A cloud database-based anti-counterfeiting holographic paper authentication information management method and system

Abstract

The application provides a cloud database-based anti-counterfeiting holographic paper authentication information management method and system, applied to the anti-counterfeiting authentication technical field, by establishing the association of identity identification and physical anti-counterfeiting features, and preprocessing the physical features into first feature data and second feature data with different complexities, hierarchical verification is realized. The identity identification is acquired and it is judged whether it is in a state to be checked; only when it is in the state, real-time physical features are collected and first feature data is extracted, and first-level local comparison is performed on a terminal or an edge node. If the matching degree is between preset thresholds, the data is uploaded to the cloud, and the server uses high-precision second feature data to perform second-level comparison to determine the final result. Finally, the identity identification state is dynamically updated according to the verification result, and access control is performed on subsequent requests. Therefore, the scheme has the beneficial effects of reducing the cloud computing load, guaranteeing the verification experience of legal users, and effectively intercepting illegal requests.

Description

Technical Field

[0001] This application relates to the field of anti-counterfeiting authentication technology, and in particular to a method and system for managing anti-counterfeiting holographic paper authentication information based on cloud database. Background Technology

[0002] In the field of product anti-counterfeiting, to counterfeiters' attacks that use genuine product digital identity codes for large-scale counterfeiting, existing authentication systems have been upgraded to a composite mechanism combining spatiotemporal information analysis and secondary physical feature verification. Specifically, when the system detects an anomaly through spatiotemporal conflicts (such as the same verification code being queried from different locations within a short period), it marks the identity code as suspicious and triggers secondary verification: requiring the user to photograph the microscopic physical features of the anti-counterfeiting holographic paper and perform a high-precision comparison with pre-stored standard samples in the cloud, using this as the final basis for judgment.

[0003] However, this mechanism faces a significant technical challenge: when counterfeiters launch a large-scale copy-paste attack, a massive number of counterfeit products simultaneously trigger a massive number of secondary verification requests. A copy-paste attack involves attackers purchasing genuine products, copying their real digital identification codes, and mass-printing them on counterfeit products. When consumers scan the codes, the system returns a successful verification because the genuine code is used, leading to the user being deceived, and the mechanism relying solely on digital identification codes thus fails.

[0004] The massive volume of secondary verification requests leads to complex image processing and pattern recognition calculations, resulting in a surge in cloud server resource consumption, prolonged response times, and even service interruptions. Simultaneously, the system needs to dynamically update the identity code status to ensure that legitimate users' subsequent queries after the initial verification are not misjudged as suspicious, guaranteeing a good user experience, while also accurately identifying and blocking subsequent illegal verification requests.

[0005] Therefore, the urgent technical problem to be solved is: when faced with a massive influx of secondary verification requests, how to optimize the processing flow to reduce server computing load, ensure response speed, and establish a reliable state transition mechanism to balance the user experience of genuine products with the effectiveness of blocking illegal requests, thereby maintaining the robustness and data consistency of the entire anti-counterfeiting system. Summary of the Invention

[0006] In view of the shortcomings of the prior art, this application provides a cloud database-based method and system for managing anti-counterfeiting holographic paper authentication information. It has the advantages of effectively reducing the consumption of computing resources on cloud servers, avoiding extended verification response time and service interruption, establishing a clear and reliable mechanism to dynamically update the status of digital identity codes, ensuring legitimate user experience, and effectively distinguishing and blocking subsequent illegal verification requests to maintain system robustness and data consistency.

[0007] Firstly, a method for managing anti-counterfeiting holographic paper authentication information based on a cloud database, wherein the method establishes a correlation between identity information and physical anti-counterfeiting features, the physical anti-counterfeiting features being pre-processed into first feature data and second feature data, and the method includes the following steps: S1: Obtain the target identity information and determine whether the target identity information is in a pending verification state; S2: When the target identity information is in a state of pending verification, the terminal device collects real-time physical anti-counterfeiting features and extracts real-time first feature data; S3: Execute the first-level comparison logic on the terminal device or edge node to match the real-time first feature data with the pre-stored first feature data and calculate the first matching degree; S4: If the first matching degree is between the first preset threshold and the second preset threshold, the data to be verified is uploaded to the cloud server, and the cloud server uses the second feature data to perform the second-level comparison logic to determine the final verification result; S5: Based on the verification results of the first-level comparison logic or the second-level comparison logic, dynamically update the management status of the target identity information, and perform access control on subsequent verification requests for the identity information based on the management status.

[0008] Furthermore, step S1 includes: S11: Obtain a verification request for the target identity information, the verification request including the current spatiotemporal parameters; S12: Perform consistency analysis between the spatiotemporal parameters and the historical verification records of the target identity information to determine whether the target identity information is in a pending verification state.

[0009] Furthermore, step S12 includes: S121: Obtain the historical location information and historical time information of the most recent successful verification in the historical verification record; S122: Determine the current location information and current time information based on the current spatiotemporal parameters; S123: Calculate the geographical distance between the current location information and the historical location information, and the time difference between the current time information and the historical time information; S124: Determine the movement rate based on the ratio of the geographical distance to the time difference; S125: If the movement rate is greater than a preset logical threshold, the target identity information is determined to be in a state pending verification.

[0010] Furthermore, step S3 includes the following: S31: If the first matching degree is lower than the first preset threshold, it is determined that the verification fails and the subsequent verification process is terminated; S32: If the first matching degree is higher than the second preset threshold, it is determined that the verification is successful.

[0011] Furthermore, step S4 includes: S41: When the first matching degree is between the first preset threshold and the second preset threshold, the terminal device performs data compression processing on the collected real-time physical anti-counterfeiting features, generates data to be verified, and uploads it to the cloud server. S42: The cloud server uses the second feature data to perform a second-level comparison logic to determine the final verification result.

[0012] Furthermore, step S42 includes: S421: The cloud server calculates the second matching degree between the data to be verified and the pre-stored second feature data; S422: If the second matching degree is higher than the third preset threshold, the verification is deemed successful; S423: If the second matching degree is not higher than the third preset threshold, the verification is deemed unsuccessful.

[0013] Furthermore, step S5 includes: S51: Based on the verification result of the first-level comparison logic or the second-level comparison logic, when the verification result is that the verification is passed, the management status of the target identity information is changed from the pending verification status to the normal verification status. S52: When the verification result is that the verification fails, the management status of the target identity information is changed from the pending verification status to the abnormal lock status; S53: If the management status is updated to an abnormal lock status, the verification process will be directly intercepted and terminated for subsequent verification requests for the same target identity information, and the terminal device will not be triggered to collect real-time physical anti-counterfeiting features. S54: If the management status is updated to normal verification status, record the device identification information and geographical location information of this verification request, and use them as trust reference data for the target identity information.

