Commodity traceability management intelligent anti-counterfeiting method and system based on Internet of Things

By performing multi-dimensional feature vectorization processing and binding of product flow chain structure, combined with irreversible hash transformation and verification rules, the problem of poor anti-counterfeiting effect in existing traceability systems has been solved, realizing reliable traceability and anti-counterfeiting verification throughout the entire product lifecycle.

CN121391301APending Publication Date: 2026-01-23BEIJING SOOLE INFORMATION TECH CORP LTD

Patent Information

Application Number
CN202511555247.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-10-29
Publication Date
2026-01-23

AI Technical Summary

Technical Problem

Existing product traceability systems lack multi-dimensional feature extraction and vectorization processing mechanisms, making it difficult to effectively distinguish counterfeit products. Furthermore, the lack of a verifiable binding mechanism between product information and circulation events results in poor anti-counterfeiting effects and an inability to achieve comprehensive verification of the entire circulation chain of products from production to sales.

Method used

By performing multi-dimensional feature vectorization on the physical characteristic data of goods, a two-way binding relationship between the unique feature vector of goods and the circulation link structure is established. Irreversible hash transformation is performed in each circulation event to generate tamper-proof traceability certificates. A set of verification rules is constructed to verify the consistency of the unique feature vector of goods and the integrity of the circulation link structure.

Benefits of technology

It enables reliable anti-counterfeiting and traceability throughout the entire product lifecycle, effectively preventing counterfeiting and tampering, improving the authenticity and reliability of product traceability, and enhancing the security of the anti-counterfeiting system.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121391301A_ABST
    Figure CN121391301A_ABST
Patent Text Reader

Abstract

The invention provides a commodity traceability management intelligent anti-counterfeiting method and system based on the Internet of Things, and relates to the technical field of anti-counterfeiting traceability of the Internet of Things, and the method comprises the steps: carrying out the vectorization processing of commodity physical feature data, obtaining a unique feature vector, constructing a circulation link structure, and building a bidirectional binding relation; and irreversible Hash transformation is carried out in the circulation event, a node record is embedded, a tamper-proof traceability voucher is formed, and a verification rule set is constructed to carry out consistency and integrity verification. According to the invention, the commodity information can be effectively prevented from being tampered, and the reliability and anti-counterfeiting efficiency of the traceability process are improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of IoT anti-counterfeiting and traceability technology, and in particular to an intelligent anti-counterfeiting method and system for commodity traceability management based on IoT. Background Technology

[0002] With the rapid development of global trade, the production, distribution, and sales of goods have become increasingly complex. Problems such as difficulty in distinguishing genuine from counterfeit goods and the proliferation of counterfeit and substandard products in the market are becoming increasingly serious, severely damaging consumer rights and corporate brand reputation. Traditional anti-counterfeiting technologies mainly rely on physical carriers such as anti-counterfeiting labels and QR codes, but these methods are easily copied and counterfeited. The rapid development of IoT technology has brought new solutions to the field of product traceability and anti-counterfeiting. By collecting physical characteristic data of goods and combining it with their circulation information, a more reliable anti-counterfeiting and traceability system can be built. Currently, IoT-based product traceability systems have been applied in areas such as food safety, drug regulation, and luxury goods anti-counterfeiting. Their core lies in effectively linking product information with the circulation process and ensuring the authenticity and integrity of the data.

[0003] Existing technologies lack multi-dimensional feature extraction and vectorization mechanisms for product uniqueness identification, relying mostly on single or limited feature parameters. This makes it difficult to effectively distinguish counterfeit goods, resulting in poor anti-counterfeiting effects. When criminals use advanced counterfeiting techniques to replicate product appearance features, single-dimensional feature identification is easily bypassed. Traditional traceability systems lack a verifiable binding mechanism between product information and circulation events; data storage is often independent, failing to guarantee consistency between the physical product and recorded information. Even with blockchain technology for storing circulation data, the correlation between the product's physical characteristics and on-chain data is weak, making true tamper-proof verification difficult. Existing traceability verification processes lack a systematic verification rule framework; most systems can only verify the authenticity of information at a specific stage, failing to comprehensively and effectively verify the entire circulation chain from production to sales. This allows criminals to bypass system detection by tampering with information at certain stages, reducing the reliability of the entire traceability and anti-counterfeiting system. Summary of the Invention

[0004] This invention provides an intelligent anti-counterfeiting method and system for product traceability management based on the Internet of Things, which can solve the problems in the prior art.

[0005] A first aspect of this invention provides an intelligent anti-counterfeiting method for product traceability management based on the Internet of Things, comprising:

[0006] The physical characteristic data of the goods are processed into multi-dimensional feature vectors to obtain unique feature vectors of the goods. The flow event data of the goods are analyzed by time-series correlation to obtain the flow link structure.

[0007] A bidirectional binding relationship is established between the unique feature vector of the product and each node in the circulation link structure. When each circulation event occurs, the unique feature vector of the product is subjected to an irreversible hash transformation and the transformation result is embedded into the circulation record of the corresponding node to obtain an anti-tampering traceability certificate.

[0008] Construct a verification rule set based on the tamper-proof traceability certificate. The verification rule set includes consistency verification rules for the unique feature vector of the product and integrity verification rules for the circulation link structure.

[0009] According to the consistency verification rules, the physical feature data of the product to be verified is compared with the anti-tampering traceability certificate, and the circulation link structure is traversed and verified according to the verification rule set. When the transformation sequence of the product unique feature vector is completely matched with the embedding result of each node in the circulation link structure and the circulation link structure satisfies the integrity verification rules, the anti-counterfeiting verification is passed.

[0010] The physical characteristic data of the goods are processed into multi-dimensional feature vectors to obtain unique feature vectors for each product. The flow event data of the goods is analyzed using time-series correlation to obtain the flow link structure, which includes:

[0011] Multiple physical dimension parameters are collected from the physical feature data. Feature components are extracted for each physical dimension parameter. The feature components are combined according to a preset dimensional mapping relationship to form an initial feature vector. The initial feature vector is weighted based on spatial distribution differences. A weight coefficient is calculated based on the dispersion of each feature component in the multidimensional feature space. The initial feature vector is then weighted and adjusted using the weight coefficient to obtain the product uniqueness feature vector.

[0012] The timestamp information and node identifier information in the flow event data are parsed. The timestamp information is converted into a time-series index value, and the node identifier information is mapped to a node location code. The sequential relationship of the flow events is determined based on the time-series index value, and the spatial topology relationship of the flow events is determined based on the node location code, thus obtaining the directed dependency relationship between nodes. The transitivity of the directed dependency relationship is constrained and verified. It is checked whether the time-series index value satisfies the monotonically increasing requirement and whether the node location code satisfies the path reachability requirement. The directed dependency relationship that passes the constraint verification is taken as the flow link structure.

[0013] Establishing a bidirectional binding relationship between the product's unique feature vector and each node in the circulation link structure includes:

[0014] Extract the feature components from the product uniqueness feature vector, associate and map the feature components with the node identification information of each node in the flow link structure, and generate a positive index relationship between the feature vector and the node;

[0015] Based on the node identification information of each node in the flow link structure, a reverse query is performed to find the unique feature vector of the product associated with it, and a reverse index relationship between the node and the feature vector is generated.

[0016] The forward index relationship and the reverse index relationship are combined to form a two-way binding relationship, and the two-way binding relationship is stored in the binding relationship record table of the flow link structure.

