Jewelry handicraft anti-counterfeiting traceability method and system based on big data

Abstract

The invention relates to the technical field of anti-counterfeiting traceability, in particular to a jewelry handicraft anti-counterfeiting traceability method and system based on big data, and the method comprises the following steps: collecting surface scratches, inclusion distribution, optical reflection and concave-convex data of jewelry handicrafts when the jewelry handicrafts leave a factory, comparing the weight and size of circulation nodes and an outer package mark to form an offset signal, and sending the offset signal to a database; and screening the identity and the time sequence of the recovery node, matching voucher number identification abnormity, and outputting link parameter mapping characteristics. According to the method, unique digital fingerprints are formed by synchronously collecting multi-dimensional jewelry real object microstructure information and circulation full-link identity data, dynamic coupling of surface detail features, identity codes and node sequences is achieved on a data layer, and by means of automatic parameter consistency judgment, link offset screening and node behavior cross verification, the dynamic coupling of the surface detail features, the identity codes and the node sequences is achieved. Identity abnormity and circulation breakpoints are found in real time, a risk source is accurately positioned, whole-process data mapping is constructed, and authenticity guarantee of a jewelry handicraft traceability chain and safety trust of a circulation link are improved.

Description

Technical Field

[0001] This invention relates to the field of anti-counterfeiting and traceability technology, and in particular to a method and system for anti-counterfeiting and traceability of jewelry and handicrafts based on big data. Background Technology

[0002] The field of anti-counterfeiting and traceability involves identifying, recording, and managing the identity, information, and pathways of goods and items during their production and distribution processes. This aims to prevent counterfeiting and ensure the traceability of genuine products. It is widely applied in high-value or high-risk sectors such as food, pharmaceuticals, jewelry, and art. Through information collection, coding, database management, identity verification, and full-process recording, a technical system for anti-counterfeiting identification and full-chain information traceability is constructed. Traditional anti-counterfeiting and traceability methods for jewelry and handicrafts involve attaching physical anti-counterfeiting marks to the items and registering them throughout their distribution process using an information management system. This approach focuses on ensuring the authenticity and traceability of the jewelry's identity information throughout the entire distribution process. It typically uses barcodes or QR codes, recording product information, transaction records, and distribution paths manually or through local area networks. Fixed query devices then complete identity verification and traceability of information at certain points.

[0003] Existing technologies rely on static labeling and manual operation, which have bottlenecks in data synchronization and link consistency management. They are prone to data disconnect and identity verification failure due to omission of node information, label replacement, or time sequence disorder during circulation. Manual input is prone to registration errors, and the status of goods is difficult to trace when nodes switch. Information silos lead to the accumulation of potential risks. In actual business, problems such as product substitution, identity distortion, and link breakage frequently occur, making it difficult to support the full-process identity verification and security traceability of high-value jewelry in multiple scenarios. Summary of the Invention

[0004] The purpose of this invention is to address the shortcomings of existing technologies by proposing a big data-based anti-counterfeiting and traceability method and system for jewelry and handicrafts.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: a method for anti-counterfeiting and traceability of jewelry and handicrafts based on big data, comprising the following steps: S1: Based on the manufacturing process of jewelry and craft products, collect data on the direction and length of scratches on the surface of jewelry and craft products, distribution of natural inclusions, optical reflection intensity and micron-level unevenness data, integrate the information into text and hash code, and then bind it with the finished product certificate number to obtain a unique identifier of the microstructure. S2: Based on the unique identifier of the microstructure, compare the weight, size and unique mark of the outer packaging of the logistics node, analyze the changes in the corresponding data of the previous node, identify the key nodes of the change range, and obtain the node parameter offset signal. S3: Based on the node parameter offset signal, filter the microstructure code sequence and the unique mark of the outer packaging of the recycled nodes, manually compare the node identification number with the historical record, and judge the node consistency through handover time data and time series analysis to obtain the node time series abnormal characteristics; S4: Based on the node time-series anomaly characteristics, determine the correspondence between the recycling voucher number and the finished product certificate number, match the voucher number and certificate number, identify the abnormal number, and integrate the node deviation data for interactive verification to obtain the voucher mapping anomaly index.

[0006] The present invention is improved in that the unique identifier of the microstructure includes a feature summary, an identifier index, and a data collection label; the node parameter offset signal includes a state discrimination code, a flow change item, and node marker data; the node temporal anomaly feature includes an anomaly type, a temporal identification code, and a node comparison label; and the credential mapping anomaly index includes a verification result item, a certificate anomaly label, and a credential comparison identifier.

[0007] The present invention is improved in that the step of obtaining the unique identifier of the microstructure is specifically as follows: S111: Based on the manufacturing process of jewelry and handicrafts, the data on the direction of scratches collected are analyzed. Combined with the scratch trajectory and natural inclusion distribution information under the standard sequence, all features in the measurement area are aggregated and analyzed to screen the morphological parameters with correlation and obtain the scratch inclusion combination factor. S112: Based on the scratch inclusion combination factor, compare it with the collected optical reflection intensity data, analyze the surface micron-level concavity and convexity morphology of each spatial region, calculate the joint distribution of each parameter, judge the rationality of the encoding order, and obtain the structural feature mapping segment. S113: Based on the structural feature mapping fragment, analyze its structural rules, select the mapping content that establishes a one-to-one relationship with the finished product certificate number, judge the uniqueness binding effect, and adjust the data mapping structure between the text and the number to obtain the unique identifier of the microstructure.

[0008] The present invention is improved in that the step of obtaining the node parameter offset signal is specifically as follows: S211: Based on the unique identifier of the microstructure, analyze the physical weight and size of the jewelry and handicrafts at the current logistics node, compare the same data from the previous logistics node, calculate the change range between the two, determine the weight and size change trend of each link, and obtain the node physical parameter change sequence. S212: Based on the sequence of changes in the physical parameters of the nodes, determine the intersection of the unique identifier of the outer packaging of the current node with the identifier of the previous node, analyze the trend of the change magnitude and the matching of the sets, and obtain the set of parameter offset nodes; S213: Based on the set of parameter offset nodes, compare the weight change range, size change range and the number of intersections of the outer packaging markings of each node, calculate the parameter offset degree, sort all nodes, and obtain the node parameter offset signal.

[0009] The present invention is improved in that the step of obtaining the node timing anomaly features is specifically as follows: S311: Based on the node parameter offset signal, analyze the microstructure code sequence of the recycling node, compare the current outer packaging unique mark and node identity number with the historical node identity number item by item, determine whether the node identity number has been registered in the historical path, filter out the node identity numbers that have not appeared, and obtain a set of non-duplicate node numbers. S312: Based on the set of non-repeating node numbers, analyze the handover order of adjacent nodes, compare the time offset of the node identity sequence with the time difference of the packaging mark, adjust the time association method, optimize the coupling relationship, obtain the average temporal coupling difference, and then determine the coupling difference category to obtain the time offset identification signal. S313: Based on the time offset identification signal, determine its distribution in the node sequence, analyze the order of the corresponding nodes, compare the identity number and registration order of adjacent nodes, filter out nodes with reversed or skipped order, optimize the time sequence discrimination criteria, and obtain node time sequence abnormal characteristics.