[0014] Furthermore, prior to step S1, the following is included: S01: Obtain the original image data of the physical anti-counterfeiting feature; S02: Extract the first feature data and the second feature data with different computational complexities from the original image data; S03: The identity information is associated with and stored with the first feature data and the second feature data, and the first feature data is synchronized to the terminal device or the edge node for distributed caching.

[0015] Furthermore, step S2 includes: S21: Send a secondary verification activation command to the terminal device to enable the image acquisition guidance function of the terminal device; S22: Obtain a real-time image containing the physical anti-counterfeiting features through the terminal device, and preprocess the real-time image to extract the real-time first feature data.

[0016] Secondly, a cloud database-based anti-counterfeiting holographic paper authentication information management system, the system comprising: Acquisition module: Acquires target identity information and determines whether the target identity information is in a pending verification state; Data Acquisition Module: When the target identity information is in a state of pending verification, the terminal device acquires real-time physical anti-counterfeiting features and extracts real-time first feature data; Calculation module: Executes first-level comparison logic on the terminal device or edge node, matches the real-time first feature data with the pre-stored first feature data, and calculates the first matching degree; Verification module: If the first matching degree is between the first preset threshold and the second preset threshold, the data to be verified is uploaded to the cloud server, and the cloud server uses the second feature data to perform the second-level comparison logic to determine the final verification result; Update control module: Based on the verification results of the first-level comparison logic or the second-level comparison logic, dynamically update the management status of the target identity information, and perform access control on subsequent verification requests for the identity information based on the management status.

[0017] Beneficial Effects: This application proposes a cloud database-based method and system for managing anti-counterfeiting holographic paper authentication information. By establishing a link between identity identifiers and physical anti-counterfeiting features, and preprocessing the physical features into first and second feature data of varying complexity, hierarchical verification is achieved. In terms of process, the identity identifier is first acquired and its status as pending verification is determined. Only when in this status is the terminal collecting real-time physical features and extracting low-complexity first feature data, performing a first-level local comparison at the terminal or edge node. If the matching degree is within a preset threshold, the data is uploaded to the cloud, where the server uses high-precision second feature data for a second-level comparison to determine the final result. Finally, the identity identifier status is dynamically updated based on the verification result, and subsequent requests are subject to access control. This scheme significantly reduces the cloud computing load through local screening, avoiding response delays and service interruptions. Simultaneously, through dynamic status management, it ensures a reliable verification experience for legitimate users while effectively blocking illegal requests, improving system robustness and data consistency. Attached Figure Description

[0018] Figure 1 This is a flowchart illustrating a cloud database-based method for managing anti-counterfeiting holographic paper authentication information.

[0019] Figure 2 This is a structural diagram of a cloud database-based anti-counterfeiting holographic paper authentication information management system proposed in this application.

[0020] Figure 3 This is a schematic diagram of a cloud database-based anti-counterfeiting holographic paper authentication information management system proposed in this application.

[0021] Labeling Explanation: 201. Acquisition Module; 202. Data Acquisition Module; 203. Calculation Module; 204. Verification Module; 205. Update Control Module. Detailed Implementation

[0022] 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 a part of the embodiments of this application, and not all of the embodiments. The components of the embodiments of this application described and marked in the accompanying drawings can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0023] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0024] Please refer to Figure 1 To address large-scale copy-paste attacks while ensuring a reliable verification experience for legitimate users and establishing a clear and reliable mechanism for dynamically managing identity status, this application proposes a cloud database-based method for managing authentication information on anti-counterfeiting holographic paper. The core idea of ​​this method is to move away from centralizing all complex verification tasks on cloud servers and instead establish a distributed, layered intelligent verification system. By decomposing the inherently difficult-to-copy physical characteristics of anti-counterfeiting holographic paper into two levels of computational complexity, and utilizing the computing power of terminal devices or edge nodes for preliminary pre-screening, the method effectively intercepts the vast majority of crude counterfeits, distributing the computational burden from the cloud to the edge. Only requests that have passed the preliminary screening and have a high probability of authenticity are uploaded to the cloud for final detailed comparison, thereby significantly reducing the computational load on cloud servers and ensuring stable operation under large-scale attacks and a smooth experience for legitimate users.

[0025] Specifically, this method establishes a correlation between identity information and physical anti-counterfeiting features. The physical anti-counterfeiting features are pre-processed into first feature data and second feature data. The method includes the following steps: S1: Obtain the target identity information and determine whether the target identity information is in a pending verification state; S2: When the target identity information is in a state of pending verification, the terminal device collects real-time physical anti-counterfeiting features and extracts real-time first feature data; S3: Execute the first-level comparison logic on the terminal device or edge node, match the real-time first feature data with the pre-stored first feature data, and calculate the first matching degree; S4: If the first matching degree is between the first preset threshold and the second preset threshold, the data to be verified will be uploaded to the cloud server, and the cloud server will use the second feature data to execute the second-level comparison logic to determine the final verification result. S5: Based on the verification results of the first-level comparison logic or the second-level comparison logic, dynamically update the management status of the target identity information, and perform access control on subsequent verification requests for the identity information according to the management status.

[0026] Before the entire authentication information management process begins, each product bearing the anti-counterfeiting holographic paper label must first be registered, establishing a strong binding relationship between its unique digital identity and physical characteristics. This process forms the foundation for all subsequent verification activities. Specifically, on the product production line, when an anti-counterfeiting holographic paper label is affixed to a product, a series of initialization operations are performed.

[0027] Specifically, before step S1, the following are included: S01: Obtain the original image data of the physical anti-counterfeiting features; S02: Extract first and second feature data with different computational complexities from the original image data; S03: Associate and store the identity information with the first feature data and the second feature data, and synchronize the first feature data to the terminal device or edge node for distributed caching.

[0028] First, step S01 is completed using high-precision imaging equipment. For example, an industrial-grade camera equipped with a CMOS sensor of over 20 million pixels can be used, along with a macro lens, to photograph the anti-counterfeiting holographic paper label under controlled lighting conditions, such as using a shadowless ring LED light source, at a fixed focal length and angle. The image resolution is set to an extremely high standard, such as at least 5,000 pixels per millimeter, to ensure that the microscopic structural details in the holographic pattern that are difficult to distinguish with the naked eye can be clearly captured.

[0029] Next, from the acquired raw image data, first and second feature data with different computational complexities are extracted. This step is the core of the hierarchical verification strategy, which aims to decompose a complex physical feature into two different levels of digital fingerprints.

