key elements: the operator's face, the shape of the goods, and the scale reading; extracting facial features and matching them with a pre-built authorization database; performing OCR recognition on the invoice photos to extract category and weight information, and comparing it with the weighing reading and the entered category; and determining non-compliance and blocking the warehousing process in real time when any key element is missing, the identity is unauthorized, the weight deviation exceeds a threshold, or the category is inconsistent. This invention transforms compliance judgment from post-event manual sampling to real-time control during the process, ensuring the integrity and validity of the stored photos, achieving automatic cross-verification of invoice information and weighing data, effectively preventing misrecording, omissions, and forgery, and reducing regulatory costs.

Description

Technical Field

[0001] This application relates to the field of intelligent warehouse management technology, specifically to an automatic compliance verification method, system, equipment, and medium for inbound scenarios. Background Technology

[0002] In the field of canteen food and warehouse material warehousing management, the current common practice is to use intelligent weighing terminals to collect photos of the on-site environment and warehousing documents as evidence of operations, mainly for post-event review. The existing warehousing management model relies on manual verification of documents, weighing data, and warehousing item information, and compliance judgment relies on manual spot checks and post-event review, failing to achieve real-time control of the operation process.

[0003] In practical applications, on-site environmental photos often suffer from problems such as improper shooting angles and incomplete content. Key evidentiary elements such as the operator's face, items to be stored, and weighing equipment readings are often missing or obscured, rendering the photos ineffective as evidence. Furthermore, the identity of operators in the storage process is not linked to the on-site operation screen in real time, making it impossible to automatically verify operating permissions and posing a risk of unauthorized personnel operating illegally. In addition, there is a lack of automated verification mechanisms between storage invoice information, system-entered information, and actual weighing data. Manual entry is prone to errors such as incorrect product categories and weight deviations, and even the potential for data tampering and fraudulent storage. Once anomalies occur, they are difficult to detect in a timely manner during the storage process, making subsequent tracing and liability determination challenging. Therefore, the existing storage supervision methods suffer from low automation, delayed verification, and unstable evidence quality, failing to meet the real-time compliance management and safety supervision needs of canteen food and other supplies entering the warehouse. Supervision efficiency and risk control capabilities need to be improved. Summary of the Invention

[0004] In view of this, this application aims to provide a method, system, device and medium for automatic verification of compliance in warehousing scenarios, in order to solve the problem of how to transform the compliance judgment of warehousing operations from post-event manual spot checks to real-time automatic verification during the process, so as to solve the technical problems of real-time control of non-compliant operations, which leads to regulatory lag, invalid evidence, and data errors.

[0005] The first aspect of this application provides an automatic compliance verification method for an inbound scenario, the method comprising: Collect environmental photos and close-up photos of invoices during the warehousing process; Image recognition is performed on the environmental photo to determine whether the environmental photo simultaneously contains the following three key elements: the operator's face area, the appearance of the items to be put into storage, and the reading display area of ​​the weighing equipment; The operator's facial features are extracted from the environmental photos and matched with a pre-established authorized identity feature database to determine whether the operator is an authorized person. The close-up photo of the ticket is subjected to OCR recognition to extract the product category and labeled weight information, and the extracted information is compared with the actual reading of the weighing equipment and the product category entered into the system. If any of the three key elements is missing, or the matching operator is not authorized, or the weight deviation exceeds a preset threshold or the product category is inconsistent, the current warehousing operation is determined to be non-compliant in real time, a prompt message is output and the warehousing process is blocked; otherwise, the warehousing process is completed.

[0006] In one possible implementation of this application, the process of establishing the authorized identity feature database includes: The system acquires multiple frames of facial images of authorized operators collected by the administrator, extracts facial feature vectors, and stores them in an encrypted manner. It also binds the employee's employee number, name, department, and authorization validity period to establish an authorized identity feature database. The addition, deletion, and modification operations of the authorized identity database can only be performed by the administrator account, and each authorization record has an validity period field. Expired or revoked authorizations are directly judged as non-compliant. In scenarios where terminals and the cloud are deployed collaboratively, the terminal and the cloud synchronize the authorization library on a periodic or event-triggered basis.

[0007] In one possible implementation of this application, the step of extracting the operator's facial features from the environmental photograph and matching them with a pre-established authorized identity feature database to determine whether the operator is an authorized person includes: Detect face regions from the environmental photos, and extract face feature vectors from each detected face region; The similarity between the extracted facial feature vector and the feature vector corresponding to the valid authorization record in the authorized identity feature database is calculated. When the highest similarity score obtained exceeds the preset threshold, the identity is determined to be matched successfully, and the operator is an authorized person; otherwise, it is determined to be an unauthorized person.

