A two-dimensional code anti-counterfeiting intelligent verification system

Abstract

The application relates to the field of information security and anti-counterfeiting technology and discloses a two-dimensional code anti-counterfeiting intelligent verification system which comprises a code assignment management module, a physical fingerprint registration module and an offline verification module. The code assignment management module is used for generating a two-dimensional code load containing a digital signature and a product serial number and transmitting the two-dimensional code load to a printing device to execute a printing instruction. The physical fingerprint registration module collects a microtexture feature vector of a two-dimensional code area during the execution of the printing instruction and binds the microtexture feature vector with the product serial number to generate a preset fingerprint template. The offline verification module is used for acquiring a real-time microtexture and performing feature alignment by referring to the preset fingerprint template. The physical fingerprint registration and feature alignment verification scheme is adopted, the printing microtexture is extracted as a unique identifier, the double binding of the digital signature and the physical feature is realized, the effect of distinguishing the original and the copied copy is achieved, and the problem that the existing static graphics are easily cloned by high-precision photographing or copying is solved.

Description

Technical Field

[0001] This invention relates to the field of information security and anti-counterfeiting technology, specifically a QR code anti-counterfeiting intelligent verification system. Background Technology

[0002] Against the backdrop of the current global supply chain digital transformation, QR code technology, with its significant advantages such as large storage capacity, fast reading speed, low production cost, and ease of mobile interaction, has become a link connecting physical goods and the digital world. In the fields of brand protection, product traceability, and anti-counterfeiting, assigning codes to product packaging and combining this with backend database verification has become the mainstream security measure. However, due to the widespread adoption of high-precision optical acquisition equipment and digital image processing technology, counterfeiting methods are becoming increasingly high-tech, integrated, and chain-like. Traditional QR code anti-counterfeiting systems are facing unprecedented technological challenges and security crises, as follows:

[0003] First, existing QR code anti-counterfeiting technologies mainly rely on static encoded graphics, making the anti-counterfeiting carrier extremely easy to physically clone through high-precision photography, copying, or screen scanning. Since the backend system can only recognize the information content carried by the graphic and cannot distinguish whether the graphic originates from the original product or is an illegally copied copy, the anti-counterfeiting verification fails at the physical level.

[0004] Second, traditional QR code verification systems rely heavily on real-time data interaction between mobile terminals and backend servers. If a user is in a scenario with no network signal, such as a cross-border transportation warehouse, underground storage room, or remote port, the database information cannot be retrieved for verification. The system's heavy reliance on real-time network connectivity means that the anti-counterfeiting verification system cannot provide security in certain environments.

[0005] Third, in the entire product lifecycle circulation process, customs, consumers and brands have different data requirements. Moreover, the existing QR code verification system adopts a full data display mode and lacks permission distribution logic, which leads to the exposure of sensitive business information in unnecessary links. It cannot protect the identity privacy and trade secrets of all parties while meeting the requirements of regulatory audits.

[0006] Fourth, current verification solutions are mostly passive feedback based on single scanning behavior, lacking comprehensive correlation analysis of scanning time, geographical location, and device fingerprints. When the same QR code is scanned abnormally on a large scale in different locations, it is impossible to identify the potential counterfeit cluster behavior in time and issue an early warning, making it difficult for brands to accurately crack down on chain-like counterfeit dens.

[0007] To address the aforementioned issues, this invention proposes a QR code anti-counterfeiting intelligent verification system. By combining physical micro-textures and digital signatures through dual binding, along with offline incremental updates of material packages and hierarchical desensitization output, it achieves physical anti-counterfeiting and privacy protection across all environments, overcoming the shortcomings of QR codes being easily cloned and failing under weak network conditions. Summary of the Invention

[0008] To address the shortcomings of existing technologies, this invention provides a QR code anti-counterfeiting intelligent verification system to solve the problems mentioned in the background section.