[0030] The first feature data, also known as edge feature summary, is a set of features with low computational cost and lightweight data size. Its extraction process prioritizes efficiency and compactness. In a specific embodiment, the captured high-resolution image can first undergo image preprocessing, including grayscale conversion and contrast enhancement. Subsequently, the Oriented Fast and Rotated Brief (ORB) algorithm is applied to extract several key points and their corresponding descriptors in specific regions of the holographic pattern, such as the center and four corners. These descriptors are then processed into a compact feature vector; for example, the descriptors generated by ORB are themselves binary strings. The resulting first feature data typically has a data size of tens to hundreds of kilobytes, making it ideal for processing on terminal devices with limited network bandwidth or weak computing power. As a specific feasible method for anti-counterfeiting holographic paper, the extraction process of the first feature data is as follows: First, the high-resolution original image is converted to grayscale, and Gaussian filtering is used to remove high-frequency noise. Then, the Oriented Fast Rotation Brief (ORB) algorithm is applied to extract local features in the four corner regions and the central region of the holographic pattern (a total of five fixed-size 512×512 pixel sub-regions). Specifically, the upper limit of the number of feature points in the ORB algorithm is set to 500, the number of scale pyramid layers is 8, and the scale factor is 1.2. For each detected feature point, a 32-byte binary descriptor is generated. The descriptors of all feature points are concatenated in coordinate order to form a compact binary vector with a length not exceeding 16KB, which is the first feature data. This data is suitable for rapid comparison using Hamming distance in terminal devices or edge nodes.

[0031] In another embodiment, the first feature data can be obtained by performing a Haar wavelet transform on the grayscale image of the holographic pattern, retaining only the lowest frequency LL subband (compressed to 1 / 8 of the original image size), and then binarizing the coefficients of this subband to form an image hash fingerprint. The second feature data is obtained by acquiring multi-angle diffraction spot images of holographic paper under laser illumination, and extracting the morphological moments (Hu moments) and Fourier descriptors of the spots as the first feature vector, which are then stored in a cloud database.

[0032] The second feature data, also known as cloud-based detailed features, is a set of high-precision, large-volume features used for final accurate comparison in the cloud. Its extraction process strives for comprehensiveness and uniqueness. In one embodiment, deeper analysis of the holographic image can be performed based on the extracted first feature data. For example, under controlled lighting conditions, three original images of the anti-counterfeiting holographic paper are acquired using a macro lens, corresponding to annular light source incident angles of 0°, 60°, and 120°, respectively, with each image having a resolution of 2048×2048 pixels. Then, a two-dimensional Fast Fourier Transform (FFT) is performed on each image to obtain its spectrogram. On the spectrogram, with the center as the origin, the amplitude spectrum within annular regions with radii between 50 and 200 pixels is extracted and unfolded into a one-dimensional vector. The amplitude spectrum vectors at the three angles are concatenated end-to-end to form a floating-point feature vector of length (3×150) 450 dimensions. Subsequently, this vector is input into a pre-trained lightweight convolutional neural network (using the MobileNetV2 architecture, with the classification layer removed, and an output dimension of 128) to extract deep semantic features, ultimately yielding a 128-dimensional floating-point feature vector, which serves as the second feature data. This data is stored on a cloud server for high-precision cosine similarity comparison.

[0033] In another, more complex embodiment, the extraction of the second feature data can be combined with a deep learning network. Specifically, the high-resolution original image can be segmented into multiple tiny regions, and local binary pattern or gradient orientation histogram features can be extracted for each region. These features are then combined with high-dimensional feature vectors extracted by the feature extraction layers of a pre-trained convolutional neural network, such as ResNet or EfficientNet. The resulting second feature data, potentially reaching several megabytes in size, can capture microscopic details in the holographic pattern with extremely high precision that are almost impossible to replicate.

[0034] The difference in computational complexity between the first and second feature data mainly lies in the data dimensionality and the computational load of the comparison algorithm. The first feature data employs lightweight feature extraction algorithms, such as a simplified binary orientation fast rotation algorithm, to transform the physical holographic pattern into a short vector composed of zeros and ones. This feature comparison process involves only XOR operations and bit counting, resulting in low computational complexity, making it suitable for rapid execution on mobile terminals or edge nodes with limited computing power. The second feature data, on the other hand, uses high-dimensional floating-point feature vectors, such as semantic features with 512 or higher dimensions extracted through deep convolutional neural networks, or frequency domain diffraction patterns obtained through Fourier transform. This feature comparison involves large-scale matrix operations, cosine similarity calculations, or complex phase matching algorithms, typically resulting in computational complexity two orders of magnitude higher than the first feature, requiring collaborative computation by a cluster of graphics processing units on a cloud server.

[0035] As a specific implementation method, during the initial registration phase of the anti-counterfeiting label, a high-precision imaging device acquires multiple frames of original images of the holographic paper under different illumination angles. From these images, the microscopic texture distribution of the holographic pattern surface is extracted using a local binary mode algorithm. Due to the compact data structure and fast comparison speed of the texture descriptor, it is defined as the first feature data. Simultaneously, a deep residual network is used to perform feature dimensionality reduction on the interference fringes of the hologram, extracting a two-dimensional vector matrix representing optical depth information as the second feature data. During storage, the identity identifier is used as the primary key, and the first feature data is pushed to a discretely distributed cache database so that it can be read by edge nodes at microsecond speeds during verification. The second feature data, which is massive in volume and contains core secrets, is stored in a cloud-based core database protected by a hardware encryption module. This distributed storage strategy achieves a balance between performance and security, ensuring that the edge has sufficient reference data for initial screening while ensuring that the core biometric physical fingerprint does not leave the cloud security domain, thereby preventing the risk of the anti-counterfeiting system collapsing due to large-scale data leakage.

[0036] Finally, the identity information is associated with and stored in conjunction with the extracted first and second feature data, and the first feature data is synchronized to terminal devices or edge nodes for distributed caching. Specifically, the product's unique identity information, such as the string corresponding to a QR code, along with its corresponding first and second feature data, is uploaded to the cloud via a secure socket layer encrypted communication protocol. In the cloud database, a distributed object storage service can be used to store the original high-precision image and the massive amount of second feature data to meet the storage needs of massive amounts of data. The identity information and the smaller amount of first feature data can be stored in a relational database with indexes for quick retrieval. After registration, the system synchronizes the first feature data to edge computing nodes around the world via a content delivery network. Simultaneously, when a user first uses the verification application or when the application is updated, a portion of the first feature data related to the user's geographical location is pushed through a secure channel to the local storage of the user's smartphone or other terminal device, such as the database within the application's sandbox, and is updated periodically to provide a data foundation for subsequent offline or rapid comparisons.

