A blockchain and distributed storage based jewelry customization traceability system and method

Abstract

The application discloses a kind of based on blockchain and distributed storage's jewelry customization traceability system and method, it is related to jewelry processing credible traceability management technical field;The method comprises: acquisition terminal is divided into processing procedure section according to processing flow, extracts first frame evidence and tail frame evidence and records local acquisition serial number, generates process boundary abstract;Video file is uploaded asynchronously to distributed storage network to obtain content identification, and is bound to process boundary abstract to generate process video binding abstract after being written into blockchain;Adjacent process boundary relationship record is generated according to local acquisition serial number and chain confirmation time to identify adjacent process first and tail upside down;According to real process order, construct whole-process merkle root and generate permanent traceability voucher.The application can avoid that chain off-line video asynchronous upload leads to chain storage evidence sequence deviates from real processing sequence, improve the authenticity and verifiability of jewelry customization traceability archives.

Description

Technical Field

[0001] This invention relates to the field of trusted traceability management technology for jewelry processing, and in particular to a jewelry customization traceability system and method based on blockchain and distributed storage. Background Technology

[0002] With the development of e-commerce and personalized consumption, jewelry consumers' demand for customized jewelry has gradually shifted from simply purchasing finished products to participating in the design, confirming raw materials, viewing the processing, and verifying the delivery results. For high-value customized jewelry such as rings, pendants, and earrings, the processing typically involves multiple stages, including design confirmation, raw material confirmation, setting, polishing, quality inspection, and delivery confirmation. This process is lengthy and involves numerous stakeholders. To improve the transparency of the customization process, existing jewelry customization platforms usually create processing archives using images, videos, craft records, and certificate information, and combine this with blockchain technology to store key data. This allows consumers, platforms, and manufacturers to verify the authenticity of the jewelry customization process in subsequent inquiries, authentication, rights protection, or after-sales handling.

[0003] The jewelry customization process generates a large volume of data, including high-definition videos, detailed images, and design drafts. Directly writing all of this data to the blockchain would create significant on-chain storage pressure, low node synchronization efficiency, and heavy bandwidth burden on small and medium-sized processing terminals. Therefore, existing solutions typically employ a collaborative on-chain and off-chain approach, storing video files on distributed storage networks like IPFS and writing content identifiers, keyframe hashes, or evidence digests to the blockchain. However, this method still has new technical drawbacks: video files from preceding processing steps are often quite long and may be uploaded to the blockchain later than subsequent shorter videos due to local caching, segmented transmission, distributed storage processing, and on-chain confirmation delays. This results in a discrepancy between the on-chain evidence storage order and the actual processing order. Consequently, even if the on-chain hashes and off-chain video files remain unchanged, a process inversion can occur, where the last frame of a longer preceding video is anchored later, while the first frame of a shorter subsequent video is uploaded first. This affects the accuracy of the jewelry customization traceability archive in representing the actual processing process. Summary of the Invention

[0004] The purpose of this invention is to address the shortcomings of existing technologies in the on-chain and off-chain collaborative evidence preservation process. In the process of pre-processing long videos, asynchronous uploading, segmented re-uploading, and distributed storage content identifier generation delays cause the last frame evidence to be anchored on-chain later than subsequent short videos or key frame evidence. This results in an inconsistency between the on-chain evidence preservation order and the actual jewelry processing order, affecting the accurate representation of process boundary relationships in customized traceability archives. Therefore, this invention proposes a jewelry customization traceability system and method based on blockchain and distributed storage.

[0005] To address the problems existing in the prior art, the present invention adopts the following technical solution: A jewelry customization and traceability method based on blockchain and distributed storage includes: S1. The acquisition terminal divides the jewelry customization and processing video into multiple processing steps according to the processing flow. It acquires video files for each processing step, extracts the first frame evidence and the last frame evidence, and records the first frame acquisition time, the last frame acquisition time, and the local acquisition sequence number. S2. Based on the first frame evidence, the last frame evidence, the first frame acquisition time, the last frame acquisition time, the local acquisition sequence number, and the process boundary summary or fixed initial summary of the previous processing process segment, generate the process boundary summary of the current processing process segment. S3. Asynchronously upload the video file to the distributed storage network to obtain the content identifier, bind the process boundary summary with the content identifier to generate the process video binding summary, write the process video binding summary to the blockchain and record the on-chain confirmation time; S4. For adjacent processing steps, identify the inverted beginning and end based on the local collection sequence number and the on-chain confirmation time, and generate the inverted beginning and end identification result. S5. Generate records of adjacent process boundary relationships based on process boundary summaries, process video binding summaries, and first-to-last-end inversion recognition results of adjacent processing process segments; S6. Determine the actual process sequence based on the local collection sequence number, construct the full-process Merkle root based on the process boundary summary, process video binding summary and adjacent process boundary relationship record, and write the full-process Merkle root into the blockchain to generate a permanent traceability certificate.