[0017] At each transaction event, an irreversible hash transformation is performed on the product's unique feature vector, and the transformation result is embedded into the transaction record of the corresponding node to obtain an anti-tampering traceability certificate, including:

[0018] When each circulation event occurs, the node identifier information and circulation timestamp of the node corresponding to the current circulation event are extracted. The node identifier information and circulation timestamp are used as salt input parameters and combined with the product unique feature vector to generate input data to be transformed.

[0019] An irreversible hash transformation operation is performed on the input data to be transformed to obtain a hash transformation result, and a fixed-length hash digest value is extracted from the hash transformation result; the hash digest value, the flow timestamp, the node identification information, and the positive index relationship identifier in the two-way binding relationship are encapsulated into a flow record data block;

[0020] The flow record data block is embedded into the flow record field of the node corresponding to the current flow event in the flow link structure, and the flow record field is sealed to prevent subsequent modification;

[0021] The transfer record fields of all nodes in the transfer link structure are organized in a chain according to the temporal relationship of the transfer timestamp to form the tamper-proof traceability certificate.

[0022] A verification rule set is constructed based on the tamper-proof traceability certificate. The verification rule set includes consistency verification rules for the unique feature vector of the product and integrity verification rules for the circulation link structure, including:

[0023] Based on the method of generating the hash digest value in the tamper-proof traceability certificate, a feature vector comparison benchmark for consistency verification is obtained by reverse derivation. The feature vector comparison benchmark is then established to correspond with the feature components in the product uniqueness feature vector, and a feature vector difference measurement criterion is set based on the correspondence. The feature vector difference measurement criterion and the feature vector comparison benchmark are combined to form the consistency verification rule.

[0024] The directed dependencies of the flow link structure are extracted from the tamper-proof traceability certificate. A predecessor-successor association matrix between nodes is constructed based on the directed dependencies. Link continuity constraints and temporal monotonicity constraints are set based on the predecessor-successor association matrix. The link continuity constraints are used to detect whether there are missing nodes or broken paths in the flow link structure. The temporal monotonicity constraints are used to detect whether the flow timestamp of the current flow event satisfies a strict increasing relationship. The link continuity constraints, the temporal monotonicity constraints, and the predecessor-successor association matrix are combined to form the integrity verification rule.

[0025] The consistency verification rules and the integrity verification rules are encapsulated into the verification rule set.

[0026] According to the consistency verification rules, comparing the physical characteristic data of the product to be verified with the tamper-proof traceability certificate includes:

[0027] The physical characteristic data of the product to be verified is converted into a feature vector to be verified.

[0028] The feature vector comparison benchmark and the feature vector to be verified are dimensionally aligned. The feature component in the feature vector to be verified is compared with the feature component at the corresponding position in the feature vector comparison benchmark. The feature deviation of each dimension is obtained. The number of dimensions that exceed the allowable range of the feature deviation of each dimension is counted according to the feature vector difference measurement criterion. The feature consistency judgment result is determined according to the ratio of the number of dimensions to the total number of dimensions.

[0029] Extract the flow record data block of the last node of the flow link structure from the tamper-proof traceability certificate; concatenate the feature vector to be verified with the node identifier information of the last node of the flow link structure and the flow timestamp, and perform the same irreversible hash transformation operation as in the tamper-proof traceability certificate on the concatenation result to obtain the hash digest value to be verified;

[0030] The hash digest value to be verified is compared bit by bit with the hash digest value in the flow record data block to obtain the hash consistency determination result.

[0031] Perform a logical AND operation between the feature consistency determination result and the hash consistency determination result to obtain the comparison verification result.

[0032] The anti-counterfeiting verification is performed by traversing and verifying the circulation link structure according to the verification rule set. When the transformation sequence of the product unique feature vector completely matches the embedding result of each node in the circulation link structure and the circulation link structure satisfies the integrity verification rule, the anti-counterfeiting verification passes, including:

[0033] The traversal path of the flow link structure is determined according to the predecessor-successor association matrix in the integrity verification rule; each node in the flow link structure is visited sequentially according to the traversal path, and the hash digest value and flow timestamp are extracted to form an embedding result sequence;

[0034] The unique feature vector of the product is concatenated and combined with the node identification information of each node in the traversal path and the flow timestamp in sequence, and then subjected to irreversible hash transformation. A transformation sequence is generated according to the access order of the traversal path.

[0035] Each hash digest value in the transformed sequence is compared one by one with the hash digest values ​​of the corresponding node positions in the embedded result sequence to obtain a sequence matching determination result; according to the link continuity constraint in the integrity verification rule, it is detected whether there are missing nodes or path interruptions in the traversal path to obtain a continuity determination result; according to the temporal monotonicity constraint in the integrity verification rule, it is detected whether the flow timestamps in the embedded result sequence satisfy a strict increasing relationship to obtain a temporal determination result.

[0036] Perform a triple logical AND operation on the sequence matching determination result, the continuity determination result, and the timing determination result. When all three determination results are true, the anti-counterfeiting verification is confirmed to be successful.

[0037] A second aspect of this invention provides an intelligent anti-counterfeiting system for product traceability management based on the Internet of Things, comprising:

[0038] The first unit is used to perform multi-dimensional feature vectorization processing on the physical feature data of the goods to obtain the unique feature vector of the goods, and to perform time-series correlation analysis on the circulation event data of the goods to obtain the circulation link structure.

[0039] The second unit is used to establish a two-way binding relationship between the unique feature vector of the product and each node in the circulation link structure. When each circulation event occurs, the unique feature vector of the product is subjected to an irreversible hash transformation and the transformation result is embedded into the circulation record of the corresponding node to obtain an anti-tampering traceability certificate.

[0040] The third unit is used to construct a set of verification rules based on the tamper-proof traceability certificate. The set of verification rules includes consistency verification rules for the unique feature vector of the product and integrity verification rules for the circulation link structure.

[0041] The fourth unit is used to compare the physical feature data of the product to be verified with the anti-tampering traceability certificate according to the consistency verification rules, and to traverse and verify the circulation link structure according to the verification rule set. When the transformation sequence of the product unique feature vector completely matches the embedding result of each node in the circulation link structure and the circulation link structure satisfies the integrity verification rules, the anti-counterfeiting verification is passed.

[0042] A third aspect of the present invention,

[0043] An electronic device is provided, comprising:

[0044] processor;

[0045] Memory used to store processor-executable instructions;

[0046] The processor is configured to invoke instructions stored in the memory to execute the aforementioned method.

[0047] Fourth aspect of the embodiments of the present invention,

[0048] A computer-readable storage medium is provided, having stored thereon computer program instructions that, when executed by a processor, implement the aforementioned method.

[0049] The beneficial effects of this application are as follows:

[0050] This invention achieves reliable anti-counterfeiting and traceability throughout the entire product lifecycle by bidirectionally binding the vectorization of product physical features with the circulation link structure, effectively preventing counterfeiting and tampering, and improving the authenticity and reliability of product traceability.

[0051] By performing multi-dimensional feature vectorization on the physical characteristic data of goods and applying irreversible hash transformation in each circulation event, a unique tamper-proof traceability certificate is created. This certificate cannot be copied or forged, significantly enhancing the security of the anti-counterfeiting system.

[0052] The verification rule set established by this invention simultaneously verifies the consistency of the product's unique feature vector and the integrity of the circulation link structure, forming a dual verification mechanism that makes anti-counterfeiting verification more comprehensive and rigorous, providing consumers with more reliable guarantees of product authenticity. Attached Figure Description

[0053] Figure 1This is a flowchart illustrating the intelligent anti-counterfeiting method for product traceability management based on the Internet of Things, as described in an embodiment of the present invention.

[0054] Figure 2 A schematic diagram illustrating the process of building a two-way binding relationship. Detailed Implementation

[0055] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0056] The technical solution of the present invention will be described in detail below with reference to specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments.