[0008] In one possible implementation of this application, the step of performing OCR recognition on the close-up photo of the invoice to extract the product category and labeled weight information includes: The target field is extracted from the OCR recognition result using at least one of the following methods: by triggering localization through preset keywords, the text line immediately adjacent to the keyword is identified as the target field; or, by using the table space proximity rule, the target field is extracted using the row and column position relationship of the text line. Each extracted field is assigned a recognition confidence score. When the confidence score is lower than a preset threshold, the field is marked as requiring manual confirmation, and the operator is prompted to check it on the operation interface. In response to the operator's verification and confirmation operation, the recognition result of the field is marked as valid and used for subsequent consistency comparison; or, in response to the operator's correction input, the corrected value is used to replace the OCR recognition result for subsequent consistency comparison.

[0009] In one possible implementation of this application, after the data entry process is blocked, the data entry operation is resumed when at least one of the following release conditions is met: The block will be automatically lifted once the operator retakes photos of the environment and tickets as prompted, and all re-performed checks pass. The block is lifted when the operator corrects the entered category or weight information within the preset limits, and the verification is successfully performed again after correction; The block will be lifted in response to the verification and confirmation operation performed by authorized administrators through the back-end management system or terminal.

[0010] In one possible implementation of this application, the method further includes: recording all blocking operations, unblocking operations, and auditing operations, the recorded information including: triggering reason, operator identification, operation timestamp, and processing method; the records are used for subsequent auditing and tracing.

[0011] The second aspect of this application provides an automatic compliance verification system for warehousing scenarios, including: The image acquisition module is used to capture environmental photos and close-up photos of invoices during the warehousing operation. The image recognition and analysis module is used to perform image recognition on the environmental photo and determine whether the environmental photo simultaneously contains the following three key elements: the operator's face area, the appearance of the items to be put into storage, and the reading display area of ​​the weighing equipment. The authorization and identity verification module is used to extract the operator's facial features from the environmental photos and match them with a pre-established authorization and identity feature database to determine whether the operator is an authorized person. The consistency comparison module is used to perform OCR recognition on the close-up photo of the ticket, extract the product category and labeled weight information, and compare the OCR recognition information with the actual reading of the weighing equipment and the product category entered by the system. The compliance verification module is used to determine in real time that the current warehousing operation is non-compliant when any of the three key elements is missing, the matching judgment operator is not authorized, the comparison judgment weight deviation exceeds a preset threshold, or the categories are inconsistent; otherwise, the warehousing process is completed.

[0012] In one possible implementation of this application, the system is deployed locally on the intelligent weighing terminal, or it is deployed in collaboration between the terminal and the cloud.

[0013] A third aspect of this application provides an electronic device, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, the computer program being executed by the at least one processor to cause the at least one processor to perform an automatic verification method for compliance of an inbound scenario as described in the first aspect and possible implementations thereof.

[0014] The fourth aspect of this application provides a computer storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement an automatic compliance verification method for data entry scenarios, as described in the first aspect and possible implementations thereof.

[0015] The fifth aspect of this application provides a computer program product, comprising: a computer program that, when executed by a processor, implements an automatic compliance verification method for database entry scenarios as described in the first aspect and possible implementations of the first aspect.

[0016] The automatic compliance verification method for warehousing scenarios provided in this application automatically identifies and verifies the completeness of key elements, the authorized identity of operators, and the consistency between document information and on-site data by simultaneously collecting photos of the warehousing environment and close-up photos of documents. It can instantly identify non-compliant situations and block the process during warehousing, ensuring from the source that the warehousing certificate photos are complete and valid, the operators' identities are legal and traceable, and the document information is consistent with the actual weighing and system-entered data. This achieves automatic control of warehousing compliance during the process, effectively improving verification efficiency and regulatory accuracy, reducing the risk of human error and illegal operations, and forming a complete and traceable chain of evidence. It meets the safety and compliance management needs of warehousing raw materials for canteens, reduces labor costs, and improves overall regulatory efficiency and system credibility. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the specific embodiments or related technologies of this application, the drawings used in the description of the specific embodiments or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart illustrating the automatic compliance verification method for the warehousing scenario provided in the embodiments of this application.

[0019] Figure 2 This application provides a schematic diagram of the structure of an automatic compliance verification system for warehousing scenarios.