[0010] The present invention is improved in that the step of obtaining the certificate mapping anomaly indicator is specifically as follows: S411: Based on the node timing anomaly characteristics, analyze the field arrangement of the recycling voucher number and the finished product certificate number, compare the order and content of each group of number characters one by one, determine if there are inconsistencies in field order or character position, and obtain the field mapping deviation combination. S412: Based on the field mapping deviation combination, combined with the node identity number, time sequence identification code and handover time, analyze the correlation change between the number and the node identity, determine the abnormal order of the number's flow time between nodes and the matching situation of the node mark, and obtain the node path difference characteristics. S413: Based on the node path difference characteristics, compare whether the field structure anomalies and node timing anomalies of each number occur synchronously, analyze their correspondence, filter out the set of numbers that cannot correspond to each other, and obtain the voucher mapping anomaly index.

[0011] The present invention is improved in that the steps further include: S5: Based on the abnormal index of the certificate mapping, adjust the finished product certificate number and node identity number, call the surface scratch direction of the jewelry leaving the factory, establish a three-parameter comparison sequence according to the node order, judge the consistency of the link node parameter order, and obtain the link parameter mapping feature. The link parameter mapping features include sequence consistency items, node relationship indexes, and parameter comparison labels.

[0012] The present invention is improved in that the step of obtaining the link parameter mapping feature is specifically as follows: S511: Based on the aforementioned certificate mapping anomaly index, analyze the correspondence between the finished product certificate number and the node identity number, and combine the surface scratch direction in the factory data to organize the parameters of all nodes according to the factory, circulation and recycling links. The three parameters are arranged according to the order of circulation to obtain the three-parameter sequence group. S512: Based on the three-parameter sequence group, compare the certificate number, identity number and scratch direction arrangement of each node, calculate the difference in the order correspondence of the three parameters in the node, check the correlation between the parameter order one by one, and obtain the node order consistency performance. S513: Based on the consistency of the node order, judge the parameter arrangement performance, filter out nodes that are inconsistent or change in order during the order comparison, organize the number and order difference characteristics of each node, and obtain the link parameter mapping characteristics.

[0013] A big data-based anti-counterfeiting and traceability system for jewelry and handicrafts, the system comprising: The micro-feature coding module is based on the manufacturing process of jewelry crafts. It analyzes the surface scratch direction data, obtains the scratch length and natural inclusion distribution in a standard order, and then combines the optical reflection intensity parameters. It collects micron-level concavity and convexity data through a surface measuring instrument, organizes all information into a standardized text, processes it with hash encoding, and associates it one-to-one with the finished product certificate number to obtain a unique identifier for the microstructure. The node attribute comparison module compares the physical weight and size of the logistics node based on the unique identifier of the microstructure, compares the current parameters with the data of the previous node according to the standard order, analyzes the change range of each item, judges the intersection of the set of the current outer packaging unique mark and the previous node mark, identifies the key nodes of parameter change range, and obtains the node parameter offset signal. The recycling consistency determination module filters the microstructure code sequence registered by the recycling node based on the node parameter offset signal, reads the unique mark of the current outer packaging, compares the node identity number with the historical record by manual means, judges whether the time interval is abnormal based on the handover time data, and evaluates the node consistency by combining the identity number and time sequence analysis to obtain the node time sequence abnormal characteristics. Based on the node time-series anomaly characteristics, the voucher anomaly detection module determines the correspondence between the recovered voucher number and the finished product certificate number. It performs field position matching on all voucher numbers and certificate numbers to identify abnormal numbers that cannot be matched one-to-one, and performs interactive verification in conjunction with node deviation data to obtain voucher mapping anomaly indicators. Based on the certificate mapping anomaly index, the link feature extraction module adjusts the finished product certificate number and node identity number in the entire process, retrieves the surface scratch direction in the factory data, establishes a three-parameter comparison sequence according to the order of factory, circulation and recycling, calculates the consistency of parameter order of each node in the entire link, and identifies the case of order discrepancy, thereby obtaining the link parameter mapping features.

[0014] Compared with the prior art, the advantages and positive effects of the present invention are as follows: In this invention, by synchronously collecting multi-dimensional microscopic structural information of physical jewelry and identity data from the entire circulation process, a unique digital fingerprint is formed. Surface detail features, identity codes, and node sequences are dynamically coupled at the data layer. With the help of automated parameter consistency judgment, link offset screening, and node behavior cross-verification, identity anomalies and circulation breakpoints are detected in real time, accurately locating the source of risk. A full-process data mapping is constructed, comprehensively improving the authenticity guarantee of the jewelry traceability chain and the security and trust of the circulation process. It covers the entire process from factory to recycling, forming a digital risk management closed loop, and enhancing the depth and accuracy of the industry's dynamic traceability of high-value physical objects. Attached Figure Description

[0015] Figure 1 This is a flowchart of the main steps of the present invention; Figure 2 This is a flowchart illustrating the process of obtaining the unique identifier of the microstructure in this invention. Figure 3 This is a flowchart illustrating the acquisition of node parameter offset signals in this invention. Figure 4 This is a flowchart illustrating the acquisition of node timing anomaly features in this invention. Figure 5 This is a flowchart illustrating the process of obtaining abnormal indicators for voucher mapping in this invention. Figure 6 This is a flowchart illustrating the process of obtaining link parameter mapping features in this invention. Detailed Implementation

[0016] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0017] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0018] Example: Please see Figure 1 This invention provides a technical solution: a method for anti-counterfeiting and traceability of jewelry and handicrafts based on big data, comprising the following steps: S1: Based on the manufacturing process of jewelry, analyze the surface scratch direction data, obtain the scratch length and natural inclusion distribution in a standard order, and then combine the optical reflection intensity parameters. Collect micron-level concavity and convexity data through a surface measuring instrument, organize all information into a standardized text, process it with hash encoding, and associate it one-to-one with the finished product certificate number to obtain a unique identifier of the microstructure. S2: Based on the unique identifier of the microstructure, compare the physical weight and size of the logistics node, compare the current parameters with the data of the previous node according to the standard order, analyze the change range of each item, determine the intersection of the set of the current outer packaging unique mark and the previous node mark, identify the key nodes of parameter change range, and obtain the node parameter offset signal. S3: Based on the node parameter offset signal, filter the microstructure code sequence registered by the recycling node, read the unique mark of the current outer packaging, compare the node identity number with the historical record by manual means, judge whether the time interval is abnormal based on the handover time data, and evaluate the node consistency by combining the identity number and time series analysis to obtain the node time series abnormal characteristics. S4: Based on the node time-series anomaly characteristics, determine the correspondence between the recycling voucher number and the finished product certificate number. Use field position matching for all voucher numbers and certificate numbers to identify abnormal numbers that cannot be matched one by one. Combine node deviation data for interactive verification to obtain voucher mapping anomaly indicators. S5: Based on the abnormal indicators of the voucher mapping, adjust the finished product certificate number and node identity number of the entire process, retrieve the surface scratch direction in the factory data, establish a three-parameter comparison sequence according to the order of factory, circulation and recycling, calculate the consistency of the parameter order of each node in the entire chain, identify the situation of inconsistent order, and obtain the link parameter mapping characteristics.

[0019] The unique identifier of the microstructure includes feature summary, identifier index, and data collection label; the node parameter offset signal includes state discrimination code, flow change item, and node tag data; the node time sequence anomaly feature includes anomaly type, time sequence identification code, and node comparison label; the certificate mapping anomaly indicator includes verification result item, certificate anomaly label, and certificate comparison identifier; and the link parameter mapping feature includes sequence consistency item, node relationship index, and parameter comparison label.

