Intelligent jewelry traceability system and method

Abstract

The present invention proposes an intelligent jewelry traceability system and method, the method comprising: collecting gemstone data; encrypting the collected gemstone data with blockchain data and storing it in layers; matching and comparing the gemstones to be traced with the collected data; integrating natural language processing technology to simplify the data query process; and using augmented reality and virtual reality technology to display gemstone data. In the face of problems such as inaccuracy in data collection, insecurity in storage, low efficiency in comparison, and unfriendly user experience, the present invention provides a comprehensive, efficient, and user-friendly intelligent jewelry traceability solution through the comprehensive application of technology.

Description

Technical Field

[0001] The present invention belongs to the field of traceability technology, and in particular relates to an intelligent jewelry traceability system and method thereof. Background Art

[0002] Globally, the jewelry industry faces severe traceability and verification issues, which are directly related to the authenticity identification and value assessment of jewelry. At present, traditional jewelry traceability technology mainly relies on physical markings (such as laser engraving) and paper certificates. Although these methods provide basic tracking functions, they have many deficiencies in data security, anti-counterfeiting effectiveness and information transparency. For example, physical markings are easily tampered with or forged, while the information on paper certificates may be tampered with over time or during transactions. In addition, existing electronic data recording systems are usually centralized, which means that data can be modified or deleted by system administrators, making it difficult to ensure the objectivity and immutability of information.

[0003] In addition, the current technology on the market also shows limitations in processing and analyzing gemstone data. Traditional methods often require manual operations when collecting various types of information on gemstones (such as spectrum, color, and internal structure information, etc.), which is not only inefficient but also prone to errors. This method is particularly insufficient in the modern jewelry industry with high throughput and high precision requirements, especially when the characteristics of gemstones are diverse and complex.

[0004] Therefore, the jewelry industry urgently needs a new technology solution that can automatically, efficiently and securely collect and manage gemstone information. Summary of the invention

[0005] The purpose of the present invention is to design an intelligent jewelry traceability system and method. By integrating high-resolution spectral imaging and high dynamic range imaging technology, the system can automatically and efficiently collect detailed information of gemstones, including but not limited to quality, specifications, color, clarity, cut, etc. In addition, the system uses machine learning technology to optimize the processing and analysis of image and spectral data, greatly improving the accuracy and efficiency of data collection, and overcoming the limitations of traditional methods in data processing.

[0006] In order to achieve the above object, the present invention provides a smart jewelry traceability method in a first aspect, the method comprising:

[0007] S101, collecting gemstone data information; wherein the gemstone data information includes at least two of quality data, specification data, color data, clarity data, cutting data, microscopic information data, spectral information data, luminescence information data, image information data, precious metal material data, imprint data, and / or microscopic information data of matching stones, wherein the method of collecting gemstone data information includes: multi-spectral data collection and / or light field data collection;

[0008] S102, encrypting and / or storing the collected gemstone data information in layers;

[0009] S103, matching and comparing the gemstone data information that needs to be traced with the stored gemstone data information, and generating a traceability query report based on the matching and comparison results.

[0010] Furthermore, the method further comprises the following steps:

[0011] The system is adaptively optimized based on the real-time collected gemstone data, including:

[0012] Define data streams, analyze data streams in real time and extract key information; wherein the data streams include jewelry appraisal results, user query behaviors and market dynamics;

[0013] Verify the system performance in real time and provide system performance evaluation results;

[0014] Collect key information of data flow and user feedback data, and use the collected data flow and user feedback data in combination with system performance evaluation results to adjust and optimize the system;

[0015] The collected gemstone data information is encrypted, including designing data access rights on the blockchain and performing intelligent management, specifically including:

[0016] Set up a smart contract and deploy it to the blockchain platform; wherein the smart contract includes role definition, permission setting and logical judgment of data access; when a user initiates an access request, identity authentication is performed. If the authentication is successful, the contract allows the current requesting user to access and automatically records the current access event; all access events are recorded on the blockchain to form an access log for query.

[0017] Furthermore, in the step S101, the collected gemstone data information is preprocessed, wherein the preprocessing step includes:

[0018] Synchronize the collected gemstone data to align the multispectral data and light field data in time and match the spatial coordinates;

[0019] The spectral data is corrected according to the system S(λ) and the environmental factor E, which is expressed as follows:

[0020]

[0021] Where, C'(λ,x,y) represents the corrected spectral data, R(λ,x,y) represents the reflectivity of wavelength λ at position (x,y); (x,y) represents the spatial coordinates;

[0022] Extract key spectral features and encode them into feature vectors And extract light field features, use light field data for 3D reconstruction, and calculate the depth D(x,y) and surface normal vector of each point

[0023] Furthermore, the step 102 specifically includes the following steps:

[0024] Encrypt the gemstone data information, as shown below:

[0025] E(D)=AES-256-Encrypt(D,k)

[0026] Where E(D) represents the encrypted data, D represents the original data to be encrypted, k is the encryption key, and Encrypt is the encryption function;

[0027] The encryption key k is protected by asymmetric encryption technology to generate a key version k' that can be securely stored, as shown below:

[0028] k′=RSA-Encrypt(k,PK)

[0029] Wherein, k' represents the key encrypted by the public key, RSA represents the asymmetric encryption algorithm, Encrypt is the encryption function; PK represents the public key;

[0030] Packing the encrypted data E(D) into a blockchain transaction; wherein each data item is recorded as a transaction, and the transaction includes a timestamp and a transaction source;

[0031] Generate a block; wherein the block includes a transaction hash value, a previous block hash value and a timestamp;

[0032] Distribute new blocks through the blockchain network and use consensus algorithms to verify the validity of new blocks;

[0033] The authorized user uses his own private key SK to decrypt the encrypted key k', and uses the decrypted key k to decrypt the encrypted data E(D).