[0037] Through the above registration process, a complete file containing a digital identity, a lightweight physical fingerprint, and a high-precision physical fingerprint is created for each genuine product, and the lightweight fingerprint is pre-deployed at the edge of the network.

[0038] When a user needs to verify the authenticity of a product, the entire verification process is initiated. The first step is to obtain the target's identity information and determine if that information is in a pending verification state. This step is the entry point and first line of defense for the entire layered verification system. Its purpose is to quickly identify abnormal verification behavior caused by copy-paste attacks by analyzing the movement patterns of the items.

[0039] Furthermore, step S1 includes: S11: Obtain a verification request for the target identity information. The verification request includes the current spatiotemporal parameters. S12: Perform consistency analysis based on the spatiotemporal parameters and historical verification records of the target identity information to determine whether the target identity information is in a pending verification state.

[0040] Specifically, in step S11, when a user uses a smartphone or other terminal device to scan the QR code on the anti-counterfeiting label using its built-in scanning function, the application on the terminal device automatically obtains the device's GPS coordinates, i.e., longitude and latitude, as well as the current system timestamp. This information, along with the identity information decoded from the QR code, is packaged into a data request, such as a JSON-formatted data packet, and sent to the cloud authentication service interface via a secure HTTPS protocol.

[0041] Subsequently, in step S12, after receiving the request, the cloud-based authentication service first checks the database to see if the identity information exists. If it exists, it further retrieves all verification records of the identity information over a period of time, such as the last thirty days. These records contain the geographical location and timestamp of each verification.

[0042] To perform consistency analysis, it is necessary to quantitatively evaluate the spatiotemporal relationship between the current verification request and historical verification records. Specifically, step S12 includes: S121: Obtain the historical location and time information of the most recent successful verification from the historical verification records; S122: Determine the current location information and current time information based on the current spatiotemporal parameters; S123: Calculate the geographical distance between the current location information and the historical location information, as well as the time difference between the current time information and the historical time information; S124: Determine the movement rate based on the ratio of geographical distance to time difference; S125: If the movement speed is greater than the preset logical threshold, the target identity information is determined to be in a state of pending verification.

[0043] The logical threshold is used for spatiotemporal consistency analysis and can be determined based on the physical limits of actual logistics transportation. It calculates the fastest possible movement speed of normal goods from production to sales (e.g., no more than 900 km / h for air express), and, considering a certain margin, sets the logical threshold to 1000 km / h or a more conservative 800 km / h. The system also allows administrators to manually adjust this threshold based on industry experience.

[0044] In step S121, selecting the most recent successful record as the comparison benchmark ensures that the historical data used for analysis is valid and reflects the item's most recent true state. Then, based on the spatiotemporal parameters included in the current request, the current location information and current time information are determined.

[0045] In step S123, the geographical distance can be calculated using the Havers formula, which calculates the shortest distance between two locations on the Earth's surface based on their latitude and longitude. For example, assuming the latitude and longitude of the two verifications are lat1,lon1 and lat2,lon2 respectively, and the average radius of the Earth is R, the distance calculation process is as follows: first, calculate the square of the sine of half the difference in latitude and longitude; then, calculate the central angle between the two points using inverse trigonometric functions; finally, multiply by the Earth's radius to obtain the final distance.

[0046] After obtaining the geographical distance and time difference, the movement rate is determined based on the ratio of these two values. This movement rate, in a physical sense, represents the average speed at which the item moves between the two verifications.

[0047] Finally, in step S125, the logical threshold is set based on common sense and physical laws. For example, it could be set to 800 kilometers per hour, a speed far exceeding the maximum speed of conventional land or water transport. If the calculated movement rate exceeds this threshold, it means that the identity identifier could not possibly have moved such a long distance in such a short time. This is highly likely because the identity identifier has been copied and used on counterfeit items in different locations. In this case, the authentication service will mark the identity identifier information in the database as pending verification. If the identity identifier is in the initial state or normal verification state, and the item's movement pattern is normal, the verification result will be returned directly to the terminal.

[0048] In summary, spatiotemporal parameter consistency analysis is a process of identifying logical conflicts by comparing the physical spatial displacement and time intervals in the current verification request with those in historical verification records. Specifically, the latitude and longitude coordinates and timestamp of the last successful verification recorded in the cloud database are retrieved for the target identity information. The spherical distance between the current request's latitude and longitude and the historical latitude and longitude is calculated using the Havers formula, and the time difference between the current and historical times is also calculated. Dividing the calculated spherical distance by the time difference yields the average movement rate of the identity during the two verification periods. If this rate exceeds the maximum physical speed limit of conventional transportation such as commercial airliners or high-speed trains, the identity is considered to have an unreasonable jump in spatial distribution, resulting in a failed consistency analysis and the identity being placed in a pending verification state.

[0049] As a specific implementation method, in the process of tracking goods in cross-border trade, the consistency analysis of spatiotemporal parameters can be combined with a more complex logical judgment model. The obtained verification requests not only include latitude and longitude, but also the service set identifier of the network access point and the cell identifier of the base station. In the consistency analysis process, a dynamic physical reachability model is established instead of a single rate threshold. If an identity appears for verification at a boutique in country A at 2:00 AM, and then appears for verification at a duty-free shop in country B three hours later, although this time span allows for the possibility of cross-border flight, a comprehensive judgment is made by combining the time pattern of commercial operations, the actual flight time of international flights, and the customs clearance cycle of goods. If the time difference is much smaller than the shortest possible physical flow cycle, it is judged as an abnormal movement rate. In the analysis process, it is also considered whether the identity is frequently accessed by a large number of devices from different network segments in a short period of time. Such clustered verification requests will be identified as a potential copying attack pattern. Through this multi-dimensional spatiotemporal consistency analysis, the risk of identity being illegally cloned can be warned earlier, and the first round of logical screening can be completed before physical anti-counterfeiting measures are intervened.

[0050] If the target identity information is determined to be in a pending verification state, it means that simple digital identity verification is insufficient to determine authenticity, and a higher-level physical feature verification process must be initiated. At this point, the process enters the second step: the terminal device collects real-time physical anti-counterfeiting features and extracts real-time first feature data.

[0051] Specifically, step S2 includes: S21: Send a secondary verification activation command to the terminal device to enable the image acquisition guidance function on the terminal device; S22: Obtain a real-time image containing physical anti-counterfeiting features through a terminal device, and preprocess the real-time image to extract real-time first feature data.