[0009] To achieve the above objectives, the present invention provides the following technical solution: a QR code anti-counterfeiting intelligent verification system, comprising:

[0010] The coding management module is used to generate a QR code payload containing a digital signature and product serial number and transmit it to the printing equipment to execute printing instructions;

[0011] The physical fingerprint registration module collects the micro-texture feature vector of the QR code area during the printing instruction execution process, and associates and binds the micro-texture feature vector with the product serial number to generate a preset fingerprint template.

[0012] The offline verification module is used to acquire real-time micro-textures and perform feature alignment using a preset fingerprint template. At the same time, it calls a local algorithm to perform integrity verification on the digital signature in the QR code payload to obtain a judgment result.

[0013] The risk assessment module is used to extract the real-time environmental parameters of the terminal from the judgment results and substitute them into the multi-dimensional risk function for calculation to obtain the abnormal probability for the current scanning behavior.

[0014] The hierarchical output module is used to reference the anomaly probability and perform information desensitization processing on the judgment result according to the preset verification subject role permissions, generating verification result receipts of different dimensions.

[0015] The cloud scheduling module is used to aggregate product serial numbers, pre-set fingerprint templates, and anomaly probabilities, generate offline material packages containing verification rules through cluster analysis, and distribute the offline material packages to the offline verification module.

[0016] Preferably, the anomaly probability calculation performed in the risk assessment module is determined by the comprehensive confidence score formula:

[0017] ,

[0018] in, To calculate the overall confidence score, For normalized mapping functions, To verify the weight of digital signatures, The feature distance between the real-time microtexture and the pre-set fingerprint template. For real-time microtexturing, For pre-set fingerprint templates, To preset dynamic threshold, For a multidimensional risk function, , , These are the weighting coefficients;

[0019] The multidimensional risk function The independent variable is defined as: For scanning timestamps, For geographical coordinates, The scanning frequency per unit time. For mobile terminal hardware identification, This is to verify the spatiotemporal distribution pattern of the same product serial number in history.

[0020] Preferably, the microtexture feature vector in the physical fingerprint registration module refers to the nonlinear distribution characteristics of the QR code pixel edges captured by the image sensor due to printing pressure and medium penetration.

[0021] Preferably, in the hierarchical output module, the information desensitization processing refers to masking or digesting the original data fields in the judgment result according to the role and permissions of the verification subject. When the role and permissions are consumer, the verification result receipt retains the authenticity judgment conclusion.

[0022] In the hierarchical output module, the verification subject role permissions include consumer role, regulatory subject role, logistics node role and brand management role, and the data disclosure dimensions of the verification result receipt corresponding to each role are different.

[0023] Preferably, in the cloud scheduling module, the offline material package is limited to a public key certificate containing an asymmetric encryption algorithm, a feature digest of the pre-set fingerprint template, and a risk blacklist data set identified by the clustering analysis;

[0024] In the cloud scheduling module, the clustering analysis refers to extracting the anomaly probability reported by several mobile terminals. When multiple verification requests are detected for the same product serial number within a time period exceeding a preset spatial span, it is determined to be a cluster fraud risk.

[0025] Preferably, in the coding management module, the digital signature is a binary data stream generated after signing the product serial number with a private key using an asymmetric encryption algorithm.

[0026] Preferably, in the offline verification module, feature alignment refers to extracting topological feature points from the real-time microtexture and calculating the affine transformation parameters of the topological feature points relative to the corresponding feature points in the preset fingerprint template.

[0027] Preferably, in the offline verification module, the structural similarity index between the real-time microtexture after feature alignment and the preset fingerprint template is calculated. To assist in the determination, the logic for determining the legality of the real-time micro-texture is expressed as follows:

[0028] ,

[0029] in, For real-time microtexturing, For pre-set fingerprint templates, The preset similarity threshold constant is used when the calculation result is... When this happens, the determination result is marked as a physical failure state.

[0030] Preferably, the offline verification module has an ambient light compensation function. During the acquisition of the real-time micro-texture, the offline verification module calls the flash hardware of the mobile terminal to perform color temperature compensation lighting, and uses a homomorphic filtering algorithm to perform brightness uniformization processing on the acquired image to eliminate the feature point extraction deviation caused by uneven ambient lighting.