[0034] Further, the step S103 specifically includes the following steps: S103: matching and comparing the gemstone data information to be traced with the stored pre-processed gemstone data information, and generating a traceability query report according to the matching and comparison results, specifically including: cleaning and standardizing the collected pre-processed gemstone data information, and then matching and comparing the data with the stored collected gemstone data information; wherein the cleaning and standardization processing is specifically expressed as follows:

[0035]

[0036] in, and L norm (x, y, θ, φ) represent the normalized spectral vector and light field data, respectively. L(x, y, θ, φ) represents the data captured by the light field camera. represents the spectral feature vector; represents the minimum spectral vector; The expression is the maximum spectral vector; max(L) represents the maximum light field data; min(L) represents the minimum light field data, and L(x, y, θ, φ) is the coordinate of the data captured by the light field camera;

[0037] The multi-scale waveform analysis function MSWT is applied to the standardized spectral data to calculate the structural changes of the gemstones. At the same time, the angle data information of the standardized light field data is used to capture the optical properties of the surface and interior of the gemstones through Fourier transform FT. Finally, the principal component analysis function PCA is used to reduce the data dimension by combining the standardized spectral characteristics and light field characteristics, extract key information, and obtain a comprehensive feature vector.

[0038] The multi-scale waveform analysis function MSWT is specifically expressed as follows to calculate the structural changes of gemstones:

[0039]

[0040] in, represents the feature set after waveform transformation;

[0041] The Fourier transform FT captures the optical properties of the surface and interior of the gemstone, which are specifically expressed as follows:

[0042]

[0043] in, represents the set of features after Fourier transformation;

[0044] Use the principal component analysis function PCA to reduce the data dimension, extract key information, and obtain a comprehensive feature vector It is expressed as follows:

[0045]

[0046] Among them, PCA is the principal component analysis function, represents the comprehensive feature vector;

[0047] Machine learning algorithms are applied to classify and identify gemstones based on comprehensive feature vectors, and the results and processes of classification and identification are recorded on the blockchain.

[0048] Furthermore, the method further comprises the following steps: the user performs a quick query based on spoken language, which specifically comprises the following sub-steps:

[0049] Extracting search fields based on natural language: receiving natural language query information from a user, preprocessing the query information, extracting target entity information and intent information to be queried in the natural language query information, and using the target entity information and intent information as search fields;

[0050] Constructing a mapping database: The mapping database contains graphical interface information, and specifically uses the knowledge graph to establish a mapping relationship between entities and search fields. Then, according to the search fields containing intent information and entity information, the graphical interface information stored in the database related to the target entity and intent information can be queried;

[0051] Fast search query: Use the search field containing the target entity information and intent information as the search entry to search the mapping database, obtain the query results, and present the query results to the user in the form of a graphical interface;

[0052] Mapping database optimization based on user queries and feedback: Receive and store user query feedback information, record and store user historical query information, analyze user query preferences and prioritize based on user query feedback information and historical query information using machine learning algorithm models, store the prioritized user query information in the mapping database, and when users conduct subsequent queries, for the same entity information, prioritize the recommendation of high-priority graphical interface information to the querying user;

[0053] The mapping database uses dynamic SQL generation technology to combine the user's query intention information and entity information to generate a specific SQL query statement, which is expressed as follows:

[0054] SQL=GenerateSQL(I,M)

[0055] Among them, I represents the user's query intention, M represents the knowledge graph, and GenerateSQL is a function that combines the user's query intention information and entity information to generate a specific SQL query statement.

[0056] Furthermore, the method further comprises the following steps: the graphical interface information includes displaying gemstone data using augmented reality and virtual reality technology;

[0057] It includes the following sub-steps:

[0058] Design a 3D Gemstone Display Platform M in Combination with Blockchain 3D , which is expressed as follows:

[0059]

[0060] Among them, f construct represents a function for constructing a three-dimensional model, and D(x,y) represents the coordinates of the gemstone data information to be encrypted; represents the normalized spectral vector;

[0061] Using graphics rendering technology to display three-dimensional gemstones on the M 3D The display is as follows:

[0062] R 3D =f render (M 3D ,LightModel,Texture)

[0063] where f render Represents a function for rendering a three-dimensional model, LightModel represents an optical model, and Texture represents a texture.

[0064] Furthermore, the step S103: performs data matching and comparison between the data information of the gemstones that need to be traced and the collected data information of the gemstones that have been stored, and generates a traceability query report based on the matching and comparison results, wherein the data matching and comparison at least includes: comparison of two types of information among the before and after data comparison of the composition information of the gemstones that need to be traced, the before and after data comparison of the spectral characteristic data information, the before and after data comparison of the luminescent image data information, and the before and after data comparison of the appearance information, as a basis for the judgment result of the query report of the gemstones that need to be traced.