[0052] In step S21, the cloud authentication service sends a secondary verification activation command to the requesting terminal device to enable the image acquisition guidance function. This command triggers the application on the terminal device to enter a special working mode. From the user's perspective, the application interface changes; for example, a prompt box will pop up stating: "This product has an abnormal verification record. Please take a hologram for in-depth verification." Simultaneously, the application activates the device's camera and displays a real-time viewfinder on the screen. This viewfinder may include a rectangle, alignment marks, or corner indicators to guide the user to accurately place a specific area of ​​the anti-counterfeiting hologram within the frame, ensuring uniform lighting and a clear image. The application can utilize the device's built-in image signal processor for autofocus, auto exposure, and white balance adjustments to maximize the quality of the captured image.

[0053] Guided by the user, the terminal device acquires a real-time image containing physical anti-counterfeiting features and preprocesses it to extract real-time first feature data. After the user takes the picture, the terminal device does not immediately upload the raw image, which may be several megabytes in size. Instead, it utilizes its own computing power, such as the device's digital signal processor or neural network processing unit, to run a lightweight feature extraction algorithm. This algorithm corresponds to the one used to extract the first feature data during the registration phase. For example, if the registration phase used a fast orientation and a simplified rotation algorithm, the terminal device will also run the same algorithm to extract key points and their descriptors from the newly captured real-time image, forming a compact set of real-time first feature data. This process transforms a complex image into a digital fingerprint with a very small data volume, preparing it for subsequent edge pre-screening.

[0054] After successfully extracting the real-time first feature data on the terminal device, the process enters the most innovative part of this method: executing the first-level comparison logic on the terminal device or edge node, matching the real-time first feature data with the pre-stored first feature data, and calculating the first matching degree. This step is also known as terminal edge pre-screening, and its core purpose is to utilize distributed computing resources to perform a fast and efficient filtering of counterfeit products before the data is uploaded to the cloud.

[0055] The terminal device compares the newly extracted real-time first feature data with pre-stored first feature data corresponding to the current identity information, either cached locally or obtained from the nearest edge node via a low-latency network. The comparison method can be chosen based on the type of the first feature data. For example, if the feature data is a binary string, Hamming distance can be used to calculate its difference; if the feature data is a feature vector, cosine similarity can be used to calculate its matching degree. Cosine similarity is calculated by dividing the dot product of the two vectors by the product of their magnitudes. The result ranges from -1 to 1; the closer the value is to 1, the more similar the two vectors are.

[0056] The calculated matching result, i.e., the first matching degree, is compared with two preset thresholds: the first preset threshold and the second preset threshold. These two thresholds together define a three-interval judgment logic.

[0057] Specifically, step S3 is followed by: S31: If the first matching degree is lower than the first preset threshold, it is determined that the verification fails and the subsequent verification process is terminated; S32: If the first matching degree is higher than the second preset threshold, the verification is deemed successful.

[0058] This involves collecting at least 1000 images of genuine anti-counterfeiting holographic paper under various actual shooting conditions (different light intensities, angles, focal lengths, and shooting equipment models), and at least 1000 images of counterfeit products. Real-time first feature data is extracted from each image and compared with corresponding pre-stored first feature data to calculate the first matching degree. A receiver operating characteristic curve (ROC curve) is plotted, using the false rejection rate (misclassifying genuine products as counterfeit) and the false acceptance rate (misclassifying counterfeit products as genuine) as indicators. The minimum matching degree when the false rejection rate is below 1% is set as the first preset threshold, and the minimum matching degree when the false acceptance rate is below 0.1% is set as the second preset threshold. This method ensures that crude counterfeit products can be safely intercepted below the low threshold, and genuine products can be safely confirmed above the high threshold.

[0059] The first preset threshold is a low threshold, for example, set to 0.3. If the calculated first matching degree is lower than this value, it means that the physical features collected in real time differ significantly from the feature summary pre-stored for genuine products. This strongly suggests that the product in the user's possession is a poorly made counterfeit. In this case, the terminal device will immediately make a judgment, display a warning message to the user on the screen indicating verification failure or counterfeit, and, crucially, terminate the entire verification process, refusing to upload any data, including the original image or extracted features, to the cloud server. This mechanism can directly intercept the vast majority of counterfeit verification requests at the edge, greatly saving network bandwidth and valuable cloud computing resources.

[0060] On the other hand, if the first matching degree is higher than the second preset threshold, the verification is considered successful. The second preset threshold is an extremely high threshold, for example, set to 0.7. If the calculated first matching degree is higher than this value, it means that the physical features collected in real time are highly consistent with the feature summary pre-stored on the genuine product, which can almost be considered a perfect match. In this case, the terminal device will directly determine that it is a genuine product and send a lightweight confirmation request to the cloud. This request only contains identity information and a borderline verification pass flag. After receiving this request, the cloud will directly update the status of the identity without performing subsequent detailed comparisons. This mechanism provides users holding genuine products with a fast and smooth verification experience, while also avoiding unnecessary cloud resource consumption.

[0061] When the result of the first-level comparison is neither a significant mismatch nor a perfect match—that is, when the first matching degree is between the first and second preset thresholds, for example, within the range of 0.3 to 0.7—this indicates that the current product may be genuine. However, due to factors such as shooting angle, lighting conditions, or the high precision of counterfeit production, a 100% certain judgment cannot be made based solely on the first feature data. Such issues require submission to the cloud for a higher-level adjudication.

[0062] When the first matching degree falls between the first and second preset thresholds, it represents a fuzzy judgment range. The first preset threshold represents the boundary between obviously counterfeit and suspicious products. If the matching degree is lower than this value, it indicates that the collected features are fundamentally different from the standard template, and it is directly judged as counterfeit. The second preset threshold represents the absolute boundary for confirming a genuine product. If the matching degree is higher than this value, it indicates that the feature overlap is extremely high, and no further verification is needed. When the matching degree is between the two, it means that the collected image may have been affected by environmental factors such as light intensity, shooting angle shift, and background noise interference, causing the lightweight feature extraction algorithm to be unable to give a definitive conclusion, or the counterfeit product's manufacturing process has reached a certain level of simulation, capable of simulating some macroscopic features. In this case, to ensure the fairness and accuracy of the verification, a higher-dimensional second-level comparison logic must be introduced for final adjudication.