[0031] Preferably, in the cloud scheduling module, the offline material package is distributed using an asynchronous incremental update mechanism. The cloud scheduling module calculates the feature difference packet between the offline data version currently held by the mobile terminal and the latest version on the cloud. By limiting the upper limit of the number of bytes in the offline material package, the data synchronization consistency of the offline verification module in a weak network environment is ensured.

[0032] This invention provides a QR code anti-counterfeiting intelligent verification system. It has the following beneficial effects:

[0033] 1. This invention adopts a physical fingerprint registration and feature alignment verification scheme. By extracting printed micro-textures as unique identifiers, it achieves dual binding of digital signatures and physical features, thereby distinguishing between original documents and photocopied copies and solving the problem that existing static graphics are easily photographed or cloned with high precision.

[0034] 2. This invention adopts an offline verification and asynchronous incremental update material package scheme. By pre-setting the verification algorithm and fingerprint digest on the terminal and dynamically synchronizing differential data, it realizes autonomous verification in scenarios without network signal, solving the shortcomings of traditional systems that rely on real-time network connection, which makes the verification function prone to failure.

[0035] 3. This invention adopts a role-based access control and structured minimum disclosure scheme. It de-identifies the judgment data based on the identity of the verification subject, realizes on-demand disclosure and differentiated display of sensitive information, and solves the shortcomings of the full data display mode that leads to the leakage of trade secrets and inadequate privacy protection.

[0036] 4. This invention adopts a multi-dimensional risk assessment and cloud-based clustering analysis scheme. By associating the spatial and temporal information of barcode scanning, device fingerprints, and serial number trajectories to identify cluster characteristics, it realizes the transformation from passive feedback to proactive identification of counterfeit dens, solving the problem that the lack of correlation analysis in single verification makes it difficult to crack down on the counterfeit chain. Attached Figure Description

[0037] Figure 1 This is a schematic diagram of the system framework of the present invention. Detailed Implementation

[0038] To enable those skilled in the art to understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort should fall within the scope of protection of the present invention.

[0039] The present invention will now be described in detail with reference to the accompanying drawings:

[0040] A QR code anti-counterfeiting intelligent verification system includes:

[0041] The coding management module is used to generate a QR code payload containing a digital signature and product serial number and transmit it to the printing equipment to execute printing instructions;

[0042] In the code management module, the digital signature is a binary data stream generated after signing the product serial number with a private key using an asymmetric encryption algorithm;

[0043] The physical fingerprint registration module collects the micro-texture feature vector of the QR code area during the printing instruction execution process, and associates and binds the micro-texture feature vector with the product serial number to generate a preset fingerprint template.

[0044] In the physical fingerprint registration module, the micro-texture feature vector refers to the nonlinear distribution characteristics of QR code pixel edges captured by the image sensor due to printing pressure and medium penetration;

[0045] The offline verification module is used to acquire real-time micro-textures and perform feature alignment using a preset fingerprint template. At the same time, it calls a local algorithm to perform integrity verification on the digital signature in the QR code payload to obtain a judgment result.

[0046] In the offline verification module, feature alignment refers to extracting topological feature points from real-time micro-textures and calculating the affine transformation parameters of the topological feature points relative to the corresponding feature points in the preset fingerprint template.

[0047] In the offline verification module, the structural similarity index between the real-time micro-texture after feature alignment and the pre-set fingerprint template is calculated. To assist in the determination, the logic for determining the legality of real-time micro-textures is expressed as follows:

[0048] ,

[0049] in, For real-time microtexturing, For pre-set fingerprint templates, The preset similarity threshold constant is used when the calculation result is... When this happens, the judgment result is marked as a physical failure state.

[0050] The offline verification module has an ambient light compensation function. In the process of acquiring real-time micro-textures, the offline verification module calls the flash hardware of the mobile terminal to perform color temperature compensation lighting, and uses the homomorphic filtering algorithm to perform brightness uniformization processing on the acquired image to eliminate the feature point extraction deviation caused by uneven ambient light.