[0006] Preferably, the acquisition terminal divides the jewelry customization processing video into multiple processing steps according to the processing flow, including: taking the process steps registered according to the processing flow of the jewelry customization order as processing steps, the process steps include at least two of the following: design confirmation, raw material confirmation, setting, grinding, polishing, quality inspection and delivery confirmation, and generating a local acquisition sequence number before uploading the video file of each processing step to the distributed storage network.

[0007] Preferably, the extraction of the first frame evidence and the last frame evidence includes: extracting the first frame evidence representing the state of the jewelry at the beginning of the processing segment from the video file corresponding to the processing segment, and extracting the last frame evidence representing the state of the jewelry at the end of the processing segment, so that the beginning and end states of the same processing segment form boundary evidence.

[0008] Preferably, generating the process boundary summary of the current processing segment includes: performing hash calculations on the first frame evidence and the last frame evidence respectively to obtain the first frame hash and the last frame hash; taking the order identifier, processing segment identifier, first frame hash, last frame hash, first frame acquisition time, last frame acquisition time, local acquisition sequence number, and the process boundary summary of the previous processing segment as input, performing hash calculations to obtain the process boundary summary of the current processing segment; when the current processing segment is the first processing segment, the fixed initial summary is used as the process boundary summary of the previous processing segment to participate in the hash calculation.

[0009] Preferably, the step of binding the process boundary summary with the content identifier to generate the process video binding summary includes: after the content identifier of the video file is returned from the distributed storage network, the content identifier and the process boundary summary of the same processing process segment are used as input to perform hash calculation to obtain the process video binding summary, so that the delayed content identifier corresponds to the generated process boundary summary.

[0010] Preferably, the step of identifying the inversion of the first and last processes based on the local acquisition sequence number and the on-chain confirmation time includes: when the local acquisition sequence number of the preceding processing process is earlier than the local acquisition sequence number of the following processing process, and the on-chain confirmation time of the preceding processing process is later than the on-chain confirmation time of the following processing process, a first and last process inversion identification result is generated indicating the existence of the inversion; otherwise, a first and last process inversion identification result is generated indicating that the inversion does not exist.

[0011] Preferably, generating the adjacent process boundary relationship record includes: performing hash calculation on the process boundary summary of two adjacent processing process segments, the process video binding summary of two adjacent processing process segments, and the corresponding first and last inverted recognition results as input to obtain the adjacent process boundary relationship record; the adjacent process boundary relationship record contains the adjacent boundary relationship jointly defined by the process boundary summary of the previous processing process segment and the process boundary summary of the next processing process segment.

[0012] Preferably, the construction of the full-process Merkle root includes: arranging each processing segment according to the local acquisition sequence number; for each processing segment, generating Merkle leaf nodes based on the process boundary summary, process video binding summary, adjacent process boundary relationship record corresponding to the previous processing segment, and adjacent process boundary relationship record corresponding to the next processing segment; constructing the full-process Merkle root based on the Merkle leaf nodes arranged according to the actual process sequence; when the first processing segment does not have an adjacent process boundary relationship record corresponding to the previous processing segment, setting the adjacent process boundary relationship record as a preset fixed empty field; when the last processing segment does not have an adjacent process boundary relationship record corresponding to the next processing segment, setting the adjacent process boundary relationship record as a preset fixed empty field, and participating the preset fixed empty field in the generation of the corresponding Merkle leaf node.

[0013] Preferably, the verification of the permanent traceability certificate includes: reading the video file in the distributed storage network according to the content identifier, and recalculating the corresponding process boundary summary, process video binding summary and adjacent process boundary relationship record according to the stored order identifier, processing process segment identifier, first frame acquisition time, last frame acquisition time, local acquisition sequence number and on-chain confirmation time; constructing a full-process Merkle root for verification based on the recalculated data; when the full-process Merkle root for verification is consistent with the full-process Merkle root in the permanent traceability certificate, it is confirmed that the video file, process boundary relationship and real process sequence have not been tampered with.