[0020] In S1, the "factory exit stage" refers to the operation and data collection stage when jewelry is manufactured, inspected, and packaged, and ready to be shipped from the manufacturer to the next stage (such as distribution companies or warehousing centers). Surface scratch direction data refers to the spatial distribution direction information of tiny scratches on the jewelry surface collected by professional imaging or testing equipment. These scratches are usually natural or generated during the manufacturing process, difficult to replicate, and are one of the core physical characteristics for traceability. The "standard sequence" refers to the sequential collection and arrangement of various parameters according to a pre-defined fixed order, ensuring consistent data format for subsequent processing and hash coding. The "optical reflection intensity parameter" refers to the intensity of light reflected from the surface of the jewelry under specific lighting conditions, measured by a spectrometer or professional equipment, and is one of the unique attributes that distinguishes individual pieces. The "surface measuring instrument" is a professional instrument used to measure the microscopic undulations (concavity and convexity) of the jewelry surface, such as a 3D laser scanner or white light interferometer, which can accurately reconstruct the surface morphology and provide basic data for subsequent coding. "One-to-one association" refers to uniquely binding each set of collected physical characteristic hash codes to the finished product certificate number of the jewelry, preventing multiple pieces of jewelry from corresponding to the same code or number, and achieving unique identification.

[0021] In S2, logistics nodes refer to every physical location (such as warehousing, sorting, transportation, stores, and inspection points) that jewelry and handicrafts pass through during the circulation process. Each node collects and records relevant data. Comparing current parameters with data from the previous node means that after collecting the jewelry's weight, size, and unique outer packaging identifier at the current node, the data is compared one by one with the data collected at the previous node to verify the consistency of the physical goods during the logistics process. Variation range refers to the range or difference in the value of the same parameter (such as weight and size) between the current node and the previous node, used to determine whether there are any abnormal changes in the jewelry during circulation. Set intersection refers to the set intersection operation between the unique outer packaging identifier collected at the current node and the corresponding parameter at the previous node to detect whether there are phenomena such as packaging replacement or substitution. Nodes with critical variation range refer to nodes whose parameter variation is significantly higher than other nodes among all logistics nodes. After screening, these nodes are designated as key monitoring nodes or nodes for abnormal warning.

[0022] In S3, the recycling node refers to the designated or authorized recycling enterprise, store, or recycling center when jewelry enters the recirculation, recycling, or second-hand trading stages. It serves as a data collection and identity verification node in the later stages of the jewelry's lifecycle. The microstructure code sequence refers to the unique identifier registered at the recycling node, obtained by encoding microscopic surface feature parameters, used for full-process traceability comparison. The node identity number refers to the unique authentication number of each circulation and recycling node, used to confirm the actual circulation path, authorization information, and responsibility attribution of the jewelry. The historical record refers to the node identity numbers and related circulation records of all flows of the jewelry from its manufacturing stage to the present, facilitating comparison with current records. The system compares the recycling node number; the handover time data refers to the specific timestamp of each jewelry transfer between nodes, facilitating the tracing of time sequence and circulation link; whether the time interval is abnormal refers to whether the interval between the current handover time and the previous node's handover time exceeds the set rules, used to determine whether there is unauthorized or abnormal flow in the process; time sequence analysis refers to the time sequence structure analysis of the circulation path based on the handover time and sequence of each node, to verify the compliance and anomalies of the process; node consistency refers to the consistency or inconsistency between the current recycling node and all nodes that the historical flow has passed through in terms of identity, time, etc., which is an important basis for judging whether the link is trustworthy.

[0023] In S4, the correspondence refers to whether there is a unique, one-to-one correspondence between the recycling voucher number and the finished product certificate number, which is used for identity authentication and anti-counterfeiting; field position matching refers to comparing the arrangement, order, and content of each character, number, or field of the two numbers to see if they are completely consistent, and to check for counterfeit or incorrect numbers; abnormal numbers refer to voucher or certificate numbers that have no corresponding relationship or do not match after matching, which are usually used as risk markers or traceability doubts; node deviation data refers to abnormal and different data generated during the comparison of the aforementioned recycling nodes with historical nodes in terms of identity, time sequence, parameters, etc.; interactive verification refers to comprehensively analyzing and verifying the abnormal numbers and node deviation data to determine whether there is a correlation between the two, supporting risk judgment.

[0024] In S5, the entire process refers to the complete chain and all operational processes of jewelry and handicrafts from the factory, circulation to recycling; the three-parameter comparison sequence refers to the comparison data sequence established in chronological order based on three core parameters: finished product certificate number, node identification number, and surface scratch direction; each node in the entire chain refers to each actual node in all stages that specifically participates in data collection, recording, and transfer; sequence consistency refers to judging whether the recording order of each node in the three-parameter comparison sequence is completely consistent with the standard chain, which is used to verify the authenticity of the traceability chain; sequence discrepancies refer to differences or errors in the node order, parameter registration order, etc., after actual comparison, compared with the normal process standard, which is a direct manifestation of system risk or potential anomalies.

[0025] Please see Figure 2The specific steps for obtaining the unique identifier of the microstructure are as follows: S111: Based on the manufacturing process of jewelry and handicrafts, the data on the direction of scratches collected are analyzed. Combined with the scratch trajectory and natural inclusion distribution information under the standard sequence, all features in the measurement area are aggregated and analyzed to screen the morphological parameters with correlation and obtain the scratch inclusion combination factor. Based on surface scratch direction data obtained during the manufacturing process of jewelry, a microscopic imaging device was used to collect the start and end coordinates of all scratches within a 2mm × 2mm measurement area. By sequentially connecting these coordinates, a scratch trajectory data set was formed. Under a standard order, the length of all scratches was calculated and sorted from shortest to longest. For example, in a sample, 47 scratches were obtained, ranging in length from 2.1 micrometers to 18.4 micrometers. After sorting, they were numbered, and a direction distribution table was generated based on the orientation of each scratch within the area. The width and depth parameters of the corresponding scratches were recorded. Simultaneously, the distribution of natural inclusions was detected within the same area, and the coordinates and dimensions of the center point of each inclusion were recorded. The spatial distance between each inclusion and the nearest scratch was calculated sequentially. For example, inclusion number B13 was 11.2 micrometers away from the nearest scratch. If this distance was less than 15... Micrometers are considered relevant and are included in the initial combination. Among all the initially screened combinations, combinations with a scratch width to inclusion radius between 0.5 and 2.5 are further selected. For example, a scratch width of 9 micrometers and an inclusion radius of 4.2 micrometers have a ratio of 2.14, which meets the condition and is retained for inclusion. Then, for each scratch-inclusion combination pair, its spatial proximity (distance less than 10 micrometers is considered perfectly close), length difference (length difference from the sample median is less than 3 micrometers is considered close), and geometric shape (width-to-height ratio and inclusion similarity greater than 80% are considered a match) are evaluated separately. Scores are assigned according to whether the three conditions are met. If the total score of the combination pair is greater than 75 out of 100, it is selected as a formal combination factor. In this analysis, 9 combinations were confirmed to meet the set conditions, and their morphological characteristics were summarized to form the scratch-inclusion combination factor.