[0051] The risk assessment module is used to extract the real-time environmental parameters of the terminal from the judgment results and substitute them into the multi-dimensional risk function for calculation to obtain the abnormal probability for the current scanning behavior.

[0052] The anomaly probability calculation performed in the risk assessment module is determined by the comprehensive confidence score formula:

[0053] ,

[0054] in, To calculate the overall confidence score, For normalized mapping functions, To verify the weight of digital signatures, The feature distance between the real-time microtexture and the pre-set fingerprint template. For real-time microtexturing, For pre-set fingerprint templates, To preset dynamic threshold, For a multidimensional risk function, , , These are the weighting coefficients;

[0055] Multidimensional risk function The independent variable is defined as: For scanning timestamps, For geographical coordinates, The scanning frequency per unit time. For mobile terminal hardware identification, To verify the spatiotemporal distribution pattern of the same product serial number in history;

[0056] The hierarchical output module is used to reference the anomaly probability and perform information desensitization processing on the judgment result according to the preset verification subject role permissions, generating verification result receipts of different dimensions.

[0057] In the hierarchical output module, information desensitization refers to masking or digesting the original data fields in the judgment result according to the role and permissions of the verification subject. When the role and permissions are consumer, the verification result receipt retains the true or false judgment conclusion.

[0058] In the tiered output module, the permissions of the verification subjects include consumer roles, regulatory subjects, logistics node roles, and brand management roles. The data disclosure dimensions of the verification result receipts for each role are different.

[0059] The cloud scheduling module is used to aggregate product serial numbers, pre-set fingerprint templates, and anomaly probabilities, generate offline material packages containing verification rules through cluster analysis, and distribute the offline material packages to the offline verification module.

[0060] In the cloud scheduling module, the offline material package is limited to a public key certificate containing an asymmetric encryption algorithm, a feature digest of a pre-set fingerprint template, and a risk blacklist data set identified by cluster analysis;

[0061] In the cloud scheduling module, cluster analysis refers to extracting the probability of anomalies reported by several mobile terminals. When multiple verification requests for the same product serial number are detected within a time period exceeding the preset spatial span, it is determined to be a risk of cluster fraud.

[0062] In the cloud scheduling module, the offline material package is issued using an asynchronous incremental update mechanism. The cloud scheduling module calculates the feature difference package between the offline data version currently held by the mobile terminal and the latest version on the cloud. By limiting the upper limit of the number of bytes in the offline material package, the data synchronization consistency of the offline verification module in a weak network environment is ensured.

[0063] The code management module embeds digital signatures and product serial numbers into the QR code payload, ensuring the legality and uniqueness of the data payload from the source. This provides reliable original data support for physical binding and logical verification, enhancing the security of anti-counterfeiting information generation.

[0064] The physical fingerprint registration module collects the micro-textures naturally formed during the QR code printing process and generates a pre-set fingerprint template, giving each QR code uncopyable physical and biological characteristics, blocking the possibility of physical cloning through high-precision scanning or copying, and improving the anti-copying strength of the carrier.

[0065] The offline verification module performs feature alignment and digital signature verification locally, enabling the verification process to operate independently of a real-time network connection. This ensures complete verification functionality even in shielded environments such as underground warehouses or cross-border logistics, greatly expanding the application scenarios and reliability of anti-counterfeiting verification.

[0066] The risk assessment module calculates multi-dimensional risk functions by inputting real-time environmental parameters, enabling a quantitative assessment of the security of a single scanning behavior. It can identify abnormal operational behaviors hidden behind legitimate data, upgrading the anti-counterfeiting logic from static information comparison to dynamic behavior monitoring.

[0067] The tiered output module performs information anonymization processing on the judgment results according to the subject's permissions, realizing differentiated data disclosure for different roles. While meeting the needs of regulatory audits and consumer traceability, it effectively protects the brand's trade secrets and the privacy of all parties involved.