[0065] In a second aspect of the present invention, a smart jewelry traceability system is provided, the system comprising:

[0066] A data acquisition module is used to collect gemstone data information; wherein the gemstone data information includes at least two of quality data, specification data, color data, clarity data, cutting data, microscopic information data, spectral information data, luminescence information data, image information data, precious metal material data, imprint data, and / or microscopic information data of matching stones, wherein the method of collecting gemstone data information includes: multi-spectral data collection and / or light field data collection;

[0067] An encryption storage module is used to encrypt and / or store the collected gemstone data information in layers;

[0068] The data query module is used to match and compare the gemstone data information that needs to be traced with the stored collected gemstone data information, and generate a traceability query report based on the matching and comparison results.

[0069] The third aspect of the present invention also provides a storage medium that stores non-transitory computer-readable instructions. When the non-transitory computer-readable instructions are executed by a computer, the steps in the aforementioned smart jewelry traceability method can be executed.

[0070] The beneficial technical effects of the present invention are at least as follows:

[0071] (1) The present invention integrates high-resolution spectral imaging and high dynamic range imaging technology. The system can automatically and efficiently collect detailed information about precious stones, including but not limited to quality, specifications, color, clarity, cut, etc. In addition, the system uses machine learning technology to optimize the processing and analysis of image and spectral data, greatly improving the accuracy and efficiency of data collection and overcoming the limitations of traditional methods in data processing. Using high-resolution spectral imaging and high dynamic range imaging technology, detailed information of precious stones is automatically captured. The introduction of this technology significantly improves the efficiency and accuracy of data collection and solves the problems of low efficiency and error-proneness of traditional manual collection methods. The machine learning algorithm further optimizes the processing of image and spectral data, ensuring that only high-quality data is input into the system, providing a solid data foundation for subsequent steps.

[0072] (2) The present invention uses distributed ledger technology based on blockchain. The information of each piece of jewelry is recorded in an independent, tamper-proof block. All transactions and update histories are transparent and can be audited by all authorized network participants but cannot be modified unilaterally. The application of this technology not only improves the security and immutability of data, but also provides a reliable audit trail for every jewelry transaction. By utilizing blockchain technology, each data update or query must be verified by the network, ensuring the immutability and long-term security of the data. This not only solves the security risks of traditional centralized data storage that may be modified or deleted, but also improves the transparency and trust of the entire system.

[0073] The application of deep learning and pattern recognition technology enables the system to automatically perform complex feature matching and identity verification, which significantly improves the automation level and accuracy of data comparison. This intelligent comparison method provides a reliable scientific and technological support for the authenticity identification and traceability of jewelry, and directly solves the efficiency and accuracy problems of traditional methods.

[0074] (3) The present invention introduces natural language processing (NLP) technology, allowing users to perform complex data queries through natural language. At the same time, the integration of augmented reality (AR) and virtual reality (VR) technologies provides users with an interactive and immersive data viewing experience. These innovations just solve the shortcomings of existing technologies in user experience and data interaction, and bring a new technical solution to the jewelry industry. With the help of natural language processing technology, users can simply query jewelry information through natural language. This intuitive query method greatly improves the user experience and also improves the availability and accessibility of the query system. This step solves the problem that traditional query systems are complex and user-unfriendly. By providing virtual reality technology with three-dimensional simulation and detailed views, users can view every detail of the jewelry in an immersive environment, enhancing the visualization and interactivity of the information, which not only makes the user experience richer and more attractive, but also makes the display of jewelry characteristics and traceability information more vivid and accurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0075] The present invention is further described using the accompanying drawings, but the embodiments in the accompanying drawings do not constitute any limitation to the present invention. A person skilled in the art can obtain other drawings based on the following drawings without creative work.

[0076] Figure 1 This is a flow chart of the smart jewelry traceability method of the present invention.

[0077] Figure 2 Schematic diagram of spectrum data calculation in an embodiment of the present invention Figure 1 .

[0078] Figure 3 Schematic diagram of spectrum data calculation in an embodiment of the present invention Figure 2 .

[0079] Figure 4 Schematic diagram of the three-dimensional gemstone display of the present invention Figure 1 .

[0080] Figure 5 Schematic diagram of the three-dimensional gemstone display of the present invention Figure 2 .

[0081] Figure 6 This is a framework diagram of the smart jewelry traceability system of the present invention. DETAILED DESCRIPTION

[0082] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be understood as limiting the present invention.

[0083] In one or more embodiments, Figure 1 As shown, the smart jewelry traceability method of the present invention is disclosed, and the method includes the following steps S101-S107:

[0084] S101. Collecting gemstone data information; wherein the gemstone data information includes quality data, specification data, color data, clarity data, cutting data, microscopic information data, spectral information data, luminescence information data, image information data, precious metal material data, imprint data, and / or microscopic information data of matching stones, wherein the method of collecting gemstone data information includes: multi-spectral data collection and / or light field data collection.

[0085] In one embodiment of the present invention, data acquisition is performed using a broadband light source that can cover the ultraviolet to near infrared bands, with the wavelength range set to 200nm to 2500nm. A multi-spectral scanner is also configured, with each band width of about 10nm, and the entire band is continuously scanned to capture comprehensive spectral data.

[0086] Preferably, a light field camera with a microlens array is selected to record the intensity and direction of light emitted or reflected by the gemstone from different angles, and the angular resolution and spatial resolution captured by the light field camera are set to optimize the capture of the surface details and internal structure of the gemstone.