[0014] Compared with the prior art, the beneficial effects of the present invention are: 1. This invention divides the jewelry customization process into processing stages according to the processing flow, and extracts the first frame and last frame evidence for each processing stage. Simultaneously, it records the first frame acquisition time, the last frame acquisition time, and the local acquisition sequence number, ensuring that each processing stage forms verifiable process boundary data before the video file is uploaded to the distributed storage network. Furthermore, based on the first frame evidence, last frame evidence, local acquisition sequence number, and the process boundary summary of the previous processing stage, a process boundary summary for the current processing stage is generated. This process boundary summary is then bound to the content identifier returned by the distributed storage network to generate a process video binding summary. This allows even large volumes of processing videos stored using off-chain asynchronous upload to establish a stable correspondence between the on-chain summary and the actual processing stage boundaries, avoiding the risk of distorted traceability order due to off-chain video upload delays.

[0015] 2. This invention also identifies the "tail-to-tail inversion" formed when the last frame of a preceding long video is later than the first frame of a subsequent short video, by comparing the local acquisition sequence number and the on-chain confirmation time of adjacent processing steps. Based on the process boundary summary, process video binding summary, and tail-to-tail inversion identification results of adjacent processing steps, it generates a record of adjacent process boundary relationships, ensuring that inconsistencies between the on-chain confirmation order and the actual processing order are clearly recorded and not misjudged as actual process reversal. Furthermore, this invention determines the actual process order according to the local acquisition sequence number and constructs a full-process Merkle root permanent traceability certificate based on the process boundary summary, process video binding summary, and adjacent process boundary relationship record. This allows consumers, platforms, or regulators to verify not only whether the video file has been replaced, but also whether the adjacent process boundary relationships and the actual process order have been tampered with during the jewelry customization process, thereby improving the authenticity, completeness, and long-term verifiability of the jewelry customization traceability archive. Attached Figure Description

[0016] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings: Figure 1 This is a flowchart illustrating a jewelry customization traceability method based on blockchain and distributed storage, as provided in an embodiment of the present invention. Detailed Implementation

[0017] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0018] Example: Figure 1 As shown, this embodiment provides a jewelry customization traceability method based on blockchain and distributed storage. This method is applied to jewelry customization processing scenarios, and is especially suitable for scenarios where the authenticity of the jewelry processing needs to be jointly proved by high-definition video, key frame images and on-chain evidence records. The jewelry customization processing scenario includes the processing of rings, pendants, earrings, bracelets or other customized jewelry from design confirmation, raw material confirmation, setting, grinding, polishing, quality inspection to delivery confirmation. In this embodiment, the acquisition terminal is deployed at the jewelry processing site to acquire videos of jewelry customization and processing. The acquisition terminal can be a camera device installed at the processing station, a mobile terminal with camera function, or a processing recording terminal that can locally archive processing videos. The central platform server is used to receive the process records uploaded by the acquisition terminal, generate process boundary summaries, call the distributed storage network to save the video files, and write the process video binding summary into the blockchain. The distributed storage network is used to save the video files and return content identifiers. The blockchain is used to record the process video binding summary, on-chain confirmation time, and permanent traceability certificate. In this embodiment, the distributed storage network is implemented using an IPFS network, and the content identifier returned by the IPFS network is denoted as... In other embodiments, the distributed storage network may also be a distributed object storage network that can address and verify the integrity of files based on content identifiers; the blockchain may be a consortium blockchain, a platform main storage chain, or a public blockchain, and the specific chain type does not affect the process inversion identification logic of this embodiment. The core of this embodiment is that the processing sequence is not based on the completion time of uploading the video file to the distributed storage network, nor on the blockchain confirmation time. Instead, the acquisition terminal records the local acquisition sequence number before the video is uploaded, and generates a process boundary summary based on the first frame evidence, the last frame evidence, and the local acquisition sequence number. When the order of the local acquisition sequence numbers of adjacent processing segments is opposite to the order of the on-chain confirmation time, it is identified as a head-to-tail inversion caused by asynchronous upload, and the actual process sequence is fixed through the record of adjacent process boundary relationships and the Merkle root of the entire process.