[0020] Figure 3 This is a schematic diagram of the hardware structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0021] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0022] Example scenario: The automatic compliance verification method for the warehousing scenario in this embodiment is applied to an automatic compliance verification system for the warehousing scenario. This embodiment is applied to the raw material warehousing scenario of school / corporate canteens. The system adopts a terminal + cloud collaborative deployment, and the hardware includes: Intelligent weighing terminal: equipped with a touch screen, processor, and memory, and runs a verification client; Image acquisition module: Two high-definition industrial cameras, one mounted above the weighing platform to take photos of the warehousing environment; the other facing the ticket placement area to take close-up photos of the tickets. Smart electronic scale: accuracy of 0.01kg, real-time uploading of weighing data; Cloud server: Stores authorized identity feature database, entry records, and verification logs, and supports terminal synchronization by period / event.

[0023] The software employs a lightweight image recognition model, a face feature extraction model, and an OCR recognition model, supporting both local offline recognition and real-time verification.

[0024] Before the system is put into use, the administrator first establishes a database of authorized user identity features through the backend management system. The specific process is as follows: 1. The administrator logs into the cloud management backend (only the administrator account has the operation permission), enters the "Authorized Personnel Management" module, and clicks "Add Authorized Personnel".

[0025] 2. The administrator collects multiple frames of facial images (at least 3 frames, covering different angles such as frontal and slightly profile views) of the authorized operators on-site. The images are captured by the camera of the smart terminal, and the system automatically extracts the facial feature vector of each frame. The images are then encrypted and stored using the AES encryption algorithm to prevent the leakage of feature information.

[0026] 3. Bind the relevant information of the operator, including employee number (e.g., “XXX”), name (e.g., “Zhang San”), department (e.g., “Inbound Management Department”), and authorization validity period (e.g., “20XX-01-01 to 20XX-12-31”), and complete the creation of a single authorization record.

[0027] 4. Repeat steps 2-3 to complete the authorization entry for all personnel involved in the data entry process and establish a complete authorized identity feature database. If it is necessary to add, delete, or modify authorization records in the future, only the administrator can operate through the backend; ordinary operators do not have this permission.

[0028] 5. The terminal and cloud automatically synchronize the authorization database daily, and the administrator can trigger the synchronization immediately after making changes.

[0029] Exemplary method: This application provides a flowchart illustrating the automatic compliance verification method for inbound scenarios. This embodiment uses an industrial raw material inbound scenario (such as chemical raw materials or mechanical parts) as its application scenario. This scenario has high requirements for the standardization of inbound operations and the accuracy of data, and needs to avoid compliance risks such as unauthorized operations, discrepancies between goods and documents, and excessive weight deviations, achieving automatic compliance verification throughout the entire inbound process. Figure 1 As shown, the method specifically includes: S101: Collect environmental photos and close-up photos of invoices for the warehousing operation.

[0030] Specifically, the operator (e.g., Zhang San) places the raw materials to be stored on the smart electronic scale, arranging them neatly so as not to obstruct the scale's reading area; simultaneously, the storage invoice is laid flat in the invoice placement area, without wrinkles or obstructions. The operator clicks the "Start Verification" button on the smart terminal, triggering the image acquisition module to start. Two cameras simultaneously take pictures: the upper camera captures an environmental photo (covering the operator's face, the appearance of the sodium hydroxide solid to be stored, and the electronic scale's reading display area), while the front camera captures a close-up photo of the invoice (clearly showing all the text information on the invoice). After the photos are taken, the images are automatically uploaded to the smart terminal's image recognition and analysis module.

[0031] S102: Perform image recognition on the environmental photo to determine whether the environmental photo simultaneously contains the following three key elements: the operator's face area, the appearance of the items to be put into storage, and the reading display area of ​​the weighing equipment.

[0032] Specifically, the image recognition and analysis module performs real-time recognition on environmental photos to determine whether they simultaneously contain three key elements: 1. A clear, unobstructed view of the operator's face; 2. Identifiable physical appearance of items to be received into the warehouse; 3. A fully visible display area for weighing equipment readings.

[0033] After identification, all three types of elements were found to be present in their entirety, and the process proceeded to the next step of verification. Through mandatory verification of key elements, it was ensured that each document photo contained complete and clear key information.

[0034] S103: Extract the operator's facial features from the environmental photo and match them with a pre-established authorized identity feature database to determine whether the operator is an authorized person.

[0035] Specifically, facial regions are detected from the environmental photos, and facial feature vectors are extracted from each detected facial region. The extracted facial feature vectors are then compared with the feature vectors corresponding to the valid authorization records in the authorized identity feature database. When the calculated highest similarity score exceeds a preset threshold, the identity is determined to be successfully matched, and the operator is an authorized person. Otherwise, the operation is determined to be performed by an unauthorized person.