[0087] In one embodiment of the present invention, the data acquisition method includes multispectral data acquisition and light field data acquisition; wherein the multispectral data acquisition is to define the incident light intensity I i (λ) and the received reflected light intensity I r (λ,x,y), the acquisition formula is

[0088]

[0089] Where R(λ,x,y) represents the reflectivity of wavelength λ at position (x,y).

[0090] The data captured by the light field camera is represented as L(x, y, θ, φ), where (x, y) is the spatial coordinate and (θ, φ) is the directional coordinate. The light field data provides a full view of the gemstone and can be used for further 3D reconstruction.

[0091] Preferably, ensure that the multispectral data and light field data are aligned in time and the spatial coordinates match to ensure data consistency. The spectral data is corrected, and the correction formula takes into account the system response S(λ) and environmental factors E:

[0092]

[0093] Wherein, C'(λ,x,y) represents the corrected spectral data.

[0094] In one embodiment of the present invention, it is necessary to extract key spectral features and encode them into feature vectors It is expressed as follows:

[0095]

[0096] Then extract the light field features, use the light field data for 3D reconstruction, and calculate the depth D(x,y) and surface normal vector of each point It is expressed as follows:

[0097] D(x,y)=f(L(x,y,θ,φ)) (4)

[0098] in, It can be calculated from the parallax information in the light field data and then used to reconstruct the surface details of the gemstone.

[0099] S102, encrypting and / or storing the collected gemstone data information in layers, including:

[0100] The gemstone data is encrypted using the Advanced Encryption Standard (AES-256), as shown below:

[0101] E(D)=AES-256-Encrypt(D,k) (5)

[0102] Where E(D) represents the encrypted data, D represents the original data to be encrypted, k is the encryption key, and Encrypt represents the encryption process;

[0103] As one embodiment of the present invention, AES-256 is selected because it provides military-grade security against most known attack methods.

[0104] The encryption key k is protected by asymmetric encryption technology to generate a key version k' that can be securely stored, as shown below:

[0105] k′=RSA-Encrypt(k,PK) (6)

[0106] Among them, k' represents the key encrypted by the public key, which can be safely stored or distributed with the data;

[0107] The encrypted data E(D) is packaged into a blockchain transaction, wherein each data item is recorded as a transaction, and the transaction includes a timestamp and a transaction source, as shown below:

[0108] Transaction={E(D),k′,Timestamp,Metadata} (7)

[0109] Metadata includes data sources and other related information to enhance data traceability; Transaction refers to a record of an operation on the blockchain network, which can include data transfer, currency exchange, etc. Each transaction will be recorded on the blockchain and verified through the network; Timestamp refers to the specific time point when an event occurs, which is usually used to verify the creation time of a transaction. In the blockchain, timestamps help maintain the order of transaction records and are used to verify the validity of transactions.

[0110] Generate a block; wherein the block includes a transaction hash value, a previous block hash value and a timestamp; represented as follows:

[0111] Block={Hash(Transactions),PrevHash,Timestamp} (8)

[0112] It should be noted that Hash(Transactions) is the hash value of all transactions in the block, and PrevHash is the hash value of the previous block, which ensures the integrity of the blockchain.

[0113] Distribute new blocks through the blockchain network. Use consensus algorithms (such as Proof of Work (PoW) or Proof of Stake (PoS)) to verify the validity of new blocks and ensure that there are no improper modifications;

[0114] It should be noted that the consensus process ensures that all network nodes agree on the validity of the data and the order of the blocks.

[0115] The authorized user uses his own private key SK to decrypt the encrypted key k', and uses the decrypted key k to decrypt the encrypted data E(D), as shown below:

[0116] k=RSA-Decrypt(k′,SK) (9)

[0117] D=AES-256-Decrypt(E(D),k) (10)

[0118] It should be noted that this step ensures that only users holding the appropriate keys can access sensitive data. RSA represents an asymmetric encryption algorithm that uses a pair of keys, namely a public key PK and a private key SK. The public key is used to encrypt data, and the private key is used to decrypt it.

[0119] S103, matching and comparing the data information of the gemstones to be traced with the stored data information of the collected gemstones, and generating a traceability query report according to the matching and comparison results, matching and comparing the data information of the gemstones to be traced with the stored data information of the collected gemstones, and generating a traceability query report according to the matching and comparison results, wherein the data matching and comparison at least includes: comparing the composition information of the gemstones to be traced before and after, comparing the spectral characteristic data information before and after, comparing the luminescent image data information before and after, and comparing the appearance information before and after, as the basis for the judgment result of the query report of the gemstones to be traced, including:

[0120] The collected pre-processed gemstone data information is cleaned and standardized, and then matched with the stored collected gemstone data information; wherein the cleaning and standardization processing is specifically represented as follows:

[0121]

[0122] in, and L norm (x, y, θ, φ) represent the normalized spectral vector and light field data, respectively. L(x, y, θ, φ) represents the data captured by the light field camera. represents the spectral feature vector;

[0123] Among them, this step ensures that the spectral data and light field data are aligned in spatial and temporal dimensions, so that they can be accurately associated with the same position and characteristics of the same gemstone.