[0063] As a specific implementation method, in the automated inbound process of a large-scale warehousing and logistics center, edge nodes manifest as industrial computing units deployed alongside the sorting line. When batches of goods bearing holographic anti-counterfeiting paper pass through the line, multi-angle high-sensitivity scanners mounted above the track quickly identify the QR code identification on each package. If the cloud reports that the identification is pending verification due to recent frequent cross-regional verifications, a high-speed industrial camera on the line immediately captures an optical image of the holographic paper. The edge computing unit receives the real-time image and uses a built-in lightweight feature point detection operator to extract the coordinate distribution information of key visual anchor points in the holographic pattern as real-time first feature data. Subsequently, the edge computing unit retrieves the pre-stored first feature data corresponding to the product from its local cache for rapid Euclidean distance calculation, thereby obtaining the first matching degree. In this high-concurrency industrial scenario, local comparison through edge nodes can achieve initial screening within tens of milliseconds. If the first matching degree is extremely low, the edge computing unit controls a robotic arm to directly remove the suspicious package from the normal flow line. This approach avoids network bandwidth congestion caused by uploading high-resolution images of thousands of packages to the cloud simultaneously, making full use of the warehouse's local computing resources and ensuring the continuity and efficiency of the automated logistics chain.

[0064] At this point, the process moves to the fourth step: the data to be verified is uploaded to the cloud server, and the cloud server uses the second feature data to execute the second-level comparison logic to determine the final verification result.

[0065] Specifically, step S4 includes: S41: When the first matching degree is between the first preset threshold and the second preset threshold, the terminal device performs data compression processing on the collected real-time physical anti-counterfeiting features, generates data to be verified, and uploads it to the cloud server. S42: The cloud server uses the second feature data to perform the second-level comparison logic to determine the final verification result.

[0066] In step S41, the real-time physical anti-counterfeiting feature typically refers to the previously captured original high-resolution image. To reduce the burden on network transmission, the terminal device compresses the image before uploading, for example, using a lossy compression format such as JPEG, or employing a more advanced algorithm that can efficiently compress the image while ensuring that key feature information is not lost. The compressed data, along with the identity information, is uploaded to the cloud server as the data to be verified.

[0067] Subsequently, in step S42, the cloud server uses the pre-stored second feature data to perform the second-level comparison logic to determine the final verification result. After receiving the data to be verified uploaded from the terminal, the cloud server retrieves the corresponding second feature data stored during the registration phase from its distributed object storage based on the identity information; that is, the detailed cloud features.

[0068] The cloud server utilizes its powerful computing resources, such as high-performance graphics processing unit clusters, to execute complex image processing and pattern recognition algorithms. Specifically, step S42 includes: S421: The cloud server calculates the second matching degree between the data to be verified and the pre-stored second feature data; S422: If the second matching degree is higher than the third preset threshold, the verification is deemed successful; S423: If the second matching degree is not higher than the third preset threshold, the verification is deemed unsuccessful.

[0069] This process involves collecting at least 500 high-precision images of genuine holographic paper under optimal shooting conditions, extracting their second feature data, and comparing them with pre-stored second feature data in the cloud to calculate the distribution of the second matching degree. The matching degree corresponding to the mean of this distribution minus three times the standard deviation is taken as the third preset threshold to ensure that the false judgment rate of the final cloud verification is less than 0.3%.

[0070] In step S421, the calculation process is far more complex and precise than the edge-end comparison. For example, a deep learning-based comparison method can be used, where the uploaded image is input into a pre-trained, large-scale convolutional neural network, such as ResNet-152 or InceptionV3, to extract its high-dimensional feature vector, and then compared with the pre-stored second feature data vector using cosine similarity. Alternatively, a physical optical feature-based comparison method can be used, where the uploaded image undergoes a Fourier transform to analyze the spectral differences between its spectrum and the pre-stored detailed features; or a phase correlation method can be used to accurately match the microstructure of the holographic pattern and calculate its cross-correlation peak.

[0071] After calculating the second matching degree, a final judgment is made. If the second matching degree is higher than a preset third threshold, the verification is considered successful. This third preset threshold is a very strict and precise matching threshold, for example, set to 0.95. If the comparison result is higher than this value, the product can be confirmed as genuine. Conversely, if the second matching degree is not higher than the third preset threshold, the verification is considered unsuccessful, and the product is confirmed as counterfeit.

[0072] After obtaining the final verification result through the first or second level of comparison logic, the process proceeds to step S5: dynamically updating the management status of the target identity information, and implementing access control for subsequent verification requests targeting that identity information based on the updated management status. This step establishes a closed-loop management mechanism, ensuring the robustness of the verification system and the consistency of the data.

[0073] Furthermore, step S5 includes: S51: Based on the verification result of the first-level comparison logic or the second-level comparison logic, when the verification result is successful, change the management status of the target identity information from the pending verification status to the normal verification status. S52: When the verification result is that the verification fails, the management status of the target identity information is changed from the pending verification status to the abnormal lock status; S53: If the management status is updated to an abnormal lock status, the verification process will be directly intercepted and terminated for subsequent verification requests for the same target identity information, and the terminal device will not be triggered to collect real-time physical anti-counterfeiting features. S54: If the management status is updated to normal verification status, record the device identification information and geographical location information of this verification request, and use them as trust reference data for the target identity information.

[0074] In step S51, whether the product passes the initial screening with a high score or is confirmed through detailed comparison in the cloud, once it is verified as genuine, its identity identifier in the database will change from a temporary pending verification state to a normal verification state. This ensures that legitimate users will not be misjudged as suspicious in subsequent normal verifications, thus protecting their user experience.

[0075] Meanwhile, to further enhance system security, when the management status is updated to normal verification status, the device identification information and geographical location information of this verification request are recorded and used as trust reference data for the target identity information. The device identification information can be a hashed string that uniquely represents the terminal device. This trust reference data can be used as an auxiliary basis for future judgments. For example, a trust domain can be established, and if subsequent verification requests originate from the same device or a nearby geographical location, the threshold for judging them as abnormal can be appropriately relaxed, thereby making the verification logic more intelligent.

[0076] On the other hand, when the verification result is that the verification fails, the management status of the target identity information is changed from pending verification to abnormal lock status. This means that once a product is confirmed to be counterfeit at any verification stage, its corresponding identity information will be permanently marked as abnormal lock.

[0077] This state change is irreversible. If the management state is updated to an abnormally locked state, any subsequent verification requests for the same target identity information will be directly intercepted and the verification process terminated, without triggering the terminal device to collect real-time physical anti-counterfeiting features. Specifically, when a new verification request arrives at the cloud, during the initial query phase, once the authentication service detects that the identity's state is abnormally locked, it will immediately terminate all subsequent processing and directly return a clear warning message to the requesting terminal, such as: "This product has been confirmed as counterfeit; please purchase with caution." This interception mechanism can be implemented at the gateway level of the authentication service, without forwarding the request to the backend spatiotemporal analysis or comparison service. This achieves a permanent block on confirmed counterfeit identity identities with minimal resource consumption, effectively preventing their continued circulation in the market.