[0057] Figure 1 This is a flowchart illustrating the intelligent anti-counterfeiting method for product traceability management based on the Internet of Things, as described in an embodiment of the present invention. Figure 1 As shown, the method includes:

[0058] The physical characteristic data of the goods are processed into multi-dimensional feature vectors to obtain unique feature vectors of the goods. The flow event data of the goods are analyzed by time-series correlation to obtain the flow link structure.

[0059] A bidirectional binding relationship is established between the unique feature vector of the product and each node in the circulation link structure. When each circulation event occurs, the unique feature vector of the product is subjected to an irreversible hash transformation and the transformation result is embedded into the circulation record of the corresponding node to obtain an anti-tampering traceability certificate.

[0060] Construct a verification rule set based on the tamper-proof traceability certificate. The verification rule set includes consistency verification rules for the unique feature vector of the product and integrity verification rules for the circulation link structure.

[0061] According to the consistency verification rules, the physical feature data of the product to be verified is compared with the anti-tampering traceability certificate, and the circulation link structure is traversed and verified according to the verification rule set. When the transformation sequence of the product unique feature vector is completely matched with the embedding result of each node in the circulation link structure and the circulation link structure satisfies the integrity verification rules, the anti-counterfeiting verification is passed.

[0062] In one optional implementation, the physical characteristic data of the goods are processed into multi-dimensional feature vectors to obtain unique feature vectors for the goods. The flow event data of the goods are then analyzed using time-series correlation to obtain the flow link structure, which includes:

[0063] Multiple physical dimension parameters are collected from the physical feature data. Feature components are extracted for each physical dimension parameter. The feature components are combined according to a preset dimensional mapping relationship to form an initial feature vector. The initial feature vector is weighted based on spatial distribution differences. A weight coefficient is calculated based on the dispersion of each feature component in the multidimensional feature space. The initial feature vector is then weighted and adjusted using the weight coefficient to obtain the product uniqueness feature vector.

[0064] The timestamp information and node identifier information in the flow event data are parsed. The timestamp information is converted into a time-series index value, and the node identifier information is mapped to a node location code. The sequential relationship of the flow events is determined based on the time-series index value, and the spatial topology relationship of the flow events is determined based on the node location code, thus obtaining the directed dependency relationship between nodes. The transitivity of the directed dependency relationship is constrained and verified. It is checked whether the time-series index value satisfies the monotonically increasing requirement and whether the node location code satisfies the path reachability requirement. The directed dependency relationship that passes the constraint verification is taken as the flow link structure.

[0065] Physical feature data contains multiple dimensional parameters, such as texture features, shape features, and material features. For texture features, parameters such as energy, entropy, and contrast are extracted using the gray-level co-occurrence matrix. For shape features, parameters such as contour perimeter, area ratio, and roundness are extracted. For material features, parameters such as hardness, elastic coefficient, and gloss are extracted. Feature components are extracted for each physical dimension parameter. For example, the energy value of 0.83, entropy value of 4.27, and contrast value of 0.65 of the texture are extracted as feature components.

[0066] The extracted feature components are combined according to a preset dimensional mapping relationship. Texture features are mapped to the first 8 dimensions of the vector, shape features are mapped to the middle 12 dimensions of the vector, and material features are mapped to the last 10 dimensions of the vector, thus forming a 30-dimensional initial feature vector. Specifically, the initial feature vector of a certain product can be represented as [0.83, 4.27, 0.65, 0.92, 1.47, 3.28, 0.55, 0.76, 0.95, 0.88, 0.72, 0.63, 0.57, 0.91, 0.84, 0.79, 0.68, 0.75, 0.81, 0.87, 0.94, 0.78, 0.67, 0.85, 0.73, 0.89, 0.71, 0.82, 0.93, 0.69].

[0067] To improve the discriminative power of feature vectors, the initial feature vectors are weighted based on spatial distribution differences. The dispersion of each feature component in the entire product database within the multidimensional feature space is analyzed, and weight coefficients are calculated. Feature components with high dispersion are assigned higher weights, while those with low dispersion are assigned lower weights. For example, after analyzing 1000 products, the dispersion of texture entropy is found to be 0.87, shape roundness to be 0.42, and material gloss to be 0.76. Based on these findings, a set of weight coefficients is generated. Multiply each component of the initial feature vector by its corresponding weight coefficient to obtain the weighted adjusted product uniqueness feature vector [1.41, 3.72, 0.52, 0.65, 1.25, 2.95, 0.47, 0.68, 0.40, 0.35, 0.58, 0.33, 0.24, 0.43, 0.37, 0.31, 0.35, 0.38, 0.62, 0.70, 0.82, 0.59, 0.51, 0.68, 0.55, 0.75, 0.54, 0.63, 0.74, 0.52].

[0068] During the parsing of the distribution chain structure, the distribution event data of the goods is analyzed, and the timestamp information and node identification information are extracted. For example, the distribution record of a certain product at different nodes includes: {"Node":"Manufacturer A", "Timestamp":"2023-05-10 08:30:15"}, {"Node":"Logistics Center B", "Timestamp":"2023-05-12 14:20:42"}, {"Node":"Regional Warehouse C", "Timestamp":"2023-05-15 09:45:37"}, {"Node":"Retail Store D", "Timestamp":"2023-05-17 16:18:05"}.

[0069] The timestamp information is converted into a time-series index value. Using the start time of the flow as the baseline (0), the time interval (in hours) of each subsequent flow event relative to the baseline is calculated, resulting in a time-series index value sequence [0, 53.84, 121.25, 175.80]. Simultaneously, node identifier information is mapped to node location codes, using a hash mapping method to convert the node identifier into a unique numerical code. For example, "Manufacturer A" is mapped to code "10042", "Logistics Center B" to code "20156", "Regional Warehouse C" to code "30089", and "Retail Store D" to code "40127".

[0070] By comparing the temporal index values, the order in which events occur is determined; by analyzing the node location codes, the spatial relationships between nodes are determined. This establishes directed dependencies between nodes, such as "Manufacturer A" → "Logistics Center B" → "Regional Warehouse C" → "Retail Store D".

[0071] The established directed dependencies are constrained and validated to ensure their rationality. This validation includes two aspects: first, verifying whether the temporal index values ​​meet the monotonically increasing requirement, ensuring logical consistency in time sequence; and second, verifying whether the node location codes meet the path reachability requirement, ensuring reasonable spatial flow paths. For the example above, the temporal index values ​​[0, 53.84, 121.25, 175.80] are indeed verified to show a monotonically increasing trend; simultaneously, the node code sequence [10042, 20156, 30089, 40127] is verified to have a valid path connection within the preset flow network topology. After validation, the directed dependency is confirmed to meet the constraints and is established as the final flow link structure.

[0072] Through the above processing steps, a unique feature vector and circulation link structure for the product were successfully constructed, providing technical support for product traceability and anti-counterfeiting.

[0073] In one optional implementation, establishing a bidirectional binding relationship between the product unique feature vector and each node in the circulation link structure includes:

[0074] Extract the feature components from the product uniqueness feature vector, associate and map the feature components with the node identification information of each node in the flow link structure, and generate a positive index relationship between the feature vector and the node;

[0075] Based on the node identification information of each node in the flow link structure, a reverse query is performed to find the unique feature vector of the product associated with it, and a reverse index relationship between the node and the feature vector is generated.

[0076] The forward index relationship and the reverse index relationship are combined to form a two-way binding relationship, and the two-way binding relationship is stored in the binding relationship record table of the flow link structure.