[0036] In one example, the authorized identity management module extracts Zhang San's real-time facial feature vector from an environmental photo and performs a cosine similarity calculation with the feature vector corresponding to the valid authorized identity feature library cached locally on the terminal. The system's preset similarity threshold is 0.85. The highest similarity score calculated in this instance is 0.93, which is higher than the threshold. Since the authorization is within the validity period, the identity is determined to be successfully matched, indicating that the operation was performed by an authorized person.

[0037] S104: Perform OCR recognition on the close-up photo of the ticket to extract the product category and labeled weight information, and compare the OCR recognition information with the actual reading of the weighing equipment and the product category entered into the system.

[0038] Specifically, after the system captures a close-up photo of the invoice, it performs optical character recognition (OCR) processing on the photo. In this embodiment, the invoice is a delivery note provided by the supplier, which contains text information such as "Product Name: Northeast Rice", "Net Weight: 50kg", and "Supplier Name".

[0039] In this embodiment of the invention, target fields of close-up photos of invoices are extracted by triggering location based on preset keywords or by using table space proximity rules; each extracted field is assigned a recognition confidence score, and when the confidence score is lower than a preset threshold, it is marked as requiring manual confirmation, and the operator is prompted to check on the interface.

[0040] In one example, the system extracts fields by triggering a location method using preset keywords. Specifically, the system predefines a keyword dictionary containing keywords such as "product name," "category," "net weight," and "weight," along with their common abbreviations / synonyms. The system performs keyword matching on the text set of the OCR recognition results, identifying the text line immediately adjacent to the keyword "product name" as product category information ("Northeast Rice"), and the text line immediately adjacent to the keyword "net weight" as labeled weight information ("50kg").

[0041] If the invoice uses a table format, the system can also extract data using table space proximity rules. This involves inferring the table structure based on line spacing and alignment, and extracting field values ​​according to row and column positions. Each extracted field is assigned a confidence score. For example, the confidence score for the "Product Name" field is 0.96, and the confidence score for the "Net Weight" field is 0.88, both higher than preset confidence thresholds (e.g., 0.80). The system directly uses these recognition results for subsequent comparisons.

[0042] Furthermore, the information extracted by OCR was compared with the actual reading of the weighing device and the category entered into the system. Specifically, the category extracted by OCR was "Northeast Rice," and the category entered into the system was also "Northeast Rice," indicating consistency. The labeled weight extracted by OCR was "50kg," while the actual reading of the weighing device was "50.5kg," a weight deviation of 0.5kg. The preset weight deviation threshold was 1kg, and the actual deviation did not exceed the threshold, thus the consistency verification passed.

[0043] If the OCR identifies "Northeast Rice" as the product category while the system enters "Wuchang Rice" as the product category, the system will determine that the product categories are inconsistent and trigger an anomaly alarm. If the invoice indicates a weight of 50kg but the actual weighing reading is 55kg, the deviation of 5kg exceeds the preset threshold of 1kg, and the system will also determine it as an anomaly.

[0044] S105: If any key element is missing, the authorized identity does not match, the weight deviation exceeds the preset threshold, or the product category is inconsistent, non-compliance will be determined in real time and a prompt will be output to block the warehousing process.

[0045] Specifically, when the integrity verification, identity verification, and consistency verification all pass, the system determines that the current warehousing operation is compliant and stores or uploads the verified warehousing data (including environmental photos, invoice photos, weighing data, operator information, etc.) to the backend management system.

[0046] If any verification step fails, the system immediately determines that the current inbound operation is non-compliant, outputs corresponding prompts, and blocks the inbound process. For example, during image recognition of environmental photos, if any element is missing, blurred, or obscured, the system immediately determines it is non-compliant, displays the message "Key element missing, please retake the photo," and blocks the inbound process. During identity verification, if the identity does not match, authorization has expired, or has been revoked, the system directly determines it is non-compliant, displays the message "Unauthorized operation, inbound prohibited," and blocks the process. During consistency verification, if the confidence level of a field is lower than a preset threshold, the system marks the field as "Pending manual confirmation" and highlights the corresponding area on the operation interface, prompting the operator to check before submitting. If a category inconsistency is detected, an exception prompt for category inconsistency is triggered, blocking the process.