[0019] In this embodiment, the processing segment refers to a continuous processing step registered according to the processing flow of a custom jewelry order; one processing segment corresponds to one video file, or to a video clip continuously collected and sealed within the same processing step; the processing segment may include at least two of the following steps: design confirmation, raw material confirmation, setting, polishing, quality inspection, and delivery confirmation. In this embodiment, the first frame evidence refers to a video frame image or verifiable data of the video frame image extracted from a video file corresponding to a certain processing stage, used to characterize the state of the jewelry at the beginning of the processing stage; the last frame evidence refers to a video frame image or verifiable data of the video frame image extracted from the same video file, used to characterize the state of the jewelry at the end of the processing stage. In this embodiment, the local acquisition sequence number refers to the incrementing sequence number generated by the acquisition terminal when the video file is locally archived at each processing stage; the local acquisition sequence number is generated before the video file is uploaded to the distributed storage network, and does not depend on the video file upload completion time, the content identifier return time, or the blockchain confirmation time. In this embodiment, the process boundary summary refers to a hash digest used to fix the first frame evidence, last frame evidence, acquisition time, local acquisition sequence number, and chain relationship between preceding and following processes for a certain processing process segment; the process video binding summary refers to a hash digest generated after binding the process boundary summary with the content identifier returned by the distributed storage network; the adjacent process boundary relationship record refers to a hash record used to fix the boundary relationship between two adjacent processing process segments and the first-to-last inverted identification result; the full-process Merkle root refers to the root hash obtained by constructing a Merkle tree after arranging each processing process segment according to the actual process sequence; In this embodiment, the permanent traceability certificate refers to a verifiable certificate generated after the entire process Merkle root is written into the blockchain; the permanent traceability certificate includes at least an order identifier, the entire process Merkle root, the certificate generation time, and a blockchain transaction identifier, which are used by consumers, platforms, or regulators to verify whether the video files, process boundary relationships, and actual process sequence in the jewelry customization file have been tampered with. In this embodiment, The hash calculation function can be SHA-256, SM3, or other cryptographic hash algorithms that can achieve data integrity verification. This indicates field concatenation. When concatenating fields, the fields are arranged in a preset field order, and each field is assigned a field identifier and field length to avoid ambiguity caused by different field combinations.

[0020] This embodiment provides a jewelry customization traceability method based on blockchain and distributed storage, which includes the following steps; S1. Divide the jewelry processing steps and extract evidence of the beginning and end boundaries; Before jewelry customization processing begins, the central platform server generates an order identifier ID based on the jewelry customization order. The order identifier ID is used to uniquely identify a jewelry customization business. The order identifier ID can be formed by the platform order number, the consumer confirmation number and the random digest generated by the platform, or it can be formed by an existing order number on the platform, as long as it can uniquely correspond to a jewelry customization business. During the jewelry customization and processing process, the data acquisition terminal divides the jewelry customization and processing video into multiple processing steps according to the processing flow, denoted as: in, This represents the i-th processing step, where i is an integer greater than or equal to 1 and less than or equal to n, and n represents the total number of processing steps included in this jewelry customization process. The processing steps are divided according to the processing process registered in the jewelry customization order. For example, for a customized diamond ring, the processing steps may include material confirmation, setting, grinding, polishing and quality inspection in sequence. For each processing step The acquisition terminal obtains the video file corresponding to this processing step. ; In this embodiment, local archiving of video files means that after the acquisition terminal finishes acquiring video of a processing step, it writes the video file corresponding to that processing step into the local storage of the acquisition terminal and generates a local acquisition record before the video file is uploaded to the distributed storage network. The local acquisition record includes an order identifier, a processing step identifier, a video file identifier, the first frame acquisition time, the last frame acquisition time, and a local acquisition sequence number.