[0124] like Figure 2-3 As shown, the multi-scale waveform analysis function MSWT is applied to the standardized spectral data to calculate the structural changes of the gemstone, and the angle data information of the standardized light field data is used to capture the optical properties of the surface and interior of the gemstone through Fourier transform FT. Finally, the principal component analysis function PCA is used to reduce the data dimension by combining the standardized spectral characteristics and light field characteristics, extract key information, and obtain a comprehensive feature vector

[0125]

[0126] in, It represents the feature set after waveform transformation, which can reveal the subtle structural differences of gemstones;

[0127] The Fourier Transform (FT) is expressed as follows:

[0128]

[0129] in, Represents a set of features after Fourier transformation, which helps reveal the optical response of light field data under different lighting and observation angles;

[0130] The principal component analysis (PCA) is expressed as follows:

[0131]

[0132] Among them, PCA stands for principal component analysis, represents the comprehensive feature vector;

[0133] It should be noted that PCA can effectively integrate data from different physical measurements to ensure that the final feature set contains the most representative and discriminative information.

[0134] Applying machine learning algorithms such as Random Forest (RF) or Deep Neural Networks (DNNs), expressed as follows:

[0135]

[0136] Among them, Classification represents the classification result. These machine learning models can process high-dimensional data and learn the classification patterns of gemstones from complex feature relationships.

[0137] Gemstones are classified and identified based on comprehensive feature vectors, and the results and process of classification and identification are recorded on the blockchain.

[0138] The results of identification and classification are recorded on the blockchain, using the aforementioned blockchain entry technology to ensure the traceability and non-tamperability of each operation. This step increases the trust and transparency of the entire system:

[0139]

[0140] Among them, Blockchain-Entry means recording the single event of identification and classification results on the blockchain. At the same time, the identification results and their related data are recorded on the blockchain to ensure the transparency of the authentication process and the immutability of the results.

[0141]

[0142] Blockchain-Record means recording the recognition results and related data on the blockchain. In this way, the detailed information of each authentication is stored securely and irreversibly for future query and verification.

[0143] This step overcomes the limitations of a single data source by innovatively fusing spectral and light field data, significantly improving the accuracy and reliability of gemstone authentication. At the same time, the integration of advanced machine learning and blockchain technologies ensures the efficiency of the authentication process and the security of the data.

[0144] S104, extracting search fields based on natural language: receiving natural language query information of the user, preprocessing the query information, extracting target entity information and intent information to be queried in the natural language query information, and using the target entity information and intent information as search fields, including:

[0145] Extracting search fields based on natural language: receiving natural language query information from a user, preprocessing the query information, extracting target entity information and intent information to be queried in the natural language query information, and using the target entity information and intent information as search fields:

[0146]

[0147] Among them, this expression means that the original query Q raw Through a cleaning process, unnecessary characters or words are removed to obtain a cleaner query text Q clean At the same time, the cleaned query text Q clean After further stemming, the word deformation is removed to simplify subsequent processing, and Q stemmed .

[0148] Then, a natural language processing model such as BERT (Bidirectional Encoder Representations from Transformers) is applied to perform deep semantic analysis on the preprocessed language to extract the key entities and intents in the query, as shown below:

[0149] (E,I)=BERTNLP(Q stemmed ) (19)

[0150] Among them, E represents the entity set, I represents the user's query intention, and BERTNLP represents the use of the BERT model for natural language processing.

[0151] Build a mapping database: The mapping database contains graphical interface information, and specifically uses the knowledge graph to establish a mapping relationship between entities and search fields. Then, based on the search fields containing intent information and entity information, the graphical interface information stored in the database related to the target entity and intent information can be queried.

[0152] The knowledge graph is used to establish a mapping between entities and database fields. By mapping the database schema to the domain ontology, it is ensured that queries of various expressions can be parsed, even when the terminology is inconsistent, as shown below:

[0153] M=MapOntologytoDB(E) (20)

[0154] Among them, formula (20) maps the entity E extracted from the query to the corresponding field M in the database.

[0155] Fast search query: Use the search field containing the target entity information and intent information as the search entry to search the mapping database, obtain the query results, and present the query results to the user in the form of a graphical interface;

[0156] Then, based on the parsed intent and entity mapping, an advanced query for the database is constructed; wherein the database uses dynamic SQL generation technology to combine the user's query intent and entity attributes to generate a specific SQL query statement, as shown below:

[0157] SQL=GenerateSQL(I,M) (21)

[0158] Where I represents the user's query intent and M represents the knowledge graph. The formula represents the generation of SQL statements for executing database queries based on the identified intent I and the mapping M.

[0159] Execute SQL queries to retrieve data from the database that meets the user's needs. Perform preliminary sorting and filtering of the results to provide the most relevant response, as shown below:

[0160] Results=ExecuteSQL(SQL) (22)

[0161] Mapping database optimization based on user queries and feedback: receive and store user query feedback information, record and store user historical query information, analyze user query preferences and prioritize them based on user query feedback information and historical query information using machine learning algorithm models, store the sorted user query information in the mapping database, and when the user makes subsequent queries, for the same entity information, prioritize high-priority graphical interface information to the querying user.

[0162] Combined with machine learning algorithms, query results are sorted and optimized based on user historical behavior and preferences, as shown below:

[0163] RankedResults=MLOptimize(Results,UserProfile) (23)

[0164] Specifically, the current formula uses a machine learning algorithm to optimize the query results and performs personalized sorting of the results according to the user's profile UserProfile UserProfile. RankedResults is the sorted result.

[0165] The query results are presented to the user in a graphical interface and a feedback function is designed to provide feedback to the user. The natural language processing model parameters are adjusted and the query mapping logic is improved based on the user feedback, as shown below:

[0166] Feedback=CaptureUserFeedback(RankedResults) (24)

[0167] Update_System(Feedback)

[0168] Among them, the formula captures user feedback on the ranking results, which can be used to further optimize the system's performance and user experience.