[0047] In this embodiment of the invention, if the warehousing process is blocked, the system allows the warehousing operation to be resumed after any of the following release conditions are met: Condition 1: Re-capture verification passed. When the system blocks the process due to missing key elements in the environmental photo (such as the scale reading being obscured) or low confidence in the OCR recognition of the invoice photo, the operator can retake the environmental and invoice photos according to the system's output prompts (such as "The reading is not clearly displayed, please retake the photo" or "The text on the invoice is blurry, please retake the photo"). The system will automatically lift the blockage once the operator retakes the environmental and invoice photos as prompted, and all re-performed verifications (in response to the operator retaking the environmental and invoice photos as prompted) pass.

[0048] For example, when operator Zhang San took the first photo, the electronic scale's display screen was obstructed by his arm, and the system prompted, "The scale reading is not displayed. Please take another photo." After Zhang San adjusted the shooting angle and took another photo, the system detected a clear reading area, the verification passed, and the blockage was automatically lifted.

[0049] Condition 2: Re-verification after data correction. When the system triggers a blockage due to inconsistencies between the OCR recognition information and the actual reading of the weighing equipment or the category entered into the system, if the inconsistency is caused by an operator's input error, the system allows the operator to correct the input information within a preset limit.

[0050] Specifically, the operating terminal interface will highlight inconsistent fields and provide an editing entry. Operators can correct the category information (by selecting the correct category from a drop-down menu) or the weight information (e.g., by manually entering a correction value). After correction, the system will re-trigger the verification process, comparing the corrected information with the actual reading from the weighing equipment. Once the verification passes, the system will remove the blockage.

[0051] For example, when operator Zhang San was putting "Northeast Rice" into the warehouse, he mistakenly selected "Wuchang Rice" as the category entered into the system. After the system's OCR recognized "Northeast Rice" on the document and compared it with the entered "Wuchang Rice," it found a mismatch and triggered a block. Zhang San corrected the category to "Northeast Rice" on the terminal interface, the system re-verified and passed the check, the block was lifted, and the warehousing process continued.

[0052] It should be noted that the correction range is limited. For example, the deviation of the weight correction must not exceed ±10% of the original weighing reading to prevent malicious tampering. This limit can be pre-configured by the administrator in the backend system.

[0053] Condition 3: Manual authorization revocation. In actual business operations, there may be special circumstances that cause automatic verification to fail, even though the business itself is compliant. For example: The original invoices are illegible due to improper transportation or storage, and the OCR recognition confidence level is lower than the preset threshold; the product name on the invoice provided by the supplier differs from the standard product name in the internal system (e.g., "potato" vs. "wheat gluten"); due to seasonality or batch differences, the actual weight deviates from the weight marked on the invoice from the preset threshold, but is within a reasonable range.

[0054] In the above scenario, if the operator fails to verify multiple times consecutively (e.g., 3 times), the system will mark the failure as "pending manual review" and automatically push a review request to the backend management system. Authorized managers (such as the canteen supervisor or warehouse manager) can view relevant environmental photos, invoice photos, weighing data, and reasons for verification failure through the backend management system or authorized terminals to make a comprehensive judgment.

[0055] If the administrator deems the inbound operation compliant, they can perform an audit confirmation. The system will respond to this confirmation, lifting the block and allowing the inbound process to continue. If the administrator deems the operation non-compliant, they can maintain the block and record the reason for rejection.

[0056] For example, a batch of potatoes was labeled as weighing 100kg on the invoice, but actually weighed 108kg, a deviation of 8%, exceeding the preset 5% threshold. After three consecutive failed verifications by the operator, the system sent an audit request to the administrator. The administrator reviewed the photos and found that the higher moisture content of the potatoes caused the weight increase, which was within a reasonable range. Therefore, the administrator approved the request through the backend and lifted the block.

[0057] This invention also includes complete logging of all blocking, unblocking, and auditing operations to ensure traceability and audit compliance. Each record contains at least the following information: 1. Triggering reason: The specific type of verification failure that is blocked, such as "weight deviation exceeds threshold".

[0058] 2. Operator: The identifier of the person performing the warehousing operation. For example, "Zhang San (employee number XXX)".

[0059] 3. Operation Time: The timestamp of the blocking / unblocking / review. For example, "20XX-0X-0X 10:32:15".

[0060] 4. Processing method: The type of conditions used to lift the block, such as "manual authorization to lift".

[0061] 5. Reviewer: If the removal is done manually, record the reviewer as "Li Si (Administrator)".

[0062] 6. Remarks: Additional explanatory information, such as "The handwriting on the invoice is illegible, but the transaction has been verified to be genuine".