[0078] As a specific implementation method, in the verification scenario for the secondhand circulation of luxury goods, the dynamic updating of the management status involves the process of establishing a trust domain. When the holographic tag of a luxury bag passes the second-level high-precision comparison in the cloud, the update control module not only changes the status of the identity identifier from the pending verification state to the normal verification state, but also associates and stores the fingerprint of the terminal device that successfully verified the device, the user account weight, and the geofence information, forming a set of trust reference data. In subsequent verification requests, if the request comes from a device or location within the same trust domain, its normal verification state is maintained first. Conversely, if the second-level comparison determines that the verification fails, the update control module will immediately change the status of the identity identifier to an abnormal locked state. At this time, the identifier will be marked with an unremovable blacklist tag in the global index of the cloud database. When other users scan this locked QR code again in the secondhand trading market, the authentication request will be directly intercepted when it reaches the application gateway level. This access control mechanism does not require the backend to perform any image comparison or spatiotemporal analysis again, and directly informs the user that the product has been confirmed as counterfeit. This state-based interception mechanism greatly reduces the response cost to repeated attacks, achieves permanent bans through a single accurate determination, and builds a reliable digital defense in complex secondary market environments.

[0079] Through the above series of steps, the method proposed in this application not only effectively addresses the computational pressure brought about by large-scale concurrent attacks through layered verification, but also establishes a dynamic learning and self-reinforcing anti-counterfeiting authentication system through refined state management and strict access control. While ensuring the legitimate user experience, it delivers a precise and powerful blow to counterfeiting behavior.

[0080] Please refer to Figure 2 , Figure 3 This application also provides a cloud database-based anti-counterfeiting holographic paper authentication information management system, the system comprising: Module 201: Acquires target identity information and determines whether the target identity information is in a pending verification state; Collection module 202: When the target identity information is in a state of pending verification, the terminal device collects real-time physical anti-counterfeiting features and extracts real-time first feature data; Calculation module 203: Executes the first-level comparison logic on the terminal device or edge node, matches the real-time first feature data with the pre-stored first feature data, and calculates the first matching degree; Verification module 204: If the first matching degree is between the first preset threshold and the second preset threshold, the data to be verified is uploaded to the cloud server, and the cloud server uses the second feature data to execute the second-level comparison logic to determine the final verification result; Update control module 205: Based on the verification results of the first-level comparison logic or the second-level comparison logic, dynamically update the management status of the target identity information, and perform access control on subsequent verification requests for the identity information based on the management status.

[0081] This application proposes a cloud database-based anti-counterfeiting holographic paper authentication information management system. Its core is to realize the association management of identity information and physical anti-counterfeiting features through the collaborative work of various functional modules, and to achieve efficient and reliable anti-counterfeiting authentication through hierarchical comparison and dynamic status management.

[0082] Specifically, the acquisition module 201 can be understood as a hardware unit, software module, or a combination thereof that implements the above functions. For example, the acquisition module can be configured as a software service integrated into the terminal device, responsible for listening to user input or scanning events and calling local storage or remote query interfaces to obtain identity information and its status. As another implementation, the acquisition module can also be an independent microservice deployed on an edge node, receiving requests from the terminal device through an API interface and performing preliminary status judgments.

[0083] The acquisition module 202 can be implemented as an image acquisition and processing unit in a terminal device. For example, the acquisition module can use the high-resolution camera built into the terminal device to capture images and extract real-time first feature data through a pre-built image processing algorithm library (such as OpenCV). Alternatively, the acquisition module can also be a stand-alone hardware device, such as a dedicated optical scanner that integrates an image sensor and a feature extraction processor, communicating with the terminal device via wired or wireless means.

[0084] The computing module 203 can be deployed on terminal devices or edge nodes. For example, the computing module can be a lightweight comparison algorithm library integrated into the application of the terminal device, utilizing the device's CPU or GPU resources for fast feature matching. As a preferred implementation, the computing module can also be deployed on edge nodes, connecting to multiple terminal devices to centrally process real-time first feature data from these devices, thereby achieving more efficient resource utilization and faster response speed.

[0085] The verification module 204 is primarily responsible for high-precision comparison in the cloud. For example, the verification module could be a high-performance computing cluster deployed on a cloud server, using deep learning models or complex pattern recognition algorithms to accurately compare the uploaded data to be verified with pre-stored second feature data. Furthermore, the verification module is also responsible for managing the second feature data stored in the cloud database, ensuring the security and integrity of the data.

[0086] The update control module 205 can be a core business logic service deployed on a cloud server. For example, the update control module can maintain an identity information status database and update the management status of each identity information (such as normal verification status, abnormal lock status) in real time based on the verification results. Simultaneously, this module is also responsible for implementing access control policies. For instance, when an identity information is in an abnormal lock status, all subsequent verification requests are directly blocked, and no further comparison processes are triggered, thereby effectively preventing malicious attacks. As a preferred implementation, the update control module can also be integrated with a log service to record all status changes and access control events for auditing and analysis.

[0087] The cloud database-based anti-counterfeiting holographic paper authentication information management system proposed in this application demonstrates significant innovation and technological advancement in addressing the problem of massive secondary verification request surges faced by existing technologies. Traditional anti-counterfeiting systems, upon detecting anomalies, often directly redirect all suspicious requests to the cloud for high-precision secondary verification. This one-size-fits-all approach, when facing large-scale copy-paste attacks, leads to a sharp increase in the computational load on cloud servers, prolonged response times, and may even cause service interruptions, severely impacting system availability and user experience.

[0088] The core innovation of this application lies in its modular system design, which implements hierarchical comparison logic and dynamic state management mechanism. First, through the collaborative work of the acquisition module 201, the collection module 202, and the calculation module 203, the system preprocesses physical anti-counterfeiting features into first and second feature data with different computational complexities, and designs a first-level comparison logic executed on terminal devices or edge nodes. This innovation allows most preliminary, computationally less demanding comparison tasks to be completed locally. Only when the first matching degree is in an uncertain range is the data to be verified handed over to the verification module 204 for uploading to the cloud for second-level comparison. Compared to the closest existing technology, this application's system significantly reduces the computational pressure on the cloud server and the network bandwidth requirements, thereby ensuring the system's response speed and stability when facing massive requests. For example, in a large-scale attack, if millions of counterfeit products simultaneously trigger second-level verification, existing technology may cause the cloud to crash; while this application's system can quickly filter out most counterfeits locally through distributed first-level comparison, uploading only a small number of questionable requests to the cloud, greatly alleviating cloud pressure.