[0169] S105, the graphical interface information includes using augmented reality and virtual reality technology to display gemstone data, including:

[0170] Design a 3D Gemstone Display Platform M in Combination with Blockchain 3D , the platform can display the gemstone's origin, historical processing and certification information, as shown below:

[0171]

[0172] Among them, f construct Represents a function for constructing a three-dimensional model;

[0173] like Figure 4-5 , using graphics rendering technology to display the three-dimensional gemstones on the M 3D The display includes the internal features of the gemstone, such as inclusions, cracks, etc., as well as a dynamic demonstration of its optical properties, as shown below:

[0174] R 3D =f render (M 3D ,LightModel,Texture)

[0175] where f renderRepresents a function for rendering a three-dimensional model, LightModel represents an optical model, and Texture represents a texture.

[0176] S106, adaptively optimizing the system according to the real-time collected gemstone data, including:

[0177] Define data input streams, including new jewelry appraisal results, user query behaviors, and market dynamics. Use a data stream management system (such as Apache Kafka) to process and distribute these real-time data.

[0178] DataStream in =f capture (new data) (27)

[0179] Among them, f capture Represents a data capture function, used to collect data from various sources; DataStream in Represents data input stream; new data represents new data;

[0180] Then apply a stream data processing framework (such as Apache Flink) to analyze and extract key information in real time, ensuring that the system can respond immediately to changes in market and user needs.

[0181] DataStream processed =f process (DataStream in ) (28)

[0182] f process It is a real-time data processing function responsible for analyzing and converting input data; DataStream processed Indicates the result of data stream processing.

[0183] At the same time, the collected data and user feedback are used to continuously adjust and optimize the machine learning model. For example, the classification or prediction model is regularly updated using an online learning algorithm (such as online stochastic gradient descent).

[0184] Model updated =f learn (Model current ,DataStream processed ) (29)

[0185] f learn Represents an online learning function that updates the current model based on the processed data stream; Model updated Represents the updated model, Model current Represents the current model.

[0186] Validate the performance of the system using model evaluation techniques, such as cross-validation and A / B testing, to verify the performance of the new model and decide whether to deploy it as a production model.

[0187] Performance = f evaluate (Model updated ) (30)

[0188] f evaluate Used to evaluate the performance of the model after the update to ensure that the model update brings positive improvements; Performanc represents performance;

[0189] Design a user feedback mechanism to allow users to evaluate query results and system performance. The feedback will be used as input to further optimize the system:

[0190] f feedbackcapture (UserFeedback)

[0191] Among them, f feedbackcapture A function to capture feedback, responsible for collecting feedback information from users;

[0192] Based on user feedback and system performance evaluation results, system configuration and parameter adjustments are made to improve user satisfaction and overall system performance.

[0193] f adjust (Systemsettings,Feedback,Performance) (32)

[0194] f adjust It is a system tuning function that adjusts system settings based on feedback and performance data.

[0195] S107. Design intelligent management of data access rights on the blockchain, including:

[0196] Design smart contracts and then deploy them to the blockchain platform; the smart contracts contain specific role definitions and permission settings for each role. Roles are defined based on system requirements, such as appraisers, sales staff, managers, etc. and logical judgments on data access;

[0197] Contract define = setupRoles(R,P) (33)

[0198] Among them, R represents the role definition, P represents the permission list of each role; Contract define Represents the defined smart contract; setupRoles represents the function that creates a mapping of roles and permissions within the smart contract.

[0199] As an embodiment of the present invention, the smart contract includes a logical judgment of data access, and determines whether to allow access based on the role of the requester and the type of data requested. Where D represents the data category, such as raw data, processed data, reports, etc. Deploying the smart contract to a blockchain platform, such as Ethereum, ensures that all data access requests are managed through the contract.

[0200] deploy(Contract AccessControl ) (34)

[0201] Among them, the current formula represents the deployment of smart contracts to the blockchain

[0202] Users initiate access requests, including their role and the type of data they need. Smart contracts automatically process these requests and verify the user's permissions.

[0203] RequestHandle=processRequest(UserRole,DataType) (35)

[0204] When a user initiates an access request, if the verification is successful, the contract allows access and automatically records the current event, as shown below:

[0205] logAccessEvent(User_ID,Data_Type,AccessDecision,Timestamp) (36)

[0206] Preferably, AccessDecision can be 'Granted' or 'Denied'.

[0207] Among them, logAccessEvent represents a function that records specific events during the contract execution process, especially access control events.

[0208] As an embodiment of the present invention, all access events are recorded on the blockchain. These logs are public and cannot be tampered with, ensuring the transparency and security of the system.

[0209] BlockchainLog=record(AccessLog) (37)

[0210] At the same time, an audit module is provided so that authorized auditors can query access logs and verify the compliance of system use.

[0211] AuditResults=performAudit(AuditQuery) (38)

[0212] AuditQuery includes requests for a specific time period, a specific user, etc.