[0063] The aforementioned log records are stored in the system database and can be retrieved and exported by time, operator, processing method, and other dimensions. They can be used for subsequent internal audits, compliance checks, or accountability tracing.

[0064] This invention, through the setting of multiple blocking and lifting conditions, ensures compliance in warehousing while also balancing operational flexibility and the ability to handle exceptional situations. Specifically: (1) The re-acquisition and verification mechanism solves the problem of false blocking caused by temporary issues such as shooting angle and obstruction, and improves the fault tolerance of the system; (2) The data correction and verification mechanism allows operators to correct input errors, avoids process interruption caused by human error, and improves operational efficiency; (3) The manual authorization release mechanism provides a compliant exception handling channel for special business scenarios, avoiding business blockage caused by rigid system judgment; (4) A full-process traceability mechanism ensures that all blocking and unblocking operations are traceable, meeting the audit compliance requirements in the fields of food safety and warehousing management.

[0065] Exemplary system: Figure 2 This is a schematic diagram of the automatic compliance verification system for the warehousing scenario provided in the embodiments of this application, as shown below. Figure 2 As shown: Image acquisition module 201 is used to acquire environmental photos and close-up photos of invoices during the warehousing operation; Image recognition and analysis module 202 is used to perform image recognition on the environmental photo and determine whether the environmental photo contains the following three key elements at the same time: the operator's face area, the appearance of the item to be put into storage, and the reading display area of ​​the weighing equipment. The authorization and identity verification module 203 is used to extract the operator's facial features from the environmental photo and match them with a pre-established authorization identity feature database to determine whether the operator is an authorized person. The consistency comparison module 204 is used to perform OCR recognition on the close-up photo of the ticket, extract the product category and labeled weight information, and compare the extracted information with the actual reading of the weighing equipment and the product category entered in the system. The compliance verification module 205 is used to determine in real time that the current warehousing operation is non-compliant when any of the three key elements is missing, the matching judgment operator is not authorized, the comparison judgment weight deviation exceeds the preset threshold, or the categories are inconsistent; otherwise, the warehousing process is completed.

[0066] In one or more embodiments of this application, the process of establishing the authorized identity feature database in the authorized identity authentication module 203 includes: The system acquires multiple frames of facial images of authorized operators collected by the administrator, extracts facial feature vectors, and stores them in an encrypted manner. It also binds the employee's employee number, name, department, and authorization validity period to establish an authorized identity feature database. The addition, deletion, and modification operations of the authorized identity database can only be performed by the administrator account, and each authorization record has an validity period field. Expired or revoked authorizations are directly judged as non-compliant. In scenarios where terminals and the cloud are deployed collaboratively, the terminal and the cloud synchronize the authorization library on a periodic or event-triggered basis.

[0067] In one or more embodiments of this application, the authorization and authentication module 203 includes: A face feature vector extraction unit is used to detect face regions from the environmental photo and extract face feature vectors from each detected face region; The similarity calculation unit is used to calculate the similarity between the extracted facial feature vector and the feature vector corresponding to the valid authorization record in the authorized identity feature database; The identity matching unit determines that the identity matching is successful and the operator is an authorized person when the calculated highest similarity score exceeds a preset threshold; otherwise, it determines that the operation is performed by an unauthorized person.

[0068] In one or more embodiments of this application, the consistency comparison module 204 includes: The target field extraction unit is used to extract the target field from the OCR recognition result using at least one of the following methods: by triggering positioning through preset keywords, the text line immediately adjacent to the keyword is identified as the target field; or, by using the row and column position relationship of the text line through the table space proximity rule, the target field is extracted. The confidence score comparison unit is used to assign an identification confidence score to each extracted field. When the confidence score is lower than a preset threshold, the field is marked as requiring manual confirmation, and the operator is prompted to check on the operation interface. The verification and confirmation unit is used to mark the recognition result of the field as valid and use it for subsequent consistency comparison in response to the operator's verification and confirmation operation; or, in response to the operator's correction input, to replace the OCR recognition result with the corrected value for subsequent consistency comparison.

[0069] In one or more embodiments of this application, the release operation module is used to: The block will be automatically lifted once the operator retakes photos of the environment and tickets as prompted, and all re-performed checks pass. The block is lifted when the operator corrects the entered category or weight information within the preset limits, and the verification is successfully performed again after correction; The block will be lifted in response to the verification and confirmation operation performed by authorized administrators through the back-end management system or terminal.

[0070] In one or more embodiments of this application, it further includes: a recording module, used to record all blocking operations, unblocking operations and auditing operations, the recorded information including: triggering reason, operator identification, operation timestamp, and processing method; the records are used for subsequent auditing and tracing.