[0068] The cloud scheduling module, through the asynchronous distribution of cluster analysis and offline material packages, enables real-time aggregation of global risk data and dynamic updates of local verification rules, constructing a closed loop linking cloud intelligence and terminal verification, significantly improving the system's early warning and defense capabilities against clustered fraudulent activities.

[0069] Example 1: Anti-cloning verification scenario for high-end consumer products

[0070] This embodiment details how the system solves the problem of high-precision copying and cloning by using the uniqueness of physical characteristics.

[0071] During the coding stage before the product leaves the factory, the coding management module generates a QR code payload containing a digital signature and a unique serial number, and instructs the high-precision printing equipment to perform coding.

[0072] At the moment of printing, the physical fingerprint registration module uses an embedded macro camera to capture multiple frames of the paper fiber structure and the microscopic nonlinear edge formed by ink penetration in the QR code area. By extracting the edge topological feature vector, a preset fingerprint template is generated and uniquely bound to the product serial number in the cloud database.

[0073] When a product enters the market, counterfeiters use high-precision scanning equipment to clone the appearance of the QR code and affix it to counterfeit products. When consumers verify the product, their mobile devices activate an offline verification module. This module captures real-time micro-texture images using the phone's camera and references a corresponding pre-set fingerprint template from a locally cached offline material package. The system uses a topological feature point extraction algorithm to calculate the affine transformation parameters of the real-time image relative to the template to achieve precise alignment. Because copying cannot reproduce the microscopic physical noise of printing, the edge texture of the cloned product will exhibit obvious smoothing or secondary interference features, causing the structural similarity index to fail to reach the preset threshold. Based on this, the offline verification module determines that the physical characteristics do not match. Combined with the digital signature verification result, it ultimately confirms that the product is an illegally copied copy, effectively intercepting physical-level cloning attacks.

[0074] Example 2: Cross-border Logistics Shielding Environment Verification Scenario

[0075] This embodiment details the system's data synchronization and autonomous verification logic in scenarios without network signal.

[0076] In bonded warehouses or underground storage centers for cross-border logistics, network signal coverage is often poor. To address this, the mobile verification terminals used by regulatory personnel, when in a network environment, have already completed data pre-installation via a cloud-based scheduling module. Based on regulatory task instructions, the cloud scheduling module calculates and distributes offline material packages containing specific batch public key certificates, fingerprint digests, and verification rules. When regulatory personnel enter the shielded area to conduct random inspections of goods, the offline verification module operates entirely in a local environment, without needing to access a remote server.

[0077] During the verification process, the terminal first verifies the integrity of the digital signature in the QR code, then invokes the ambient light compensation function. This function uses a flash with a fixed color temperature and a homomorphic filtering algorithm to eliminate ambient shadow interference and accurately extract real-time micro-textures. Since the verification algorithm is integrated locally on the terminal, the system can provide immediate feedback on the judgment result.

[0078] After verification, all scanning records, geographical location information, and anomaly probability data are temporarily stored in the terminal's secure area. Once the terminal reconnects to the external network, the cloud scheduling module will immediately trigger an asynchronous incremental update mechanism, upload the verification receipts from the offline period, and issue differential feature packets based on the latest global risk trends to ensure the continuity of verification behavior and the timeliness of data, thus resolving the bottleneck of anti-counterfeiting verification's reliance on real-time network connectivity.

[0079] Example 3: Brand Omnichannel Tiered Management and Cluster Risk Control Scenario

[0080] This embodiment details how the system utilizes big data analytics to identify chain-like counterfeiting while protecting the privacy of all parties. Throughout the product's entire lifecycle, distributors at all levels, logistics providers, and end consumers all participate in barcode verification. The risk assessment module continuously extracts environmental parameters for each scan, including terminal hardware identifiers, scan geographic coordinates, and timestamps. The cloud scheduling module performs cluster analysis on the aggregated massive anomaly probabilities. If it detects that a product serial number receives multiple verification requests in different geographic areas thousands of kilometers apart within a short period, or that the same physical device scans a large number of discontinuous serial numbers within a short period, the system will automatically determine this as an abnormal cluster scan and identify it as a counterfeiting operation.