[0077] like Figure 2 The method includes:

[0078] A product's unique feature vector typically contains multi-dimensional feature information. For example, for a piece of clothing, its feature vector consists of 32 dimensions, including color feature components (such as numerical representations converted from RGB values), material feature components (such as encoded values ​​for materials like cotton, linen, and silk), size feature components (such as numerical mappings of S, M, and L), and production batch feature components (such as hash values ​​of batch numbers). Feature extraction algorithms scan or detect the product to generate numerical representations of these feature components. Taking a T-shirt as an example, its feature vector is represented as [254, 128, 65, 87, 32, 145, ..., 78], where the values ​​at different positions correspond to different feature attributes.

[0079] After extracting the feature components, these feature components are associated with the node identification information of each node in the flow link structure to generate a positive index relationship between the feature vector and the node. The nodes in the flow link structure include manufacturer nodes, logistics center nodes, retailer nodes, and consumer nodes, etc. Each node has unique node identification information, such as node ID, geographical location information, business role type, etc. By establishing a mapping table, the feature components in the product feature vector are associated with these node identification information. For example, for the T-shirt product mentioned above, its color feature component [254, 128, 65] is associated with the manufacturer node ID "PRO10086", the material feature component [87, 32] is associated with the logistics center node ID "LOG20521", and the size feature component

[145] is associated with the retailer node ID "RET30978". This association forms a positive index relationship from feature vectors to nodes, and the record can be represented as: {feature component: [254, 128, 65], node ID: "PRO10086"}, {feature component: [87, 32], node ID: "LOG20521"}, and so on.

[0080] Based on the node identification information of each node in the flow chain structure, a reverse index relationship is generated between the node and the associated unique product feature vector by reverse querying. The process of establishing the reverse index relationship is the reverse operation of the forward index. By using the node identification information, all product feature vectors related to the node are found. For example, when querying the logistics center node with node ID "LOG20521", it will find which product feature vectors the node is associated with. In the above example, the node is found to be associated with the material feature component [87, 32] in the T-shirt product feature vector. The record form of the reverse index relationship can be represented as: {Node ID: "LOG20521", Product feature vector components: [[87, 32], [76, 45, 23], ...]}, where [76, 45, 23] is the feature component of another product.

[0081] Combining forward and reverse index relationships forms a bidirectional binding relationship. This bidirectional binding relationship is essentially an integration of index structures in two directions, enabling both rapid location of relevant circulation link nodes from product feature vectors and rapid retrieval of all related products from circulation link nodes. The bidirectional binding relationship can be implemented using hash tables or relational database tables, where key-value pairs store index information in both directions respectively. Represented in database table form, two tables can be designed: a feature vector index table and a node index table. The feature vector index table contains fields: feature vector ID, feature component value, and associated node ID; the node index table contains fields: node ID, node type, and associated feature vector ID.

[0082] This two-way binding relationship is stored in the binding relationship record table of the flow link structure. The binding relationship record table is a persistent data storage structure that can be implemented using a distributed database system. This table stores all binding relationship information between feature vectors and flow link nodes, including fields such as binding relationship ID, feature vector ID, node ID, binding timestamp, and binding status. Taking a T-shirt as an example, its binding relationship record is: {Binding relationship ID: "BD100001", Feature vector ID: "FV200001", Node ID: "PRO10086", Binding timestamp: "2023-05-20 10:30:45", Binding status: "Valid"}.

[0083] By establishing this two-way binding relationship, it is possible to track and manage goods throughout the entire circulation process. When it is necessary to query the circulation path of a certain product, starting from the product's unique feature vector, all related nodes can be found through the forward index relationship. When it is necessary to query all products processed by a certain node, starting from the node identification information, all related product feature vectors can be found through the reverse index relationship. This two-way indexing mechanism greatly improves the system's query efficiency and data management capabilities, and provides technical support for the transparency and traceability of the product circulation chain.

[0084] In one optional implementation, an irreversible hash transformation is performed on the product's unique feature vector each time a transfer event occurs, and the transformation result is embedded into the transfer record of the corresponding node to obtain an tamper-proof traceability certificate, including:

[0085] When each circulation event occurs, the node identifier information and circulation timestamp of the node corresponding to the current circulation event are extracted. The node identifier information and circulation timestamp are used as salt input parameters and combined with the product unique feature vector to generate input data to be transformed.

[0086] An irreversible hash transformation operation is performed on the input data to be transformed to obtain a hash transformation result, and a fixed-length hash digest value is extracted from the hash transformation result; the hash digest value, the flow timestamp, the node identification information, and the positive index relationship identifier in the two-way binding relationship are encapsulated into a flow record data block;

[0087] The flow record data block is embedded into the flow record field of the node corresponding to the current flow event in the flow link structure, and the flow record field is sealed to prevent subsequent modification;

[0088] The transfer record fields of all nodes in the transfer link structure are organized in a chain according to the temporal relationship of the transfer timestamp to form the tamper-proof traceability certificate.

[0089] During the circulation of goods, the node identifier information and circulation timestamp of the node corresponding to the current circulation event are extracted. The node identifier information can be a unique identifier of the node, such as "SUPPLIER_001", and the circulation timestamp adopts a standard time format, such as "2023-04-15T14:30:25.123Z". These two pieces of information are used as salt input parameters and concatenated with the product unique feature vector. The product unique feature vector is a set of feature data used to uniquely identify the product, such as "FV_9876543210ABCDEF". The node identifier information, circulation timestamp, and product unique feature vector are concatenated in a specific order to form the input data to be transformed, such as "SUPPLIER_001|2023-04-15T14:30:25.123Z|FV_9876543210ABCDEF".

[0090] An irreversible hash transformation operation is performed on the input data to be transformed, and the hash transformation result is calculated using a hash algorithm such as SHA-256. For example, after performing a hash transformation on the above input data, a 64-bit hash value "8a7b3c9d4e2f1a0b8c7d6e5f4a3b2c1d8e7f6a5b4c3d2e1f0a9b8c7d6e5f4a3b" is obtained. From this hash transformation result, a fixed-length hash digest value, such as the first 32 bits "8a7b3c9d4e2f1a0b8c7d6e5f4a3b2c1d", is extracted to reduce the data storage burden while maintaining tamper-proof characteristics.

[0091] The hash digest value, along with the circulation timestamp, node identifier information, and the forward index identifier in the two-way binding relationship, are encapsulated into a circulation record data block. The forward index identifier is an index pointing to the physical entity of the product, such as "PROD_INDEX_123456". The encapsulated circulation record data block structure can be represented as: {"hashDigest":"8a7b3c9d4e2f1a0b8c7d6e5f4a3b2c1d", "timestamp":"2023-04-15T14:30:25.123Z", "nodeId":"SUPPLIER_001", "forwardIndex":"PROD_INDEX_123456"}.

[0092] The circulation record data block is embedded into the circulation record field of the node corresponding to the current circulation event in the circulation link structure. The circulation link structure consists of multiple nodes, each representing a link in the product circulation process. To prevent subsequent modification, the circulation record field is sealed, and digital signature technology is used to sign the circulation record data. For example, the hash value of the circulation record data block is signed using the system private key to generate a sealed signature "SEAL_SIGNATURE_98765432", which is then added to the circulation record.

[0093] An example of the sealed transfer record fields is: {"recordData":{"hashDigest":"8a7b3c9d4e2f1a0b8c7d6e5f4a3b2c1d", "timestamp":"2023-04-15T14:30:25.123Z", "nodeId":"SUPPLIER_001", "forwardIndex":"PROD_INDEX_123456"}, "sealSignature":"SEAL_SIGNATURE_98765432"}.