[0089] Secondly, the update control module 205 of this application can dynamically update the management status of the target identity information and perform access control on subsequent verification requests based on the management status, which is another important innovation. Although existing technologies can also identify anomalies and trigger secondary verification, they often lack fine-grained and dynamic system status management. The system of this application can dynamically update the identity information to a normal verification status or an abnormal locked status based on the verification result. When the identity information is locked, subsequent verification requests for that identity information will be directly intercepted by the update control module without any further comparison, thereby effectively preventing attackers from using locked identity information for repeated probing or large-scale attacks. This mechanism not only improves system security but also avoids unnecessary consumption of computing resources. At the same time, for legitimate requests that pass verification, the system records trust reference data to ensure that legitimate users have a smooth experience during subsequent queries and avoid misjudgments.

[0090] In summary, the system of this application optimizes the anti-counterfeiting authentication process through hierarchical comparison and dynamic status management, effectively solving the performance bottlenecks and security vulnerabilities of existing technologies when handling massive secondary verification requests, and significantly improving the robustness, efficiency, and user experience of the anti-counterfeiting system, demonstrating outstanding technological progress and practical value.

[0091] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A cloud database-based anti-counterfeiting holographic paper authentication information management method, characterized in that, The method establishes a correlation between identity information and physical anti-counterfeiting features, wherein the physical anti-counterfeiting features are pre-processed into first feature data and second feature data. The method includes the following steps: S1: Obtain the target identity information and determine whether the target identity information is in a pending verification state; S2: When the target identity information is in a state of pending verification, the terminal device collects real-time physical anti-counterfeiting features and extracts real-time first feature data; S3: Execute the first-level comparison logic on the terminal device or edge node to match the real-time first feature data with the pre-stored first feature data and calculate the first matching degree; S4: If the first matching degree is between the first preset threshold and the second preset threshold, the data to be verified is uploaded to the cloud server, and the cloud server uses the second feature data to perform the second-level comparison logic to determine the final verification result; S5: Based on the verification results of the first-level comparison logic or the second-level comparison logic, dynamically update the management status of the target identity information, and perform access control on subsequent verification requests for the identity information based on the management status.

2. The anti-counterfeiting holographic paper authentication information management method based on a cloud database according to claim 1, characterized in that, Step S1 includes: S11: Obtain a verification request for the target identity information, the verification request including the current spatiotemporal parameters; S12: Perform consistency analysis between the spatiotemporal parameters and the historical verification records of the target identity information to determine whether the target identity information is in a pending verification state.

3. The cloud database-based anti-counterfeiting holographic paper authentication information management method according to claim 2, characterized in that, Step S12 includes: S121: Obtain the historical location information and historical time information of the most recent successful verification in the historical verification record; S122: Determine the current location information and current time information based on the current spatiotemporal parameters; S123: Calculate the geographical distance between the current location information and the historical location information, and the time difference between the current time information and the historical time information; S124: Determine the movement rate based on the ratio of the geographical distance to the time difference; S125: If the movement rate is greater than a preset logical threshold, the target identity information is determined to be in a state pending verification.

4. The cloud database-based anti-counterfeiting holographic paper authentication information management method according to claim 1, characterized in that, Step S3 is followed by: S31: If the first matching degree is lower than the first preset threshold, it is determined that the verification fails and the subsequent verification process is terminated; S32: If the first matching degree is higher than the second preset threshold, it is determined that the verification is successful.

5. The cloud database-based anti-counterfeiting holographic paper authentication information management method according to claim 1, characterized in that, Step S4 includes: S41: When the first matching degree is between the first preset threshold and the second preset threshold, the terminal device performs data compression processing on the collected real-time physical anti-counterfeiting features, generates data to be verified, and uploads it to the cloud server. S42: The cloud server uses the second feature data to perform a second-level comparison logic to determine the final verification result.

6. The cloud database-based anti-counterfeiting holographic paper authentication information management method according to claim 1, characterized in that, Step S42 includes: S421: The cloud server calculates the second matching degree between the data to be verified and the pre-stored second feature data; S422: If the second matching degree is higher than the third preset threshold, the verification is deemed successful; S423: If the second matching degree is not higher than the third preset threshold, the verification is deemed unsuccessful.

7. The cloud database-based anti-counterfeiting holographic paper authentication information management method according to claim 1, characterized in that, Step S5 includes: S51: Based on the verification result of the first-level comparison logic or the second-level comparison logic, when the verification result is that the verification is passed, the management status of the target identity information is changed from the pending verification status to the normal verification status. S52: When the verification result is that the verification fails, the management status of the target identity information is changed from the pending verification status to the abnormal lock status; S53: If the management status is updated to an abnormal lock status, the verification process will be directly intercepted and terminated for subsequent verification requests for the same target identity information, and the terminal device will not be triggered to collect real-time physical anti-counterfeiting features. S54: If the management status is updated to normal verification status, record the device identification information and geographical location information of this verification request, and use them as trust reference data for the target identity information.

8. The cloud database-based anti-counterfeiting holographic paper authentication information management method according to claim 1, characterized in that, Before step S1, the following are included: S01: Obtain the original image data of the physical anti-counterfeiting feature; S02: Extract the first feature data and the second feature data with different computational complexities from the original image data; S03: The identity information is associated with and stored with the first feature data and the second feature data, and the first feature data is synchronized to the terminal device or the edge node for distributed caching.

9. The cloud database-based anti-counterfeiting holographic paper authentication information management method according to claim 1, characterized in that, Step S2 includes: S21: Send a secondary verification activation command to the terminal device to enable the image acquisition guidance function of the terminal device; S22: Obtain a real-time image containing the physical anti-counterfeiting features through the terminal device, and preprocess the real-time image to extract the real-time first feature data.

10. A cloud database based anti-counterfeit holographic paper authentication information management system, characterized in that, The system includes: Acquisition module: Acquires target identity information and determines whether the target identity information is in a pending verification state; Data Acquisition Module: When the target identity information is in a state of pending verification, the terminal device acquires real-time physical anti-counterfeiting features and extracts real-time first feature data; Calculation module: Executes first-level comparison logic on the terminal device or edge node, matches the real-time first feature data with the pre-stored first feature data, and calculates the first matching degree; Verification module: If the first matching degree is between the first preset threshold and the second preset threshold, the data to be verified is uploaded to the cloud server, and the cloud server uses the second feature data to perform the second-level comparison logic to determine the final verification result; Update control module: Based on the verification results of the first-level comparison logic or the second-level comparison logic, dynamically update the management status of the target identity information, and perform access control on subsequent verification requests for the identity information based on the management status.

Images