[0026] S112: Based on the scratch inclusion combination factor, compare it with the collected optical reflection intensity data, analyze the surface micron-level concavity and convexity morphology of each spatial region, calculate the joint distribution of each parameter, judge the rationality of the encoding order, and obtain the structural feature mapping segment. The corresponding surface micro-regions are mapped onto the reflected light intensity measurement data. Within the same measurement area, sub-regions are divided in 50-micrometer increments, generating a total of 1600 regional units. The reflected light intensity value of each unit is collected, ranging from 0.42 to 0.91. If the difference in reflection intensity between a unit and its adjacent regions exceeds 0.12, a significant abrupt change in optical characteristics is considered to exist. For example, region R215 has a light intensity of 0.66, while its adjacent region R216 has 0.81, a difference of 0.15, and is marked as a high-contrast region. Within each sub-region corresponding to a scratch-inclusion combination, the scratch depth, inclusion size, and light intensity are combined as joint parameters. For example... One set of data includes a scratch depth of 4.6 micrometers, an inclusion radius of 11.3 micrometers, and a light intensity of 0.74. Next, the frequency of occurrence of all parameter combinations is counted. If a combination appears more than 3 times in the same position, it is considered a stable structural feature. Its position index is extracted, and then the parameter coding order in the sample is compared with the standard order. If the 3 occurrences are all located in the same segment of the standard sequence, the coding order is judged to be reasonable. For example, if 3 out of 5 samples have a certain combination between parameter sequence numbers 150 and 170, then this segment is recorded as the feature mapping content. Structural items that meet the requirements of stable reflection, consistent combination parameters, and consistent sequence are integrated to form a structural feature mapping segment.

[0027] S113: Based on the structural feature mapping fragment, analyze its structural rules, select the mapping content that establishes a one-to-one relationship with the finished product certificate number, judge the uniqueness binding effect, and adjust the data mapping structure between the text and the number to obtain the unique identifier of the microstructure. The repetitive distribution characteristics of each segment within and between samples were analyzed one by one. First, the frequency of each segment appearing in the test area of ​​a single jewelry sample was statistically analyzed. For example, segment F17 appeared 4 times in 5 test areas of sample A, with a frequency of 80%, which meets the high-frequency segment judgment threshold. Then, the spatial orientation of the same segment in different test areas was compared, and the difference in orientation angle in different areas was recorded. If the angle difference is less than 5 degrees, the orientation is considered to be consistent. For example, the orientation of segment F17 in area 1 and area 3 is 42° and 45° respectively, with a difference of 3°, which meets the consistency standard and is retained as a spatial repetitive structure item. Then, this type of spatial repetitive structure is bound to the certificate number of the jewelry. Each group of structure mapping segments is encoded into a 32-bit character code using rules, and then linked to the certificate number. The combination of numbers forms a unique identifier combination. For example, the certificate number of sample A is JZ20250807A105, and the code generated by fragment F17 is QX9G-MN8T-W6JZ-43KR. The two are bound together to form the record JZ20250807A105#QX9G-MN8T-W6JZ-43KR. Then, it is checked whether the structure code appears repeatedly under other certificate numbers. If a duplicate is found, such as the same code appearing under JZ20250807B118, the fragment coding method needs to be adjusted. For example, a new scratch direction parameter or displacement data of the inclusion in the three-dimensional position can be added to participate in the coding to form a new round of independent coding. When there is no duplicate relationship between all structural fragment codes and certificate numbers, the binding process is considered to be complete, and a unique identifier of the microstructure is obtained.

[0028] Please see Figure 3 The specific steps for obtaining the node parameter offset signal are as follows: S211: Based on the unique identifier of the microstructure, analyze the physical weight and size of the jewelry and handicrafts at the current logistics node, compare the same data of the previous logistics node, calculate the change range between the two, determine the trend of weight and size change in each link, and obtain the sequence of changes in node physical parameters. After retrieving the unique identification number of the jewelry item from the current logistics node, the current weight and size parameters of the item are extracted. At the current node, the weight measured by an electronic scale is 15.82 grams, and the maximum length, width, and height measured by calipers are 38.5 mm, 26.2 mm, and 10.4 mm. Comparing this with historical data registered at the previous logistics node, the weight was 15.79 grams, the length was 38.6 mm, the width was 26.0 mm, and the height was 10.5 mm. Subtracting these values ​​directly yields a weight change of +0.03 grams, a length change of -0.1 mm, a width change of +0.2 mm, and a height change of -0.1 mm. Further calculations are made of the relative change percentage for each parameter; for example, the weight change rate is 0.19%, and the width change rate is 0.77%. All change data is compiled into a physical parameter change record table for the current node. Then, the change values ​​are compared with the set change range thresholds to determine the allowable range during normal jewelry logistics. The weight variation range is ±0.05 grams, and the size variation range is ±0.3 millimeters. If any variation exceeds this range, it is marked as abnormal. In this example, all variations are within the normal range, so the variation trend is recorded. Then, the variation value recorded at this node is compared with the variation value of the previous node to obtain the variation trend. That is, compared with the previous node, the weight of this node is slightly increased, the length is slightly shortened, the width is slightly increased, and the height is slightly shortened. In five consecutive logistics nodes, if a parameter is found to show an increasing trend three times in a row and the variation value gradually increases, for example, the width changes from +0.1 mm to +0.3 mm and then +0.5 mm, then the trend is marked as "positive enhancement" trend. Conversely, if the direction of change fluctuates repeatedly, such as once increasing and once decreasing, it is recorded as "no continuity". The physical parameter variation information of each node is recorded and numbered in chronological order, forming a physical parameter variation sequence with node index, parameter item, variation range, and variation direction.

[0029] S212: Based on the sequence of changes in node physical parameters, determine the intersection of the unique identifier of the current node's outer packaging with the identifier of the previous node, analyze the trend of the change magnitude and the set matching situation, and obtain the set of parameter offset nodes; The unique identifier of the current node's outer packaging is read, and the QR code information or NFC chip code attached to the packaging surface is extracted and recorded as P20250807Z305. This is compared with the outer packaging identifier P20250807Z304 recorded by the previous node. If the two are completely identical, they are considered consecutive identifiers. If there is a difference, the identifier intersection judgment procedure is entered. The intersection judgment is based on matching the character components bit by bit. For example, if the prefix P20250807 is the same in the two codes, and only the suffixes Z305 and Z304 are different, then a partial matching relationship is determined. If at least 80% of the field content is the same (e.g., 13 out of the first 16 bits are the same), it is judged as "high intersection". If it is less than 50%, it is judged as "low intersection". In actual operation, if the outer packaging code of the current node changes every If a change occurs every three times in the logistics process, one mark of difference is allowed in the rules. Within the allowed range, no abnormal record is made. Then, a comprehensive analysis is performed based on the changing trends of the various physical parameters obtained in the previous step. If the parameter changing trend is recorded as "no continuity" and the intersection level of the packaging mark is "low intersection", then the node is marked as an offset risk node. Taking a real case as an example, the width of the current node changes by +0.5 mm, which is higher than the set size offset baseline value of 0.3 mm, and the packaging mark changes to a character difference rate of 45% compared with the previous node. Both of these are judged as abnormal, so the current node is marked into the parameter offset node set. After each node analysis is completed, the node index, judgment result, change value and other data are automatically appended to this set for cross-comparison during subsequent traceability to form a complete parameter offset node set.

[0030] S213: Based on the set of parameter offset nodes, compare the weight change range, size change range, and the number of intersections with the outer packaging markings of each node, using the following formula: ; Calculate the parameter offset, sort all nodes, and obtain the node parameter offset signal, where... Indicates the first Parameter offset of each logistics node This represents the actual weight of the jewelry / craft item detected at the i-th logistics node. Indicates the first The actual weight of the same jewelry item detected at each logistics node. Indicates the first The actual dimensions of the jewelry and handicrafts detected at each logistics node. Indicates the first The actual dimensions of the same jewelry item detected at each logistics node Indicates the first The logistics node and the first The number of tags in the intersection of the sets of unique tags on the outer packaging of each logistics node.