[0081] The tiered output module plays a crucial role when entities with different permissions view the verification results. For brand management roles, the system outputs full spatiotemporal distribution charts and risk level assessments to assist in precise anti-counterfeiting efforts. For logistics node roles, it displays authorized product flow information for compliance verification. For ordinary consumers, the tiered output module performs strict information anonymization, masking or summarizing sensitive fields such as underlying supply chain trajectories and channel prices, and displaying the authenticity determination in the receipt. This mechanism utilizes big data to proactively warn against counterfeit groups and protects corporate trade secrets and user privacy through a structured disclosure mechanism, ensuring system security in a multi-entity collaborative environment.

[0082] Summary of Examples: Through in-depth practice of the above three examples, this invention comprehensively verifies the superior performance and logical closed-loop of a QR code anti-counterfeiting intelligent verification system in complex real-world scenarios. A detailed summary is as follows:

[0083] Example 1 powerfully demonstrates the effectiveness of the system by dual-binding arbitrary microscopic features of printing with an asymmetric encryption algorithm. This architecture, which anchors digital identity to physical features, breaks through the limitations of traditional QR codes that rely on graphic information for reading from the underlying logic, ensuring a strong coupling between the verification object and the original carrier. Experimental results show that the system can accurately identify nonlinear distribution features that even the most precise physical cloning methods cannot simulate, completely solving the industry problem of anti-counterfeiting carriers being easily copied illegally.

[0084] Example 2 verifies the reliability of the offline verification module and asynchronous incremental update mechanism through simulations of shielded environments and weak network scenarios. By deploying localized algorithms and scheduling differential feature packets, the system successfully eliminates its absolute dependence on real-time network conditions for anti-counterfeiting verification. This design ensures that the anti-counterfeiting service maintains a high level of response speed and judgment accuracy even in extreme environments such as cross-border logistics and underground warehousing, significantly improving the system's industrial applicability and full-scenario coverage.

[0085] Example 3 demonstrates the system's high flexibility in handling the needs of complex stakeholders. Through quantitative analysis of multidimensional risk functions and cloud-based clustering algorithms, the system achieves a leap from "single-point passive verification" to "global proactive early warning," effectively identifying chain-like and organized fraudulent activities. Simultaneously, combined with the anonymization logic of the tiered output module, the system ensures the brand's big data risk control needs while rigorously protecting trade secrets and personal privacy, achieving an organic unity of regulatory compliance, privacy protection, and brand rights protection.

[0086] In summary, the modules of this invention are not simply functionally superimposed, but rather complemented and supported by data links and algorithmic logic. Examples demonstrate that this system provides extremely strong anti-counterfeiting protection, possesses excellent operational stability and multi-dimensional security control capabilities, fully meeting the technical demands of modern global supply chains for intelligent and precise anti-counterfeiting verification.

[0087] Embodiments of the present invention have been presented and described. It will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to the embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A QR code anti-counterfeiting intelligent verification system, characterized in that, include: The coding management module is used to generate a QR code payload containing a digital signature and product serial number and transmit it to the printing equipment to execute printing instructions; The physical fingerprint registration module collects the micro-texture feature vector of the QR code area during the printing instruction execution process, and associates and binds the micro-texture feature vector with the product serial number to generate a preset fingerprint template. The offline verification module is used to acquire real-time micro-textures and perform feature alignment using a preset fingerprint template. At the same time, it calls a local algorithm to perform integrity verification on the digital signature in the QR code payload to obtain a judgment result. The risk assessment module is used to extract the real-time environmental parameters of the terminal from the judgment results and substitute them into the multi-dimensional risk function for calculation to obtain the abnormal probability for the current scanning behavior. The hierarchical output module is used to reference the anomaly probability and perform information desensitization processing on the judgment result according to the preset verification subject role permissions, generating verification result receipts of different dimensions. The cloud scheduling module is used to aggregate product serial numbers, pre-set fingerprint templates, and anomaly probabilities, generate offline material packages containing verification rules through cluster analysis, and distribute the offline material packages to the offline verification module.