[0094] The transfer record fields of all nodes in the transfer chain structure are organized in a chain according to the temporal relationship of the transfer timestamps, forming a tamper-proof traceability certificate. Specifically, each node's transfer record not only contains its own information but also a hash reference to the transfer record of the previous node. For example, the transfer record of the second node "DISTRIBUTOR_002" will contain a hash reference to the transfer record of the first node "SUPPLIER_001". This chain structure ensures that if the record of any node is tampered with, the reference of subsequent nodes will become invalid, thus enabling the detection of tampering.

[0095] A complete example of an tamper-proof traceability credential includes the flow record of multiple nodes: [{"recordData":{"hashDigest":"8a7b3c9d4e2f1a0b8c7d6e5f4a3b2c1d", "timestamp":"2023-04-15T14:30:25.123Z", "nodeId":"SUPPLIER_001", "forwardIndex":"PROD_INDEX_123456", "prevHash":null}, "sealSignature":"SEAL_SIGNATURE_98765432"}, {"recordData":{"hashDigest":"2e3f4a5b6c7d8e9f0a1b2c3d4e5f6a7b", "timestamp":"2023-04-17T09:45:18.456Z", "nodeId":"DISTRIBUTOR_002", "forwardIndex":"PROD_INDEX_123456", "prevHash":"8a7b3c9d4e2f1a0b8c7d6e5f4a3b2c1d"}, "sealSignature":"SEAL_SIGNATURE_87654321"}, {"recordData":{"hashDigest":"3d4e5f6a7b8c9d0e1f2a3b4c5d6e7f8a", "timestamp":"2023-04-19T16:20:37.789Z", "nodeId":"RETAILER_003", "forwardIndex":"PROD_INDEX_123456", "prevHash":"2e3f4a5b6c7d8e9f0a1b2c3d4e5f6a7b"}, "sealSignature":"SEAL_SIGNATURE_76543210"}].

[0096] In this way, immutable records are generated at every stage of the product circulation process. These records form a complete traceability certificate through hash links. Any attempt to tamper with the circulation records will cause the hash chain to break, and tampering can be detected by verifying the integrity of the hash chain. Furthermore, due to the use of irreversible hash transformation and digital signature technology, even if a hacker obtains partial data, they cannot forge valid circulation records or modify existing records. This mechanism ensures the authenticity and integrity of product traceability information, providing consumers with reliable information about the origin of goods.

[0097] In one optional implementation, a verification rule set based on the tamper-proof traceability certificate is constructed. The verification rule set includes consistency verification rules for the product's unique feature vector and integrity verification rules for the circulation link structure, including:

[0098] Based on the method of generating the hash digest value in the tamper-proof traceability certificate, a feature vector comparison benchmark for consistency verification is obtained by reverse derivation. The feature vector comparison benchmark is then established to correspond with the feature components in the product uniqueness feature vector, and a feature vector difference measurement criterion is set based on the correspondence. The feature vector difference measurement criterion and the feature vector comparison benchmark are combined to form the consistency verification rule.

[0099] The directed dependencies of the flow link structure are extracted from the tamper-proof traceability certificate. A predecessor-successor association matrix between nodes is constructed based on the directed dependencies. Link continuity constraints and temporal monotonicity constraints are set based on the predecessor-successor association matrix. The link continuity constraints are used to detect whether there are missing nodes or broken paths in the flow link structure. The temporal monotonicity constraints are used to detect whether the flow timestamp of the current flow event satisfies a strict increasing relationship. The link continuity constraints, the temporal monotonicity constraints, and the predecessor-successor association matrix are combined to form the integrity verification rule.

[0100] The consistency verification rules and the integrity verification rules are encapsulated into the verification rule set.

[0101] During the construction of consistency verification rules, the hash digest value in the tamper-proof traceability certificate is derived in reverse based on its generation method. For example, for a drug's traceability certificate, its hash digest value is calculated using the SHA-256 algorithm to analyze the drug's feature vector [color value, weight, size, special marking pattern features]. By extracting the hash value "7f83b1657ff1fc53b92dc18148a1d65dfc2d4b1fa3d677284addd200126d9069" and combining it with known hash algorithm parameters, the feature vector benchmark for comparison [RGB(255, 240, 240), 45.7g, 60mm×30mm×15mm, marking pattern feature code 0x3F5A2C] is derived. This feature vector benchmark is then correlated with the actual detected product feature vectors. For example, the color value correspondence is [RGB benchmark value → measured RGB value], and the weight correspondence is [benchmark weight → measured weight], etc.

[0102] Based on the above correspondence, feature vector difference measurement criteria are set. For example, color difference is measured by an RGB Euclidean distance not exceeding 10, weight difference by ±0.5g, and size difference by ±1mm. Special marking pattern features are measured by a similarity algorithm score not lower than 0.95. The feature vector comparison benchmark and difference measurement criteria are combined to form a consistency verification rule. That is, during verification, the judgment condition is: "If the Euclidean distance between the measured RGB value and the benchmark RGB value is less than 10, the weight difference is within ±0.5g, the size difference in each dimension is within ±1mm, and the similarity of the marking pattern features is not lower than 0.95, then the consistency verification is passed."

[0103] In the process of constructing integrity verification rules, directed dependencies in the circulation chain structure are extracted from the tamper-proof traceability certificate. For example, the circulation chain of a drug contains a sequence of nodes such as [manufacturer → wholesaler → retailer → consumer]. Based on the dependencies between these nodes, a predecessor-successor association matrix is ​​constructed. Assuming there are four nodes labeled A, B, C, and D, where A represents the manufacturer, B represents the wholesaler, C represents the retailer, and D represents the consumer, the association matrix can be represented as a 4×4 two-dimensional table, where the value of position (i, j) is 1 indicating that node i is the direct predecessor of node j, otherwise it is 0. For the above circulation chain, in the association matrix, (A, B) = 1 indicates that A is the predecessor of B, (B, C) = 1 indicates that B is the predecessor of C, (C, D) = 1 indicates that C is the predecessor of D, and all other positions are 0.

[0104] Based on the predecessor-successor correlation matrix, the link continuity constraint is set as "for any non-starting node, there must be at least one valid predecessor node". In practical applications, each node is checked to see if it has a corresponding predecessor node record. If a node (such as C) is found to have no predecessor node record (i.e., there is no record from B to C), it is determined that the link is broken. The temporal monotonicity constraint is set as "the timestamp of the successor node must be strictly greater than the timestamp of its predecessor node". For example, if the timestamp of wholesaler B is "2023-06-10T15:30:00", then the timestamp of retailer C must be later than this time, such as "2023-06-12T09:15:00".

[0105] Link continuity constraints are used to detect missing nodes or broken paths in the flow link structure. For example, in the expected [A→B→C→D] link, if the record of node B is missing, there is a break between A and C. Temporal monotonicity constraints can be used to detect timestamp anomalies. For example, if the timestamp "2023-06-08T10:00:00" of node C is found to be earlier than the timestamp "2023-06-10T15:30:00" of its predecessor node B, then a temporal anomaly is identified.

[0106] The link continuity constraint, the temporal monotonicity constraint, and the predecessor-successor correlation matrix are combined to form an integrity verification rule. This rule requires that each node in the flow link (except for the starting node) must have a corresponding predecessor node, and the timestamp of each node must be strictly greater than the timestamps of all its predecessor nodes.

[0107] By encapsulating consistency verification rules and integrity verification rules into a unified set of verification rules, the consistency verification is first performed to check whether the product characteristics match, and then the integrity verification is performed to check whether the circulation link is compliant. Only when both verifications pass is the product's traceability information deemed authentic and valid. This multi-layered verification mechanism significantly improves the security and reliability of the anti-counterfeiting traceability system.