[0213] In one or more embodiments, Figure 6 As shown, the intelligent jewelry traceability system of the present invention is disclosed, and the system includes:

[0214] The data acquisition module 101 is used to collect gemstone data; wherein, the methods of collecting gemstone data include: multi-spectral data acquisition and light field data acquisition, and the specific contents collected include: quality, specification, color, clarity, cut, microscopic information, spectral information, luminescence information, image information, and other information (precious metal material, mark, microscopic information and internal information of the matching stone, etc.);

[0215] The encryption and storage module 102 is used to encrypt the collected gemstone data and / or store them in layers to facilitate retrieval and prevent information leakage;

[0216] The data storage and comparison module 103 is used to perform data matching and comparison query on the gemstones to be traced and the collected data. Specifically, the gemstones to be traced are matched and compared according to the information collected. Specifically, image overlap analysis technology and other means can be used to restore them in two dimensions, and a traceability analysis report is generated for query;

[0217] Data query module 104, used to integrate natural language processing technology, simplify the data query process, and facilitate the implementation of query time nodes and traceability analysis reports;

[0218] The gemstone data display module 105 is used to display gemstone data using augmented reality and virtual reality technology.

[0219] In a specific embodiment of the present invention, the specific traceability process is: the data acquisition module 101 completes the collection of gemstone information and the secondary collection of sample gemstone information, and the encryption and storage module 102 completes the storage of various types of collected data; the data storage and comparison module 103 completes the target gemstone information index and gemstone information comparison, and generates a traceability analysis report by comparing the information matching degree, and further queries the various matching analysis contents of the traceability analysis report through the data query module 104. If the target matching analysis contents of the traceability analysis report have a matching degree ≥ 99%, it means that the match is successful, and a diagnostic conclusion can be drawn: it matches, otherwise it does not match.

[0220] Since natural or synthetic gemstones have different chemical compositions, physical properties and microscopic appearances, the traceability method and system provided by the present invention, with the help of modern testing technology, can realize information restoration and comparison of gemstones under two-dimensional information through the collection and subsequent comparison of component information of gemstones, the early collection and subsequent comparison of spectral characteristics, the early collection and subsequent comparison of luminescent image comparison, the early measurement and subsequent comparison of appearance information, etc., thereby realizing mutual verification to achieve uniqueness and identifiability, and achieving the purpose of accurate traceability.

[0221] The present invention provides a storage medium storing non-transitory computer-readable instructions. When the non-transitory computer-readable instructions are executed by a computer, the steps in the aforementioned smart jewelry traceability method can be executed.

[0222] The above disclosures are only some preferred embodiments of the present invention, which certainly cannot be used to limit the scope of rights of the present invention. Ordinary technicians in this field can understand that all or part of the processes of the above embodiments and equivalent changes made according to the claims of the present invention are still within the scope of the invention.

Claims

1. Smart jewelry traceability method, characterized in that: The method comprises: S101, collecting gemstone data information; wherein the gemstone data information includes at least two of quality data, specification data, color data, clarity data, cutting data, microscopic information data, spectral information data, luminescence information data, image information data, precious metal material data, imprint data and microscopic information data of matching stones, wherein the method of collecting gemstone data information includes: multi-spectral data collection and / or light field data collection; S102, encrypting and / or storing the collected gemstone data information in layers; S103, matching and comparing the gemstone data information to be traced with the stored gemstone data information, and generating a traceability query report according to the matching and comparison results; Wherein, in the step S101, the collected gemstone data information is further preprocessed, wherein the preprocessing step includes: Synchronize the collected gemstone data to align the multispectral data and light field data in time and match the spatial coordinates; The spectral data is corrected according to the system S(λ) and the environmental factor E, which is expressed as follows: Where, C'(λ,x,y) represents the corrected spectral data, R(λ,x,y) represents the reflectivity of wavelength λ at position (x,y); (x,y) represents the spatial coordinates; Extract key spectral features and encode them into feature vectors And extract light field features, use light field data for 3D reconstruction, and calculate the depth D(x,y) and surface normal vector of each point Wherein, the step S103 specifically includes the following steps: S103: data matching and comparing the gemstone data information to be traced with the stored pre-processed gemstone collection data information, and generating a traceability query report according to the matching and comparison results, specifically including: cleaning and standardizing the collected pre-processed gemstone data information, and then matching and comparing the data with the stored collected gemstone data information; wherein, the cleaning and standardization processing is specifically expressed as follows: in, and L norm (x, y, θ, φ) represent the normalized spectral vector and light field data, respectively. L(x, y, θ, φ) represents the data captured by the light field camera. represents the spectral feature vector; represents the minimum spectral vector; The expression is the maximum spectral vector; max(L) represents the maximum light field data; min(L) represents the minimum light field data; The multi-scale waveform analysis function MSWT is applied to the standardized spectral data to calculate the structural changes of the gemstones. At the same time, the angle data information of the standardized light field data is used to capture the optical properties of the surface and interior of the gemstones through Fourier transform FT. Finally, the principal component analysis function PCA is used to reduce the data dimension by combining the standardized spectral characteristics and light field characteristics, extract key information, and obtain a comprehensive feature vector. The multi-scale waveform analysis function MSWT is specifically expressed as follows to calculate the structural changes of gemstones: in, represents the feature set after waveform transformation; The Fourier transform FT captures the optical properties of the surface and interior of the gemstone, which are specifically expressed as follows: in, represents the set of features after Fourier transformation; Use principal component analysis function PCA to reduce data dimension, extract key information, and obtain comprehensive feature vector It is expressed as follows: Among them, PCA is the principal component analysis function, represents the comprehensive feature vector; Apply machine learning algorithms to classify and identify gemstones based on comprehensive feature vectors, and record the results and process of classification and identification on the blockchain; The smart jewelry traceability method further includes the following steps: the user performs a quick query based on spoken language, specifically including the following sub-steps: Extracting search fields based on natural language: receiving natural language query information from a user, preprocessing the query information, extracting target entity information and intent information to be queried in the natural language query information, and using the target entity information and intent information as search fields; Constructing a mapping database: The mapping database contains graphical interface information, and specifically uses the knowledge graph to establish a mapping relationship between entities and search fields. Then, according to the search fields containing intent information and entity information, the graphical interface information stored in the database related to the target entity and intent information can be queried; Fast search query: Use the search field containing the target entity information and intent information as the search entry to search the mapping database, obtain the query results, and present the query results to the user in the form of a graphical interface; Mapping database optimization based on user queries and feedback: Receive and store user query feedback information, record and store user historical query information, analyze user query preferences and prioritize based on user query feedback information and historical query information using machine learning algorithm models, store the prioritized user query information in the mapping database, and when users conduct subsequent queries, for the same entity information, prioritize the recommendation of high-priority graphical interface information to the querying user; The mapping database uses dynamic SQL generation technology to combine the user's query intention information and entity information to generate a specific SQL query statement, which is expressed as follows: SQL=GenerateSQL(I,M) Where I represents the user's query intention, M represents the knowledge graph, and GenerateSQL is a function that combines the user's query intention information and entity information to generate a specific SQL query statement; Wherein, the graphical interface information includes displaying gemstone data using augmented reality and virtual reality technology; It includes the following sub-steps: Design a 3D Gemstone Display Platform M in Combination with Blockchain 3D , which is expressed as follows: Among them, f construct represents a function for constructing a three-dimensional model, and D(x,y) represents the coordinates of the gemstone data information to be encrypted; represents the normalized spectral vector; Using graphics rendering technology to display three-dimensional gemstones on the M 3D The display is as follows: R 3D =f render (M 3D ,LightModel,Texture) where f render Represents a function for rendering a three-dimensional model, LightModel represents an optical model, and Texture represents a texture.