2. The QR code anti-counterfeiting intelligent verification system according to claim 1, characterized in that, The anomaly probability calculation performed in the risk assessment module is determined by the comprehensive confidence score formula: ， in, To calculate the overall confidence score, For normalized mapping functions, To verify the weight of digital signatures, The feature distance between the real-time microtexture and the pre-set fingerprint template. For real-time microtexture, For pre-set fingerprint templates, To preset dynamic threshold, For a multidimensional risk function, , , These are the weighting coefficients; The multidimensional risk function The independent variable is defined as: For scanning timestamps, For geographical coordinates, The scanning frequency per unit time. For mobile terminal hardware identification, This is to verify the spatiotemporal distribution pattern of the same product serial number in history.

3. The QR code anti-counterfeiting intelligent verification system according to claim 1, characterized in that, The microtexture feature vector in the physical fingerprint registration module refers to the nonlinear distribution characteristics of QR code pixel edges captured by the image sensor due to printing pressure and medium penetration.

4. The QR code anti-counterfeiting intelligent verification system according to claim 1, characterized in that, In the hierarchical output module, the information desensitization process refers to masking or digesting the original data fields in the judgment result according to the role and permissions of the verification subject. When the role and permissions are consumer, the verification result receipt retains the authenticity judgment conclusion. In the hierarchical output module, the verification subject role permissions include consumer role, regulatory subject role, logistics node role and brand management role, and the data disclosure dimensions of the verification result receipt corresponding to each role are different.

5. The QR code anti-counterfeiting intelligent verification system according to claim 1, characterized in that, In the cloud scheduling module, the offline material package is limited to a public key certificate containing an asymmetric encryption algorithm, a feature digest of the pre-set fingerprint template, and a risk blacklist data set identified by the clustering analysis; In the cloud scheduling module, the clustering analysis refers to extracting the anomaly probability reported by several mobile terminals. When multiple verification requests are detected for the same product serial number within a time period exceeding a preset spatial span, it is determined to be a cluster fraud risk.

6. The QR code anti-counterfeiting intelligent verification system according to claim 1, characterized in that, In the coding management module, the digital signature is a binary data stream generated after signing the product serial number with a private key using an asymmetric encryption algorithm.

7. The QR code anti-counterfeiting intelligent verification system according to claim 1, characterized in that, In the offline verification module, feature alignment refers to extracting topological feature points from the real-time microtexture and calculating the affine transformation parameters of the topological feature points relative to the corresponding feature points in the preset fingerprint template.

8. The QR code anti-counterfeiting intelligent verification system according to claim 1, characterized in that, In the offline verification module, the structural similarity index between the real-time micro-texture after feature alignment and the preset fingerprint template is calculated. To assist in the determination, the logic for determining the legality of the real-time micro-texture is expressed as follows: ， in, For real-time microtexture, For pre-set fingerprint templates, The preset similarity threshold constant is used when the calculation result is... When this happens, the determination result is marked as a physical failure state.

9. The QR code anti-counterfeiting intelligent verification system according to claim 1, characterized in that, The offline verification module has an ambient light compensation function. During the acquisition of the real-time micro-texture, the offline verification module calls the flash hardware of the mobile terminal to perform color temperature compensation lighting, and uses a homomorphic filtering algorithm to perform brightness uniformization processing on the acquired image to eliminate the feature point extraction deviation caused by uneven ambient lighting.

10. A QR code anti-counterfeiting intelligent verification system according to claim 5, characterized in that, In the cloud scheduling module, the offline material package is issued using an asynchronous incremental update mechanism. The cloud scheduling module calculates the feature difference packet between the offline data version currently held by the mobile terminal and the latest version on the cloud. By limiting the upper limit of the number of bytes in the offline material package, the data synchronization consistency of the offline verification module in a weak network environment is ensured.

Images