[0108] In one optional implementation, comparing the physical characteristic data of the product to be verified with the tamper-proof traceability certificate according to the consistency verification rules includes:

[0109] The physical characteristic data of the product to be verified is converted into a feature vector to be verified.

[0110] The feature vector comparison benchmark and the feature vector to be verified are dimensionally aligned. The feature component in the feature vector to be verified is compared with the feature component at the corresponding position in the feature vector comparison benchmark. The feature deviation of each dimension is obtained. The number of dimensions that exceed the allowable range of the feature deviation of each dimension is counted according to the feature vector difference measurement criterion. The feature consistency judgment result is determined according to the ratio of the number of dimensions to the total number of dimensions.

[0111] Extract the flow record data block of the last node of the flow link structure from the tamper-proof traceability certificate; concatenate the feature vector to be verified with the node identifier information of the last node of the flow link structure and the flow timestamp, and perform the same irreversible hash transformation operation as in the tamper-proof traceability certificate on the concatenation result to obtain the hash digest value to be verified;

[0112] The hash digest value to be verified is compared bit by bit with the hash digest value in the flow record data block to obtain the hash consistency determination result.

[0113] Perform a logical AND operation between the feature consistency determination result and the hash consistency determination result to obtain the comparison verification result.

[0114] According to the aforementioned consistency verification rules, the physical characteristic data of the product to be verified needs to be compared with the tamper-proof traceability certificate. The physical characteristics of the product can include quantifiable attributes such as weight, size, color distribution, and surface texture. For the product to be verified, its physical characteristic data is first collected using sensors. For example, for a perfume product, data such as its weight of 50.2 grams, bottle height of 12.5 cm, bottle diameter of 3.2 cm, and odor characteristic values ​​of [0.35, 0.42, 0.23] can be collected.

[0115] The collected physical feature data is transformed into a feature vector to be verified using a feature extraction algorithm. This feature vector is a multi-dimensional numerical array, where each dimension represents a feature attribute of the product. For example, the physical features of perfume are transformed into a 128-dimensional feature vector, where the first 10 components represent [50.2, 12.5, 3.2, 0.35, 0.42, 0.23, 0.88, 0.75, 0.62, 0.91].

[0116] The feature vectors are aligned dimensionally with the benchmark and the feature vector to be verified to ensure they have the same dimensional structure. If the original feature vectors have different lengths, they can be padded with zeros or truncated to make them the same length. In this example, the first 10 components of the benchmark feature vector are assumed to be [50.0, 12.4, 3.2, 0.34, 0.43, 0.23, 0.89, 0.76, 0.61, 0.92]. After alignment, the difference between each feature component in the feature vector to be verified and the corresponding feature component in the benchmark feature vector is calculated dimension by dimension. The difference can be calculated using the absolute difference method to obtain the feature deviation in each dimension. In this example, the deviation of the first 10 components is [0.2, 0.1, 0.0, 0.01, 0.01, 0.00, 0.01, 0.01, 0.01, 0.01].

[0117] Based on the feature vector difference measurement criteria, an allowable deviation range is set for each dimension. For example, the weight dimension allows a deviation of ±0.5 grams, the size dimension allows a deviation of ±0.2 centimeters, and the odor feature value allows a deviation of ±0.02. The number of dimensions whose feature deviation exceeds the allowable range is counted. In this example, assuming that 3 dimensions out of the 128-dimensional feature vector exceed the allowable deviation range, and the total number of dimensions is 128, then the proportion of dimensions exceeding the allowable deviation range is 3 / 128 = 0.0234. If it is stipulated that this proportion does not exceed 0.05, then the feature consistency determination result for this case is "consistent".

[0118] The data block of the last node in the circulation chain structure is extracted from the tamper-proof traceability certificate. The traceability certificate is constructed using blockchain technology. Each circulation node generates a data block containing node information, a timestamp, and a feature data hash value. The data block of the last node is obtained, which contains the node identifier "NODE_123", the circulation timestamp "2023-04-10T15:30:45Z", and the hash digest value "7f83b1657ff1fc53b92dc18148a1d65dfc2d4b1fa3d677284addd200126d9069".

[0119] The feature vector to be verified is concatenated with the node identifier information and flow timestamp of the last node in the flow link structure. The concatenated data is the feature vector to be verified (e.g., [50.2, 12.5, 3.2, ...]) + "NODE_123" + "2023-04-10T15:30:45Z". The concatenated result is subjected to the same irreversible hash transformation operation (e.g., SHA-256) as in the tamper-proof traceability certificate, to obtain the hash digest value to be verified, "7f83b1657ff1fc53b92dc18148a1d65dfc2d4b1fa3d677284addd200126d9069".

[0120] The hash digest value to be verified is compared bit by bit with the hash digest value in the flow record data block. Each bit in the two 64-character hash strings is compared. In this example, the two hash digest values ​​are completely identical, so the hash consistency determination result is "consistent".

[0121] The feature consistency judgment result and the hash consistency judgment result are logically ANDed to obtain the final comparison and verification result. Since the feature consistency judgment result is "consistent", the hash consistency judgment result is also "consistent", and the logical AND operation result is "consistent", the product to be verified is determined to be genuine.

[0122] If any judgment result is "inconsistent," the final verification result will also be "inconsistent." For example, if the proportion of dimensions with feature deviation exceeding the allowable range is 0.06, exceeding the set threshold of 0.05, then the feature consistency judgment result is "inconsistent." Or, if the hash digest value to be verified does not match the hash digest value in the record, then the hash consistency judgment result is "inconsistent." In both cases, the final comparison verification result will be "inconsistent," and the product to be verified will be determined to be counterfeit.

[0123] In one optional implementation, the circulation link structure is traversed and verified according to the verification rule set. When the transformation sequence of the product uniqueness feature vector completely matches the embedding result of each node in the circulation link structure and the circulation link structure satisfies the integrity verification rule, the anti-counterfeiting verification passes, including:

[0124] The traversal path of the flow link structure is determined according to the predecessor-successor association matrix in the integrity verification rule; each node in the flow link structure is visited sequentially according to the traversal path, and the hash digest value and flow timestamp are extracted to form an embedding result sequence;

[0125] The unique feature vector of the product is concatenated and combined with the node identification information of each node in the traversal path and the flow timestamp in sequence, and then subjected to irreversible hash transformation. A transformation sequence is generated according to the access order of the traversal path.

[0126] Each hash digest value in the transformed sequence is compared one by one with the hash digest values ​​of the corresponding node positions in the embedded result sequence to obtain a sequence matching determination result; according to the link continuity constraint in the integrity verification rule, it is detected whether there are missing nodes or path interruptions in the traversal path to obtain a continuity determination result; according to the temporal monotonicity constraint in the integrity verification rule, it is detected whether the flow timestamps in the embedded result sequence satisfy a strict increasing relationship to obtain a temporal determination result.

[0127] Perform a triple logical AND operation on the sequence matching determination result, the continuity determination result, and the timing determination result. When all three determination results are true, the anti-counterfeiting verification is confirmed to be successful.

[0128] The anti-counterfeiting verification system ensures the authenticity and integrity of goods by traversing and verifying the structure of the product circulation chain. Based on the predecessor-successor association matrix in the integrity verification rules, the system determines the traversal path of the circulation chain structure. This association matrix is ​​stored as a two-dimensional array, where the matrix element M[i][j]=1 indicates that node i is the predecessor of node j, otherwise M[i][j]=0. For example, for a circulation chain containing 5 nodes, its association matrix is ​​a 5×5 matrix, where M[0][1]=1 indicates that node 0 is the predecessor of node 1. Starting from the source node, the system gradually determines the next access node based on the connection relationships in the association matrix, forming a complete traversal path array, such as [0, 1, 3, 4, 2].