[0031] Parameter offset refers to the abnormal deviation in the anti-counterfeiting and traceability logistics process of jewelry and handicrafts, which is used to comprehensively measure the changes in physical weight, size, and unique markings on the outer packaging between a single logistics node and the previous node. This parameter can reflect the overall consistency and stability of the physical properties and outer packaging markings of jewelry and handicrafts during the flow between logistics nodes. The higher the parameter offset value, the greater the change in weight, size, and packaging markings at that node, and the higher the risk of anomalies; conversely, it means that the parameters of that node are basically consistent with those of the previous node. Extract the currently detected weight from each node. ,size And the number of intersections with the wrapper marker of the previous node. At the same time, it retrieves the weight of the same piece of jewelry recorded in the previous node. and size In a practical example, the relevant parameters for the jewelry craft with node number 3 are set as follows: , , , , The corresponding normalized input value is , , , Substituting the normalized values ​​into the formula and performing the calculations sequentially, we obtain: The first item is ; The second item is ; The third item is ; The sum of the three terms is: ; The result indicates that the parameter offset obtained for jewelry crafts with node number 3 is... When the value exceeds the upper limit of the normal logistics node offset reference range [0, 0.18], it indicates that the node has multiple synchronous changes in the intersection of weight, size and packaging markings, showing significant logistics chain parameter offset characteristics. This result is directly used to determine whether the node belongs to the offset abnormal node, and the signal serves as an important input for subsequent recycling node identity tracing and process verification.

[0032] Please see Figure 4 The specific steps for obtaining node temporal anomaly features are as follows: S311: Based on the node parameter offset signal, analyze the microstructure code sequence of the recycling node. By comparing the current outer packaging unique mark and node identity number with the historical node identity number item by item, determine whether the node identity number has been registered in the historical path, filter out the node identity numbers that have not appeared, and obtain a set of non-duplicate node numbers. Each structural code in the sequence is identified and numbered. For example, the structural code collected during the recycling of a jewelry sample is [MC4D-T19X-V7QZ]. This code is precisely compared with the historical structural code records registered on the circulation chain. The matching method uses string shift matching, comparing character by character to see if there is a completely identical item. If the comparison result confirms that the structural code appears only once in the historical record, the structural consistency match is confirmed. Next, the unique identifier of the current node's outer packaging is read, for example, WRK230807-X54, and the node's identification number NID04567 is read. This identification number is used to identify the recycling store or authorized entity. Then, the entire historical node number list is retrieved to construct a historical number set sorted by time, such as [NID00001, NID00012, NID00057, NID00103, N ...

[00321] The current node number NID04567 is compared with each number in the set item by item to confirm whether a complete match has occurred. If no match is found after comparison, it is judged as "not registered". When performing the comparison, the string matching method is used to judge the consistency of the number. The judgment standard is that the characters are completely consistent and the length is consistent. If the current number has a character length of 7 but all the historical numbers are 6 characters, it will not be included in the matching item. In one actual sample, there are a total of 32 historically registered numbers. The current node number does not match any of them, and the character prefix is ​​different from the historical items. Therefore, it is confirmed that it has never appeared. The number is added to the temporary storage pool of non-repeating node numbers. The node numbers judged as non-repeating are uniformly recorded into the set, and their corresponding structure code and current packaging mark are recorded at the same time to realize the deduplication identification of the current recycling node and the historical circulation process, and obtain the set of non-repeating node numbers.

[0033] S312: Based on the set of unique node IDs, analyze the handover order of adjacent nodes, compare the time offset of the node identity sequence with the time difference of the packaging mark, adjust the time association method, and optimize the coupling relationship using the following formula: ; Obtaining average temporal coupling difference Then, the coupling difference category is determined to obtain the time offset identification signal, where, This represents the average temporal coupling difference. Representing the The handover time interval between adjacent nodes in a group Representing the Time offset of the group node identity number sequence. Representing the The time difference between the group packaging marking intervals. Represents the number of node pairs; The average temporal coupling difference reflects the overall deviation between the actual handover time and the theoretical time offset under the combined influence of identity, packaging and other links in the jewelry circulation path. The higher the value, the weaker the synergy and continuity between the temporal sequence of each node in the circulation process and the core elements such as identity and packaging, and the more likely abnormal risks will occur. The smaller the difference, the more normal the correspondence between the node temporal sequence and identity mark and the better the continuity of the link. The handover time interval for each node pair is extracted from the registration time field in the node database to obtain the duration of the time interval for each node pair, and the unit is standardized to minutes. For example, the handover time interval for node pair 1 is 24 minutes, for node pair 2 it is 36 minutes, and for node pair 3 it is 30 minutes. Then, the sequence structure of the node identification numbers is compared. By calculating the difference in the order of the node numbers and combining it with the standard time distribution function between nodes, the time offset corresponding to the identification number is calculated, which is 16 minutes, 30 minutes, and 22 minutes respectively. Next, the node packaging mark comparison results are extracted and mapped to the difference parameters in time form according to the dimensions such as image overlap ratio and similarity and difference of graphic features, which are 9 minutes, 12 minutes, and 14 minutes respectively. On this basis, the three parameters are normalized respectively. The normalized results for the handover time interval are 0.4, 0.72, and 0.6, the normalized results for the identification number time offset are 0.32, 0.6, and 0.44, and the normalized results for the packaging mark time difference are 0.18, 0.24, and 0.28. Then, the above data are substituted into the formula: Group 1: ; Group 2: ; Group 3: ; Calculate the average: ; This result indicates that the average temporal coupling difference If the value falls within the preset baseline range [0.2, 0.35], according to the difference level standard, this range corresponds to the "medium offset level", [0≤ <0.2] represents a low offset level, [0.35 < ≤1] indicates a high offset level, which means that there is a certain degree of inconsistency between the node handover order and the temporal relationship between its identity number sequence and packaging mark, but it has not reached a state of serious disorder. It is used to identify whether there are nodes with temporal jumps or reversed order. The formula introduces two multi-source heterogeneous temporal indicators, identity offset and packaging mark interval, as a coupling basis to achieve dynamic identification of the consistency of node sequence under a unified dimension, thereby forming a quantitative judgment basis for the rationality of node handover behavior.

[0034] S313: Based on the time offset identification signal, determine its distribution in the node sequence, analyze the order of the corresponding nodes, compare the identity number and registration order of adjacent nodes, filter out nodes with reversed or skipped order, optimize the time sequence discrimination criteria, and obtain node time sequence abnormal characteristics. Read its corresponding position number in the complete logistics node sequence and establish a node index table for that time offset point in the chain. For example, if the signal indicates an abnormal time interval, the 6th node is numbered NID00230. First, find the sequence position of NID00230 in the full process node list. The two nodes before and after it are NID00221 and NID00245. Record their registration times as 08:41, 08:57, and 08:39, and arrange them in chronological order as NID00221. <nid00230>NID00245 was registered earlier than NID00230, indicating a reverse registration. The comparison continues, checking if the ID number matches the registration order in the standard sequence. If the ID number increases but the timestamp decreases, it's marked as an inverted sequence. If the ID number jumps, such as the sudden appearance of NID00280 within the consecutive ID number range NID00110 to NID00119, it's judged as a skipped sequence. In this process, three inverted sequence nodes and two skipped sequence nodes were identified and recorded in the anomaly index list. To prevent false positives, it's also necessary to analyze whether the time offset exceeds the tolerance interval. For example, if the maximum allowable registration interval is set to 15 minutes, and the time interval between the current node and the previous node exceeds this value, it should be marked as abnormal even if the numbering order is correct. After merging the above judgments, a time sequence abnormality judgment record set is formed. Then, the records are sorted out, and the type, time interval value, node number and sequence position of each abnormality are extracted to determine the frequency of occurrence. If the number of abnormal nodes accounts for more than 10% of all nodes, the time sequence discrimination standard is updated, for example, the maximum allowable interval is adjusted to 12 minutes. The above identification results are sorted and summarized, and the identity number, time record and position relationship of all nodes with reversed order and skipped order are output to obtain the node time sequence abnormality characteristics.