[0071] The system provided in this application embodiment can be used to execute the technical solutions of the above method embodiments. Its implementation principle and technical effect are similar, and will not be repeated here.

[0072] The system provided in this application is deployed locally on the intelligent weighing terminal, or it can be deployed in collaboration between the terminal and the cloud.

[0073] When deployed locally on a smart weighing terminal: all functional modules such as image acquisition, recognition, verification, and blocking control are integrated and installed on the smart weighing terminal at the canteen site, and the entire compliance verification process can be completed offline without relying on the cloud network.

[0074] When deployed collaboratively on the terminal and the cloud: the terminal is responsible for on-site image acquisition, local rapid recognition and process control, while the cloud is responsible for authorized identity database management, data storage, log auditing and remote monitoring. The terminal and the cloud automatically synchronize data according to the period or authorized change events, taking into account both on-site response speed and centralized management capabilities.

[0075] The two deployment methods can be flexibly adapted to different warehousing scenarios. Local deployment can run stably in environments without or with weak network, ensuring uninterrupted verification and faster response. Terminal + cloud collaborative deployment retains the real-time control capability on site, while also realizing unified authorization management, centralized data storage and full traceability, improving management efficiency and supervision strength. Overall, it has the advantages of strong compatibility, wide applicability and stability.

[0076] Exemplary device: Figure 3 This is a schematic diagram of the hardware structure of the electronic device provided in an embodiment of this application. Figure 3 As shown, the electronic device of this embodiment includes a processor 301 and a memory 302.

[0077] The memory 302 stores computer-executed instructions; the processor 301 executes the computer-executed instructions stored in the memory to implement the various steps performed by the electronic device in the above embodiments. For details, please refer to the relevant descriptions in the foregoing method embodiments.

[0078] Alternatively, the memory 302 can be either standalone or integrated with the processor 301.

[0079] When the memory 302 is set up independently, the electronic device also includes a bus 303 for connecting the memory 302 and the processor 301.

[0080] Exemplary media and products: This application also provides a computer storage medium storing computer execution instructions. When the processor executes the computer execution instructions, the above-mentioned automatic compliance verification method for the warehousing scenario is implemented.

[0081] This application also provides a computer program product, including a computer program, which, when executed by a processor, implements the above-described automatic compliance verification method for data entry scenarios.

[0082] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or modules, and may be electrical, mechanical, or other forms.

[0083] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to implement the solution of this embodiment according to actual needs.

[0084] Furthermore, the functional modules in the various embodiments of this application can be integrated into one processing unit, or each module can exist physically separately, or two or more modules can be integrated into one unit. The unit composed of the above modules can be implemented in hardware or in the form of hardware plus software functional units.

[0085] The integrated modules described above, implemented as software functional modules, can be stored in a computer-readable storage medium. These software functional modules, stored in a storage medium, include several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) or processor to execute some steps of the methods of the various embodiments of this application.

[0086] It should be understood that the aforementioned processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. A general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly manifested as execution by a hardware processor, or execution by a combination of hardware and software modules within the processor.

[0087] The memory may include high-speed RAM, and may also include non-volatile storage (NVM), such as at least one disk storage device, and may also be a USB flash drive, external hard drive, read-only memory, disk or optical disc, etc.

[0088] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, the buses shown in the accompanying drawings are not limited to a single bus or a single type of bus.

[0089] The aforementioned storage medium can be implemented using any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The storage medium can be any available medium accessible to a general-purpose or special-purpose computer. An exemplary storage medium is coupled to a processor, enabling the processor to read information from and write information to the storage medium. Of course, the storage medium can also be a component of the processor. The processor and storage medium can reside in application-specific integrated circuits (ASICs). Alternatively, the processor and storage medium can exist as discrete components in an electronic device or host device.

[0090] Those skilled in the art will understand that all or part of the steps of the above-described method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0091] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

Claims

1. A method for automatic compliance verification in an inbound scenario, characterized in that, The method includes: Collect environmental photos and close-up photos of invoices during the warehousing process; Image recognition is performed on the environmental photo to determine whether the environmental photo simultaneously contains the following three key elements: the operator's face area, the appearance of the items to be put into storage, and the reading display area of ​​the weighing equipment; The operator's facial features are extracted from the environmental photos and matched with a pre-established authorized identity feature database to determine whether the operator is an authorized person. The close-up photo of the ticket is subjected to OCR recognition to extract the product category and labeled weight information, and the extracted information is compared with the actual reading of the weighing equipment and the product category entered into the system. If any of the three key elements is missing, or the matching operator is not authorized, or the weight deviation exceeds a preset threshold or the product category is inconsistent, the current warehousing operation is determined to be non-compliant in real time, a prompt message is output and the warehousing process is blocked; otherwise, the warehousing process is completed.