[0129] Following the defined traversal path, each node in the flow link structure is visited sequentially. The hash digest value and flow timestamp stored on each node are extracted. For example, for the path [0, 1, 3, 4, 2], node 0 is visited first, and its hash digest value "a1b2c3d4e5f6" and timestamp "1628156800" are extracted; then node 1 is visited, and its hash digest value "g7h8i9j0k1l2" and timestamp "1628243200" are extracted; and so on. These extracted values ​​are arranged in the order of access to form the embedding result sequence, such as [("a1b2c3d4e5f6", "1628156800"), ("g7h8i9j0k1l2", "1628243200"), ...].

[0130] Then, the product's unique feature vector is concatenated with the identifiers and timestamps of each node in the traversal path and subjected to an irreversible hash transformation. The product's unique feature vector is a fixed-length string, such as "4a5b6c7d8e9f". For each node in the path, the feature vector, node identifier, and timestamp are concatenated together, such as "4a5b6c7d8e9f-0-1628156800". Then, an irreversible transformation is performed on the concatenation result using a hash algorithm such as SHA-256 to obtain a hash digest value. These hash digest values ​​are generated sequentially according to the access order of the traversal path, forming a transformation sequence, such as ["a1b2c3d4e5f6", "g7h8i9j0k1l2", ...].

[0131] Each hash digest value in the transformed sequence is compared one by one with the hash digest values ​​at the corresponding node positions in the embedded result sequence. For example, the first element "a1b2c3d4e5f6" in the transformed sequence is compared with the hash value "a1b2c3d4e5f6" in the first element of the embedded result sequence, and so on. If all comparisons match, the sequence matching result is true; otherwise, it is false.

[0132] Based on the link continuity constraints in the integrity verification rules, the system checks whether there are missing nodes or path interruptions in the traversal path, and verifies whether the connection relationships in the association matrix are completely preserved, ensuring that adjacent nodes in the traversal path are indeed directly connected in the original flow link. For example, for the path [0, 1, 3, 4, 2], it verifies whether M[0][1]=1, M[1][3]=1, M[3][4]=1, and M[4][2]=1 are all true. If all connections pass the verification, the continuity judgment result is true; otherwise, it is false.

[0133] Based on the temporal monotonicity constraint in the integrity verification rules, the system checks whether the timestamps in the embedded result sequence satisfy a strictly increasing relationship. It compares the timestamps of adjacent nodes to ensure that the timestamp of the next node is greater than the timestamp of the previous node. For example, it checks whether the timestamp "1628243200" of node 1 is greater than the timestamp "1628156800" of node 0. If all timestamps are strictly increasing, the temporal determination result is true; otherwise, it is false.

[0134] The anti-counterfeiting verification is performed using a triple logical AND operation on the sequence matching result, continuity verification result, and time order verification result. The verification passes only when all three results are true. In the implementation, boolean variables can be used to store each result, such as `isSequenceMatched=true` indicating successful sequence matching, `isContinuityValid=true` indicating successful continuity verification, and `isTimeOrderValid=true` indicating successful time order verification. The triple logical AND operation can be represented as `finalResult = isSequenceMatched && isContinuityValid && isTimeOrderValid`. If `finalResult` is true, the verification is returned as successful; if false, the verification fails, and corresponding error messages can be provided based on the specific failed condition, such as "sequence mismatch," "link discontinuity," or "timestamp order error."

[0135] This multi-layered verification mechanism effectively prevents common counterfeiting methods such as forging nodes, tampering with circulation records, or disrupting time sequences, ensuring the authenticity of goods and the integrity of the circulation chain, and providing users with reliable anti-counterfeiting verification results.

[0136] This invention relates to an intelligent anti-counterfeiting system for product traceability management based on the Internet of Things, the system comprising:

[0137] The first unit is used to perform multi-dimensional feature vectorization processing on the physical feature data of the goods to obtain the unique feature vector of the goods, and to perform time-series correlation analysis on the circulation event data of the goods to obtain the circulation link structure.

[0138] The second unit is used to establish a two-way binding relationship between the unique feature vector of the product and each node in the circulation link structure. When each circulation event occurs, the unique feature vector of the product is subjected to an irreversible hash transformation and the transformation result is embedded into the circulation record of the corresponding node to obtain an anti-tampering traceability certificate.

[0139] The third unit is used to construct a set of verification rules based on the tamper-proof traceability certificate. The set of verification rules includes consistency verification rules for the unique feature vector of the product and integrity verification rules for the circulation link structure.

[0140] The fourth unit is used to compare the physical feature data of the product to be verified with the anti-tampering traceability certificate according to the consistency verification rules, and to traverse and verify the circulation link structure according to the verification rule set. When the transformation sequence of the product unique feature vector completely matches the embedding result of each node in the circulation link structure and the circulation link structure satisfies the integrity verification rules, the anti-counterfeiting verification is passed.

[0141] A third aspect of the present invention provides an electronic device, comprising:

[0142] processor;

[0143] Memory used to store processor-executable instructions;

[0144] The processor is configured to invoke instructions stored in the memory to execute the aforementioned method.

[0145] A fourth aspect of the present invention provides a computer-readable storage medium having stored thereon computer program instructions that, when executed by a processor, implement the aforementioned method.

[0146] This invention can be a method, apparatus, system, and / or computer program product. The computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for performing various aspects of the invention.

[0147] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. An intelligent anti-counterfeiting method for product traceability management based on the Internet of Things, characterized in that, include: The physical characteristic data of the goods are processed into multi-dimensional feature vectors to obtain unique feature vectors of the goods. The flow event data of the goods are analyzed by time-series correlation to obtain the flow link structure. A bidirectional binding relationship is established between the unique feature vector of the product and each node in the circulation link structure. When each circulation event occurs, the unique feature vector of the product is subjected to an irreversible hash transformation and the transformation result is embedded into the circulation record of the corresponding node to obtain an anti-tampering traceability certificate. Construct a verification rule set based on the tamper-proof traceability certificate. The verification rule set includes consistency verification rules for the unique feature vector of the product and integrity verification rules for the circulation link structure. According to the consistency verification rules, the physical feature data of the product to be verified is compared with the anti-tampering traceability certificate, and the circulation link structure is traversed and verified according to the verification rule set. When the transformation sequence of the product unique feature vector is completely matched with the embedding result of each node in the circulation link structure and the circulation link structure satisfies the integrity verification rules, the anti-counterfeiting verification is passed.

2. The method according to claim 1, characterized in that, The physical characteristic data of the goods are processed into multi-dimensional feature vectors to obtain unique feature vectors for each product. The flow event data of the goods is analyzed using time-series correlation to obtain the flow link structure, which includes: Multiple physical dimension parameters are collected from the physical feature data. Feature components are extracted for each physical dimension parameter. The feature components are combined according to a preset dimensional mapping relationship to form an initial feature vector. The initial feature vector is weighted based on spatial distribution differences. A weight coefficient is calculated based on the dispersion of each feature component in the multidimensional feature space. The initial feature vector is then weighted and adjusted using the weight coefficient to obtain the product uniqueness feature vector. The timestamp information and node identifier information in the flow event data are parsed. The timestamp information is converted into a time-series index value, and the node identifier information is mapped to a node location code. The sequential relationship of the flow events is determined based on the time-series index value, and the spatial topology relationship of the flow events is determined based on the node location code, thus obtaining the directed dependency relationship between nodes. The transitivity of the directed dependency relationship is constrained and verified. It is checked whether the time-series index value satisfies the monotonically increasing requirement and whether the node location code satisfies the path reachability requirement. The directed dependency relationship that passes the constraint verification is taken as the flow link structure.