[0035] Please see Figure 5 The specific steps for obtaining the abnormal voucher mapping indicator are as follows: S411: Based on the node timing anomaly characteristics, analyze the field arrangement of the recycling voucher number and the finished product certificate number, compare the order and content of each group of number characters one by one, determine the situation where there is a discrepancy between the field order or the character position, and obtain the field mapping deviation combination. Retrieve the recycling voucher number and the finished product certificate number generated at the time of manufacture for the corresponding jewelry and craft items registered at the recycling node. Separate the two sets of numbers into independent character fields and compare them by character position. For example, if the recycling voucher number is "RC-20250807-A0812" and the finished product certificate number is "CF-A0812-20250807", divide them into four fields: recycling vouchers are grouped as [RC][20250807][A0812], and finished product certificates are grouped as [CF][A0812][20250807]. When comparing the field order, it can be found that the order of "20250807" and "A0812" is completely opposite in the two numbers. The character composition is the same, but the arrangement position is different, indicating a sequence offset. Continue to analyze... The system compares the character arrangement within each field. For example, it checks whether the character order of "A0812" is consistent. If a character like "A0182" appears, it determines that there is a difference in character position and records it as a position mismatch. According to the set deviation identification standard, if more than two characters in the same field have different positions, it is determined to be a structural deviation field. The threshold is set as follows: if the number of different characters is no more than 1, it is considered acceptable; if the number of different characters is greater than or equal to 2, it is considered an inconsistent field. In this example, the number of character differences between "A0182" and "A0812" is 2, so the record is an inconsistent field. After comparing each pair of numbers, all field difference records are organized into a mapping deviation combination. At the same time, if two or more fields are misaligned or have character structure mismatch in the combination, the group of numbers is defined as a field structure abnormal number pair. The cumulative numbers form a list of field mapping deviation combinations.

[0036] S412: Based on the combination of field mapping deviations, combined with node identity number, time sequence identification code and handover time, analyze the correlation changes between number and node identity, determine the abnormal order of number flow between nodes and the matching of node tags, and obtain node path difference characteristics. The system retrieves the corresponding node's identity number, time sequence identification code, and handover time, establishing a mapping table between each number combination and node attribute information. First, it reads the recovery certificate number from the deviation combination, such as "RC-20250807-A0812," and extracts its registered node number as NID00457, with a registration time of August 7, 2025, at 14:03. It then compares this with historical node paths to see if a certificate number record exists that matches the certificate's field content. If no complete match is found, it's marked as a broken link between the number and node identity. The system then continues to determine the node's time sequence identification code in the entire process node sequence. For example, if the current node's identification code is IDX-0157, comparing it with the certificate number of its upstream node IDX-0156 ("CF-A0812-20250807"), this number and the current recovery certificate are already marked as a deviation combination in terms of field order and structure. Therefore, it's determined that the current node has an anomaly in the time sequence chain. The normal sequential transmission is judged by the following criteria: if a certain number has the same field content but different order in two consecutive nodes, and the time interval between the registration of the two nodes is less than 10 minutes, it is considered "rapid abnormal flow". If the time interval is greater than 30 minutes and the fields are misaligned, it is considered "lagging offset". For example, if the upstream node registration time is 13:54 and the current node registration time is 14:03, the time difference is 9 minutes, which meets the definition of "rapid abnormal flow". The node is marked as a priority abnormal marker. The mapping deviation combination of other fields is compared. If multiple number structure deviations are found in the same node and the handover time rules are inconsistent, the path offset index of the node is accumulated and recorded. The node path difference threshold is set to the point that the number of offset combinations exceeds 2, which means that the node path has a structural difference. The path offset statistics of all compared nodes are performed to form a node path difference feature list for subsequent interactive verification and source tracing judgment.

[0037] S413: Based on the differences in node paths, compare whether the field structure anomalies and node timing anomalies of each number occur synchronously, analyze their correspondence, filter out the set of numbers that cannot correspond to each other, and obtain the voucher mapping anomaly index. The process involves comparing each field structure anomaly with a node timing anomaly to see if they occur simultaneously under the same node number. The numbers of the field structure anomalies are matched against the node anomaly table. For example, if node NID00457 shows both "RC-20250807-A0812" and "CF-A0812-20250807" in out-of-order pairs and a reversed record of the IDX-0157 timing identifier, this node is marked as a "double anomaly node (structure and timing). The process continues by searching the entire path for node numbers with the same characteristics. All nodes showing structural anomalies but no timing anomalies, or vice versa, are filtered. The judgment criterion is set as follows: if a node has more than one field structure deviation and the corresponding timing anomaly level is level one or two, it is judged as a synchronization anomaly node. If only one type of anomaly occurs, it is marked as an "asynchronous anomaly node". The number of node numbers that simultaneously exhibit both types of anomalies is counted, and a number anomaly correspondence matrix is ​​constructed. In this matrix, if a number cannot be paired with another anomaly type in any dimension, it is marked as a "non-mutually mapped number". That is, there is no one-to-one correspondence between the field structure anomaly number and the time sequence anomaly number. For example, the number "RC-20250807-A0812" is structurally abnormal but has no time sequence node anomaly records within its time period. This number is then added to the initial set of voucher mapping anomalies. Through the above process, a joint comparison table of number field misalignment and time sequence node anomalies is constructed. All mutually exclusive or disconnected number combinations are filtered out, and the voucher mapping anomaly index set is output.

[0038] Please see Figure 6 The specific steps for obtaining link parameter mapping features are as follows: S511: Based on the abnormal indicators of the voucher mapping, analyze the correspondence between the finished product certificate number and the node identity number. Combined with the surface scratch direction in the factory data, organize the parameters of all nodes according to the factory, circulation and recycling links. The three parameters are arranged according to the order of circulation to obtain the three-parameter sequence group. For each set of abnormal indicators, the finished product certificate number and node identification number are associated and extracted, forming a mapping pair for each set of numbers. For example, "CF-A0812-20250807" and "NID00534" are considered as a pair of corresponding items. Then, the surface scratch direction record corresponding to this number is retrieved from the factory database. This data record contains the scratch direction vector angle value, start and end coordinate pairs, and scratch density index. The scratch direction recorded in the factory process is 41.2 degrees, and the density is 48 scratches per square millimeter. Next, all node numbers are sorted according to the circulation path. Starting from the factory node based on the node circulation timestamp, the three core parameters corresponding to each node are organized in sequence: the first is the certificate number, the second is the node identification number, and the third is the scratch direction description data. For example, node 1 is "CF-A0812". -20250807”"NID00001”"41.2 degrees, node 2 is "CF-A0812-20250807”"NID00018”"41.1 degrees, node 3 is "CF-A0812-20250807”"NID00534”"44.9 degrees". The three parameters of each node are summarized to form the parameter flow sequence of the jewelry. A unified timeline record is established. If the time corresponding to a certain node is 12:35 on August 7, 2025, it will be placed after the previous node 12:18 on August 7, 2025. After ensuring that the order is reasonable, all node data are arranged in order, and a three-parameter sequential sequence group is output. Each group contains the finished product certificate number, node number, and scratch direction of all recorded nodes in the production, circulation and recycling stages of the jewelry.