2. The method according to claim 1, characterized in that, The process of establishing the authorized identity feature database includes: The system acquires multiple frames of facial images of authorized operators collected by the administrator, extracts facial feature vectors, and stores them in an encrypted manner. It also binds the employee's employee number, name, department, and authorization validity period to establish an authorized identity feature database. The addition, deletion, and modification operations of the authorized identity database can only be performed by the administrator account, and each authorization record has an validity period field. Expired or revoked authorizations are directly judged as non-compliant. In scenarios where terminals and the cloud are deployed collaboratively, the terminal and the cloud synchronize the authorization library on a periodic or event-triggered basis.

3. The method according to claim 1 or 2, characterized in that, The step of extracting the operator's facial features from the environmental photo and matching them with a pre-established authorized identity feature database to determine whether the operator is an authorized person includes: Detect face regions from the environmental photos, and extract face feature vectors from each detected face region; The similarity between the extracted facial feature vector and the feature vector corresponding to the valid authorization record in the authorized identity feature database is calculated. When the highest similarity score obtained exceeds the preset threshold, the identity is determined to be matched successfully, and the operator is an authorized person; otherwise, it is determined to be an unauthorized person.

4. The method according to claim 1, characterized in that, The process of performing OCR recognition on the close-up photo of the invoice to extract the product category and labeled weight information includes: The target field is extracted from the OCR recognition result using at least one of the following methods: by triggering localization through preset keywords, the text line immediately adjacent to the keyword is identified as the target field; or, by using the table space proximity rule, the target field is extracted using the row and column position relationship of the text line. Each extracted field is assigned a recognition confidence score. When the confidence score is lower than a preset threshold, the field is marked as requiring manual confirmation, and the operator is prompted to check it on the operation interface. In response to the operator's verification and confirmation operation, the recognition result of the field is marked as valid and used for subsequent consistency comparison; or, in response to the operator's correction input, the corrected value is used to replace the OCR recognition result for subsequent consistency comparison.

5. The method according to claim 1, characterized in that, After the data entry process is blocked, the data entry operation will resume when at least one of the following conditions is met: The block will be automatically lifted once the operator retakes photos of the environment and tickets as prompted, and all re-performed checks pass. The block is lifted when the operator corrects the entered category or weight information within the preset limits, and the verification is successfully performed again after correction; The block will be lifted in response to the verification and confirmation operation performed by authorized administrators through the back-end management system or terminal.

6. The method according to claim 5, characterized in that, Also includes: All blocking, unblocking, and auditing operations are recorded. The recorded information includes: trigger reason, operator identification, operation timestamp, and handling method. The records are used for subsequent auditing and tracing.

7. An automatic compliance verification system for warehousing scenarios, characterized in that, include: The image acquisition module is used to capture environmental photos and close-up photos of invoices during the warehousing operation. The image recognition and analysis module is used to perform image recognition on the environmental photo and determine whether the environmental photo simultaneously contains the following three key elements: the operator's face area, the appearance of the items to be put into storage, and the reading display area of ​​the weighing equipment. The authorization and identity verification module is used to extract the operator's facial features from the environmental photos and match them with a pre-established authorization and identity feature database to determine whether the operator is an authorized person. The consistency comparison module is used to perform OCR recognition on the close-up photo of the ticket, extract the product category and labeled weight information, and compare the OCR recognition information with the actual reading of the weighing equipment and the product category entered by the system. The compliance verification module is used to determine in real time that the current warehousing operation is non-compliant when any of the three key elements is missing, the matching judgment operator is not authorized, the comparison judgment weight deviation exceeds the preset threshold, or the categories are inconsistent; output prompt information and block the warehousing process. Otherwise, complete the warehousing process.

8. The system according to claim 7, characterized in that, The system is deployed locally on the intelligent weighing terminal, or it can be deployed in collaboration between the terminal and the cloud.

9. An electronic device, characterized in that, include: At least one processor; The system also includes a memory communicatively connected to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, the computer program being executed by the at least one processor to cause the at least one processor to perform the automatic compliance verification method for the inbound scenario as described in any one of claims 1 to 6.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the automatic compliance verification method for the inbound scenario as described in any one of claims 1 to 6.

Images