3. The method according to claim 1, characterized in that, Establishing a bidirectional binding relationship between the product's unique feature vector and each node in the circulation link structure includes: Extract the feature components from the product uniqueness feature vector, associate and map the feature components with the node identification information of each node in the flow link structure, and generate a positive index relationship between the feature vector and the node; Based on the node identification information of each node in the flow link structure, a reverse query is performed to find the unique feature vector of the product associated with it, and a reverse index relationship between the node and the feature vector is generated. The forward index relationship and the reverse index relationship are combined to form a two-way binding relationship, and the two-way binding relationship is stored in the binding relationship record table of the flow link structure.

4. The method according to claim 1, characterized in that, At each transaction event, an irreversible hash transformation is performed on the product's unique feature vector, and the transformation result is embedded into the transaction record of the corresponding node to obtain an anti-tampering traceability certificate, including: When each circulation event occurs, the node identifier information and circulation timestamp of the node corresponding to the current circulation event are extracted. The node identifier information and circulation timestamp are used as salt input parameters and combined with the product unique feature vector to generate input data to be transformed. An irreversible hash transformation operation is performed on the input data to be transformed to obtain a hash transformation result, and a fixed-length hash digest value is extracted from the hash transformation result; the hash digest value, the flow timestamp, the node identification information, and the positive index relationship identifier in the two-way binding relationship are encapsulated into a flow record data block; The flow record data block is embedded into the flow record field of the node corresponding to the current flow event in the flow link structure, and the flow record field is sealed to prevent subsequent modification; The transfer record fields of all nodes in the transfer link structure are organized in a chain according to the temporal relationship of the transfer timestamp to form the tamper-proof traceability certificate.

5. The method according to claim 1, characterized in that, A verification rule set is constructed based on the tamper-proof traceability certificate. The verification rule set includes consistency verification rules for the unique feature vector of the product and integrity verification rules for the circulation link structure, including: Based on the method of generating the hash digest value in the tamper-proof traceability certificate, a feature vector comparison benchmark for consistency verification is obtained by reverse derivation. The feature vector comparison benchmark is then established to correspond with the feature components in the product uniqueness feature vector, and a feature vector difference measurement criterion is set based on the correspondence. The feature vector difference measurement criterion and the feature vector comparison benchmark are combined to form the consistency verification rule. The directed dependencies of the flow link structure are extracted from the tamper-proof traceability certificate. A predecessor-successor association matrix between nodes is constructed based on the directed dependencies. Link continuity constraints and temporal monotonicity constraints are set based on the predecessor-successor association matrix. The link continuity constraints are used to detect whether there are missing nodes or broken paths in the flow link structure. The temporal monotonicity constraints are used to detect whether the flow timestamp of the current flow event satisfies a strict increasing relationship. The link continuity constraints, the temporal monotonicity constraints, and the predecessor-successor association matrix are combined to form the integrity verification rule. The consistency verification rules and the integrity verification rules are encapsulated into the verification rule set.

6. The method according to claim 5, characterized in that, According to the consistency verification rules, comparing the physical characteristic data of the product to be verified with the tamper-proof traceability certificate includes: The physical characteristic data of the product to be verified is converted into a feature vector to be verified. The feature vector comparison benchmark and the feature vector to be verified are dimensionally aligned. The feature component in the feature vector to be verified is compared with the feature component at the corresponding position in the feature vector comparison benchmark. The feature deviation of each dimension is obtained. The number of dimensions that exceed the allowable range of the feature deviation of each dimension is counted according to the feature vector difference measurement criterion. The feature consistency judgment result is determined according to the ratio of the number of dimensions to the total number of dimensions. Extract the flow record data block of the last node of the flow link structure from the tamper-proof traceability certificate; concatenate the feature vector to be verified with the node identifier information of the last node of the flow link structure and the flow timestamp, and perform the same irreversible hash transformation operation as in the tamper-proof traceability certificate on the concatenation result to obtain the hash digest value to be verified; The hash digest value to be verified is compared bit by bit with the hash digest value in the flow record data block to obtain the hash consistency determination result. Perform a logical AND operation between the feature consistency determination result and the hash consistency determination result to obtain the comparison verification result.

7. The method according to claim 1, characterized in that, The anti-counterfeiting verification is performed by traversing and verifying the circulation link structure according to the verification rule set. When the transformation sequence of the product unique feature vector completely matches the embedding result of each node in the circulation link structure and the circulation link structure satisfies the integrity verification rule, the anti-counterfeiting verification passes, including: The traversal path of the flow link structure is determined according to the predecessor-successor association matrix in the integrity verification rule; each node in the flow link structure is visited sequentially according to the traversal path, and the hash digest value and flow timestamp are extracted to form an embedding result sequence; The unique feature vector of the product is concatenated and combined with the node identification information of each node in the traversal path and the flow timestamp in sequence, and then subjected to irreversible hash transformation. A transformation sequence is generated according to the access order of the traversal path. Each hash digest value in the transformed sequence is compared one by one with the hash digest values ​​of the corresponding node positions in the embedded result sequence to obtain a sequence matching determination result; according to the link continuity constraint in the integrity verification rule, it is detected whether there are missing nodes or path interruptions in the traversal path to obtain a continuity determination result; according to the temporal monotonicity constraint in the integrity verification rule, it is detected whether the flow timestamps in the embedded result sequence satisfy a strict increasing relationship to obtain a temporal determination result. Perform a triple logical AND operation on the sequence matching determination result, the continuity determination result, and the timing determination result. When all three determination results are true, the anti-counterfeiting verification is confirmed to be successful.

8. An intelligent anti-counterfeiting system for product traceability management based on the Internet of Things, used to implement the method as described in any one of claims 1-7, characterized in that, include: The first unit is used to perform multi-dimensional feature vectorization processing on the physical feature data of the goods to obtain the unique feature vector of the goods, and to perform time-series correlation analysis on the circulation event data of the goods to obtain the circulation link structure. The second unit is used to establish a two-way binding relationship between the unique feature vector of the product and each node in the circulation link structure. When each circulation event occurs, the unique feature vector of the product is subjected to an irreversible hash transformation and the transformation result is embedded into the circulation record of the corresponding node to obtain an anti-tampering traceability certificate. The third unit is used to construct a set of verification rules based on the tamper-proof traceability certificate. The set of verification rules includes consistency verification rules for the unique feature vector of the product and integrity verification rules for the circulation link structure. The fourth unit is used to compare the physical feature data of the product to be verified with the anti-tampering traceability certificate according to the consistency verification rules, and to traverse and verify the circulation link structure according to the verification rule set. When the transformation sequence of the product unique feature vector completely matches the embedding result of each node in the circulation link structure and the circulation link structure satisfies the integrity verification rules, the anti-counterfeiting verification is passed.

9. An electronic device, characterized in that, include: processor; Memory used to store processor-executable instructions; The processor is configured to invoke instructions stored in the memory to execute the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having computer program instructions stored thereon, characterized in that, When the computer program instructions are executed by the processor, they implement the method described in any one of claims 1 to 7.

Citation Information

Patent Citations

  • Commodity source-tracing method and device based on block chain, and electronic equipment

    CN108694594A

  • Product traceability record processing method and device and electronic equipment

    CN108876405A

  • Quick tracing method and system based on two-dimensional code

    CN110717765A

  • Image tracing method and device

    CN119723107A

  • Food information tracing method and system based on laser code spraying

    CN119963220A

Cited By

  • Electronic commerce supply chain commodity intelligent traceability management method

    CN121581900A