[0039] S512: Based on the three-parameter sequential sequence group, compare the certificate number, identity number and scratch direction arrangement of each node, calculate the difference in the order correspondence of the three parameters in the node, check the correlation between the parameter order one by one, and obtain the node order consistency performance. For each node, the three parameters are analyzed accordingly. The order of the certificate number, identification number, and scratch direction is compared to ensure consistency. First, nodes 1, 2, and 3 are identified, as their corresponding certificate numbers are consistent: "CF-A0812-20250807". The node numbering order is "NID00001", "NID00018", and "NID00534", with scratch directions of 41.2 degrees, 41.1 degrees, and 44.9 degrees respectively. Continuing with the fourth node as an example, numbered "NID00310", the scratch direction is 39.5 degrees. This indicates a directional deviation greater than 5 degrees. The criterion for judgment is that if the difference in scratch direction between two adjacent nodes exceeds 3 degrees, it is considered an abnormal direction. If there is a reverse order in the node numbers, for example, "NID00310" appears after "NID00018", it is considered an abnormal direction. After "534", the numbering order is marked as abnormal. Then, the three parameters are compared in order. If any two parameters do not meet the sorting requirements, they are recorded as inconsistencies. A maximum of two inconsistencies are allowed in a complete chain. If more than two are allowed, it is marked as "overall order abnormality". For example, in a circulation link with 7 nodes, nodes 4 and 6 are both recorded as directional offsets and node numbers are out of order. If more than two inconsistencies are identified, the order integrity risk record is triggered. Finally, the correspondence of the three parameters is checked for each node. If any two of the three parameters change in different directions in adjacent nodes, such as the certificate number remaining unchanged while the identity number is reversed, or the scratch direction changes abruptly, the node is recorded as a "weak corresponding node". The frequency of occurrence is then counted to form a node order consistency performance table.

[0040] S513: Based on the consistency of node order, judge the parameter arrangement performance, filter nodes that are inconsistent or change in order during order comparison, organize the number and order difference characteristics of each node, and obtain the link parameter mapping characteristics. The stability of each node's parameter arrangement is assessed one by one, categorized into three states: "completely consistent," "partially consistent," and "completely inconsistent." If a node's certificate number, identification number, and scratch direction are in the same order, it is marked as "completely consistent." If only one parameter differs, it is recorded as "partially consistent." If all three parameters are offset or their directions of change are inconsistent, it is determined as "completely inconsistent." For example, node "NID00534" has a reversed number order and a scratch angle offset of 6 degrees, but its certificate number remains consistent; according to this rule, it is determined as "partially consistent." If the next node "NID00310" differs in all three parameters, it is recorded as "completely inconsistent." This process is repeated for all nodes that are "completely inconsistent." The nodes are numbered and extracted to create a set of differing nodes. The corresponding changes are recorded, such as "NID00310: reversed numbering, certificate number skipping, scratch offset 5.6 degrees". By counting the number of nodes in the differing set, a threshold is set where if the number of differing nodes in a link exceeds 30% of the total, the link is marked as a structurally disordered link. In a certain actual link, a total of 9 nodes were recorded, of which 3 nodes were marked as "completely inconsistent", accounting for 33.3%, triggering the structural disorder judgment. All node numbers exhibiting abnormal arrangement behaviors such as order changes, field skipping, and directional abrupt changes are compiled into a feature dataset and summarized as link parameter mapping features.

[0041] A big data-based anti-counterfeiting and traceability system for jewelry and handicrafts, comprising: The micro-feature coding module is based on the manufacturing process of jewelry crafts. It analyzes the surface scratch direction data, obtains the scratch length and natural inclusion distribution in a standard order, and then combines the optical reflection intensity parameters. It collects micron-level concavity and convexity data through a surface measuring instrument, organizes all information into a standardized text, processes it with hash encoding, and associates it one-to-one with the finished product certificate number to obtain a unique identifier for the microstructure. The node attribute comparison module compares the weight and size of the physical objects at the logistics nodes based on the unique identifier of the microstructure. It compares the current parameters with the data of the previous node according to the standard order, analyzes the magnitude of changes in various parameters, determines the intersection of the set of the unique identifier of the current outer packaging and the identifier of the previous node, identifies the key nodes of parameter change magnitude, and obtains the node parameter offset signal. The recycling consistency determination module filters the microstructure code sequence registered by recycling nodes based on node parameter offset signals, reads the unique mark of the current outer packaging, compares the node identity number with the historical record by manual means, judges whether the time interval is abnormal based on the handover time data, and evaluates the node consistency by combining the identity number and time series analysis to obtain the node time series abnormal characteristics. The voucher anomaly detection module determines the correspondence between the recovered voucher number and the finished product certificate number based on the node time sequence anomaly characteristics. It uses field position matching for all voucher numbers and certificate numbers to identify abnormal numbers that cannot be matched one by one, and performs interactive verification in combination with node deviation data to obtain voucher mapping anomaly indicators. The link feature extraction module adjusts the finished product certificate number and node identity number in the entire process based on the abnormal index of the certificate mapping, retrieves the surface scratch direction in the factory data, establishes a three-parameter comparison sequence according to the order of factory, circulation and recycling, calculates the consistency of the parameter order of each node in the entire link, and identifies the case of order discrepancy, so as to obtain the link parameter mapping feature.

[0042] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A method for anti-counterfeiting and traceability of jewelry and handicrafts based on big data, characterized in that, Includes the following steps: S1: Based on the manufacturing process of jewelry and craft products, collect data on the direction and length of scratches on the surface of jewelry and craft products, distribution of natural inclusions, optical reflection intensity and micron-level unevenness data, integrate the information into text and hash code it, and then bind it with the finished product certificate number to obtain a unique identifier of the microstructure. S2: Based on the unique identifier of the microstructure, compare the weight, size and unique mark of the outer packaging of the logistics node, analyze the changes in the corresponding data of the previous node, identify the key nodes of the change range, and obtain the node parameter offset signal. S3: Based on the node parameter offset signal, filter the microstructure code sequence and the unique mark of the outer packaging of the recycled nodes, manually compare the node identification number with the historical record, and judge the node consistency through handover time data and time series analysis to obtain the node time series abnormal characteristics; S4: Based on the node time-series anomaly characteristics, determine the correspondence between the recycling voucher number and the finished product certificate number, match the voucher number and certificate number, identify the abnormal number, and integrate the node deviation data for interactive verification to obtain the voucher mapping anomaly index.

2. The method for anti-counterfeiting and traceability of jewelry and handicrafts based on big data according to claim 1, characterized in that, The unique identifier of the microstructure includes a feature summary, an identifier index, and a data aggregation label; the node parameter offset signal includes a state discrimination code, a flow change item, and node marker data; the node temporal anomaly feature includes anomaly type, temporal identification code, and node comparison label; and the voucher mapping anomaly index includes a verification result item, a certificate anomaly label, and a voucher comparison identifier.