2. The smart jewelry traceability method according to claim 1, characterized in that: The method further comprises the following steps: The system is adaptively optimized based on the real-time collected gemstone data, including: Define data streams, analyze data streams in real time and extract key information; wherein the data streams include jewelry appraisal results, user query behaviors and market dynamics; Verify the system performance in real time and provide system performance evaluation results; Collect key information of data flow and user feedback data, and use the collected data flow and user feedback data in combination with system performance evaluation results to adjust and optimize the system; The collected gemstone data information is encrypted, including designing data access rights on the blockchain and performing intelligent management, specifically including: Set up a smart contract and deploy it to the blockchain platform; wherein the smart contract includes role definition, permission setting and logical judgment of data access; when a user initiates an access request, identity authentication is performed. If the authentication is successful, the contract allows the current requesting user to access and automatically records the current access event; all access events are recorded on the blockchain to form an access log for query.

3. The smart jewelry traceability method according to claim 1, characterized in that: The step 102 specifically includes the following steps: Encrypt the gemstone data information, as shown below: E(D)=AES-256-Encrypt(D,k) Where E(D) represents the encrypted data, D represents the original data to be encrypted, k is the encryption key, and Encrypt is the encryption function; The encryption key k is protected by asymmetric encryption technology to generate a key version k' that can be securely stored, as shown below: k′=RSA-Encrypt(k,PK) Wherein, k' represents the key encrypted by the public key, RSA represents the asymmetric encryption algorithm, Encrypt is the encryption function; PK represents the public key; Packing the encrypted data E(D) into a blockchain transaction; wherein each data item is recorded as a transaction, and the transaction includes a timestamp and a transaction source; Generate a block; wherein the block includes a transaction hash value, a previous block hash value and a timestamp; Distribute new blocks through the blockchain network and use consensus algorithms to verify the validity of new blocks; The authorized user uses his own private key SK to decrypt the encrypted key k', and uses the decrypted key k to decrypt the encrypted data E(D).

4. The smart jewelry traceability method according to claim 1, characterized in that: The step S103: performing data matching and comparison between the data information of the gemstones that need to be traced and the collected data information of the gemstones that have been stored, and generating a traceability query report based on the matching and comparison results, wherein the data matching and comparison at least includes: comparing two types of information among the before and after data comparison of the composition information of the gemstones that need to be traced, the before and after data comparison of the spectral characteristic data information, the before and after data comparison of the luminescent image data information, and the before and after data comparison of the appearance information, as the basis for judging the query report result of the gemstones that need to be traced.

5. Intelligent jewelry traceability system, characterized by: The system is used to execute the method according to any one of claims 1 to 4, and the system comprises: A data acquisition module is used to collect gemstone data information; wherein the gemstone data information includes at least two of quality data, specification data, color data, clarity data, cutting data, microscopic information data, spectral information data, luminescence information data, image information data, precious metal material data, imprint data and microscopic information data of matching stones, wherein the method of collecting gemstone data information includes: multi-spectral data collection and / or light field data collection; An encryption storage module is used to encrypt and / or store the collected gemstone data information in layers; The data query module is used to match and compare the gemstone data information that needs to be traced with the stored collected gemstone data information, and generate a traceability query report based on the matching and comparison results.

6. A storage medium, characterized in that: It stores non-transitory computer-readable instructions, and when the non-transitory computer-readable instructions are executed by a computer, the steps in the smart jewelry traceability method as described in any one of claims 1-4 can be executed.

Images