3. The method for anti-counterfeiting and traceability of jewelry and handicrafts based on big data according to claim 1, characterized in that, The specific steps for obtaining the unique identifier of the microstructure are as follows: S111: Based on the manufacturing process of jewelry and handicrafts, the data on the direction of scratches collected are analyzed. Combined with the scratch trajectory and natural inclusion distribution information under the standard sequence, all features in the measurement area are aggregated and analyzed to screen the morphological parameters with correlation and obtain the scratch inclusion combination factor. S112: Based on the scratch inclusion combination factor, compare it with the collected optical reflection intensity data, analyze the surface micron-level concavity and convexity morphology of each spatial region, calculate the joint distribution of each parameter, judge the rationality of the encoding order, and obtain the structural feature mapping segment. S113: Based on the structural feature mapping fragment, analyze its structural rules, select the mapping content that establishes a one-to-one relationship with the finished product certificate number, judge the uniqueness binding effect, and adjust the data mapping structure between the text and the number to obtain the unique identifier of the microstructure.

4. The method for anti-counterfeiting and traceability of jewelry and handicrafts based on big data according to claim 1, characterized in that, The specific steps for obtaining the node parameter offset signal are as follows: S211: Based on the unique identifier of the microstructure, analyze the physical weight and size of the jewelry and handicrafts at the current logistics node, compare the same data from the previous logistics node, calculate the change range between the two, determine the weight and size change trend of each link, and obtain the node physical parameter change sequence. S212: Based on the sequence of changes in the physical parameters of the nodes, determine the intersection of the unique identifier of the outer packaging of the current node with the identifier of the previous node, analyze the trend of the change magnitude and the matching of the sets, and obtain the set of parameter offset nodes; S213: Based on the set of parameter offset nodes, compare the weight change range, size change range and the number of intersections of the outer packaging markings of each node, calculate the parameter offset degree, sort all nodes, and obtain the node parameter offset signal.

5. The method for anti-counterfeiting and traceability of jewelry and handicrafts based on big data according to claim 1, characterized in that, The specific steps for obtaining the node temporal anomaly features are as follows: S311: Based on the node parameter offset signal, analyze the microstructure code sequence of the recycling node, compare the current outer packaging unique mark and node identity number with the historical node identity number item by item, determine whether the node identity number has been registered in the historical path, filter out the node identity numbers that have not appeared, and obtain a set of non-duplicate node numbers. S312: Based on the set of non-repeating node numbers, analyze the handover order of adjacent nodes, compare the time offset of the node identity sequence with the time difference of the packaging mark, adjust the time association method, optimize the coupling relationship, obtain the average temporal coupling difference, and then determine the coupling difference category to obtain the time offset identification signal. S313: Based on the time offset identification signal, determine its distribution in the node sequence, analyze the order of the corresponding nodes, compare the identity number and registration order of adjacent nodes, filter out nodes with reversed or skipped order, optimize the time sequence discrimination criteria, and obtain node time sequence abnormal characteristics.

6. The method for anti-counterfeiting and traceability of jewelry and handicrafts based on big data according to claim 1, characterized in that, The specific steps for obtaining the abnormal voucher mapping indicator are as follows: S411: Based on the node timing anomaly characteristics, analyze the field arrangement of the recycling voucher number and the finished product certificate number, compare the order and content of each group of number characters one by one, determine if there are inconsistencies in field order or character position, and obtain the field mapping deviation combination. S412: Based on the field mapping deviation combination, combined with the node identity number, time sequence identification code and handover time, analyze the correlation change between the number and the node identity, determine the abnormal order of the number's flow time between nodes and the matching situation of the node mark, and obtain the node path difference characteristics. S413: Based on the node path difference characteristics, compare whether the field structure anomalies and node timing anomalies of each number occur synchronously, analyze their correspondence, filter out the set of numbers that cannot correspond to each other, and obtain the voucher mapping anomaly index.

7. The method for anti-counterfeiting and traceability of jewelry and handicrafts based on big data according to claim 1, characterized in that, The steps also include: S5: Based on the abnormal index of the certificate mapping, adjust the finished product certificate number and node identity number, call the surface scratch direction of the jewelry leaving the factory, establish a three-parameter comparison sequence according to the node order, judge the consistency of the link node parameter order, and obtain the link parameter mapping feature. The link parameter mapping features include sequence consistency items, node relationship indexes, and parameter comparison labels.

8. The method for anti-counterfeiting and traceability of jewelry and handicrafts based on big data according to claim 7, characterized in that, The specific steps for obtaining the link parameter mapping features are as follows: S511: Based on the aforementioned certificate mapping anomaly index, analyze the correspondence between the finished product certificate number and the node identity number, and combine the surface scratch direction in the factory data to organize the parameters of all nodes according to the factory, circulation and recycling links. The three parameters are arranged according to the order of circulation to obtain the three-parameter sequence group. S512: Based on the three-parameter sequence group, compare the certificate number, identity number and scratch direction arrangement of each node, calculate the difference in the order correspondence of the three parameters in the node, check the correlation between the parameter order one by one, and obtain the node order consistency performance. S513: Based on the consistency of the node order, judge the parameter arrangement performance, filter out nodes that are inconsistent or change in order during the order comparison, organize the number and order difference characteristics of each node, and obtain the link parameter mapping characteristics.

9. A big data-based anti-counterfeiting and traceability system for jewelry and handicrafts, characterized in that, The system is used to implement the big data-based anti-counterfeiting and traceability method for jewelry and handicrafts as described in any one of claims 1-8, and the system includes: The micro-feature coding module is based on the manufacturing process of jewelry crafts. It analyzes the surface scratch direction data, obtains the scratch length and natural inclusion distribution in a standard order, and then combines the optical reflection intensity parameters. It collects micron-level concavity and convexity data through a surface measuring instrument, organizes all information into a standardized text, processes it with hash encoding, and associates it one-to-one with the finished product certificate number to obtain a unique identifier for the microstructure. The node attribute comparison module compares the physical weight and size of the logistics node based on the unique identifier of the microstructure, compares the current parameters with the data of the previous node according to the standard order, analyzes the change range of each item, judges the intersection of the set of the current outer packaging unique mark and the previous node mark, identifies the key nodes of parameter change range, and obtains the node parameter offset signal. The recycling consistency determination module filters the microstructure code sequence registered by the recycling node based on the node parameter offset signal, reads the unique mark of the current outer packaging, compares the node identity number with the historical record by manual means, judges whether the time interval is abnormal based on the handover time data, and evaluates the node consistency by combining the identity number and time sequence analysis to obtain the node time sequence abnormal characteristics. Based on the node time-series anomaly characteristics, the voucher anomaly detection module determines the correspondence between the recovered voucher number and the finished product certificate number. It performs field position matching on all voucher numbers and certificate numbers to identify abnormal numbers that cannot be matched one-to-one, and performs interactive verification in conjunction with node deviation data to obtain voucher mapping anomaly indicators. Based on the certificate mapping anomaly index, the link feature extraction module adjusts the finished product certificate number and node identity number in the entire process, retrieves the surface scratch direction in the factory data, establishes a three-parameter comparison sequence according to the order of factory, circulation and recycling, calculates the consistency of parameter order of each node in the entire link, and identifies the case of order discrepancy, thereby obtaining the link parameter mapping features.

Images