[0021] The local acquisition sequence number is generated incrementally within the same order according to the order in which video acquisition of the processing steps ends. After the acquisition terminal generates the local acquisition sequence number, it writes the local acquisition sequence number into the local acquisition record and participates in the generation of the boundary summary of the subsequent process. Thus, the local acquisition sequence number is fixed before the video file is uploaded to the distributed storage network and will not change due to changes in the subsequent video file upload completion time, content identifier return time, or on-chain confirmation time. video file After local archiving on the acquisition terminal, the acquisition terminal extracts the video file. Extracting the first frame of evidence Evidence from the end frame Among them, the first frame of evidence This represents the jewelry's state at the start of the i-th processing step, and the end frame evidence. This represents the state of the jewelry at the end of the i-th processing step; In this embodiment, the first frame evidence and the last frame evidence are extracted according to a determined frame numbering rule; the acquisition terminal extracts the video file... The frame sequence is analyzed sequentially, and the first decodeable video frame arranged by frame number is taken as the first frame evidence. The last decodeable video frame, arranged by frame number, is used as the tail frame evidence. ; When video file If the first video frame is undecodeable, the acquisition terminal searches for the first decodeable video frame in ascending order of frame number and uses it as the first frame evidence. When video file When the last video frame is undecodeable, the acquisition terminal searches for the first decodeable video frame in descending order of frame number and uses it as evidence of the last frame. ; The same frame numbering rules are used to extract the first frame evidence and the last frame evidence in both the generation and verification stages; the decodeable video frame refers to a video frame that can be parsed into image frame data by a video decoder and can be used as input for hash calculation; the extraction of the first frame evidence and the last frame evidence does not require image clarity, manual judgment results or annotations by processing personnel as necessary conditions; The acquisition terminal synchronously records the acquisition time of the first frame. Tail frame acquisition time and local collection serial number First frame acquisition time For the first frame of evidence Corresponding video capture time, last frame capture time Evidence from the end frame Corresponding video capture time; local capture sequence number It is generated by the acquisition terminal in ascending order of the video files locally archived according to the processing steps.

[0022] In this embodiment, the acquisition terminal is in the video file Generate local collection sequence number before uploading to distributed storage network Therefore, when subsequent video files experience differences in upload completion time due to variations in file size, network conditions, or distributed storage network processing time, the local acquisition sequence number... It can still reflect the actual data collection sequence of the processing steps on the data collection terminal side.

[0023] S2. Generate process boundary summaries and fix the actual data collection order; The central platform server or data acquisition terminal handles each processing step. Calculate the hash of the first frame and the hash of the last frame separately; The hash of the first frame is: The tail frame hash is: in, This represents the hash of the first frame of the i-th processing step. This represents the hash of the last frame of the i-th processing step. The system generates a process boundary summary for the i-th processing step. For the first processing step Since there is no previous processing step, the system will fix the initial summary. The process boundary summary of the previous processing step is used in the calculation; the initial summary is fixed. It can be a hash-generated digest of a pre-defined fixed byte string, or an initial order digest generated and registered by the platform when the order is created; fixed initial digest. Maintain consistency throughout the generation and verification process of the same order; For the i-th processing step, the system generates the process boundary summary as follows: in, This indicates a custom jewelry order. Indicates the identification of processing steps. Indicates the hash of the first frame. Indicates the tail frame hash. Indicates the time of the first frame acquisition. Indicates the time of the last frame acquisition. This indicates the local collection sequence number. This indicates a summary of the process boundary of the previous processing step or a fixed initial summary; Based on the above calculation method, the process boundary summary of the current processing segment is obtained. It is related not only to the first and last frame evidence of the current processing step, but also to the process boundary summary of the previous processing step. Thus, each processing step forms a chain boundary relationship according to the local acquisition sequence number. If the first frame evidence, last frame evidence, acquisition time, or local acquisition sequence number of any processing step is changed, the process boundary summary of that processing step will change, causing the process boundary summary chain of subsequent processing steps to fail verification.

[0024] S3. Asynchronously upload process videos and generate on-chain anchor records; Summary of process boundaries Then, the system will save the video file. Asynchronous upload to a distributed storage network; in this embodiment, the distributed storage network is an IPFS network; the IPFS network is based on the video file. The content generates corresponding content identifiers. Content identifier Used to locate and verify video files Contents within a distributed storage network; In this embodiment, content identifier Video files from distributed storage networks Returned upon first upload and used as the video file. Off-chain read index participates in process video binding summary The generation of the certificate; when verifying the permanent traceability certificate, the verification end identifies it with the already stored content. As a way to read off-chain video files The index, and identified by the proven content. Participation process video binding summary Recalculation; The verification end does not replace the already stored content identifier with the new content identifier generated after re-importing the same video file. When the verification end cannot verify the content identifier already stored. If the corresponding video file is read, or if the read video file cannot extract the first frame evidence and the last frame evidence according to the consistent frame numbering rule of the generation stage, the verification of the offline video file of this processing stage is determined to be unsuccessful. Because the file sizes of jewelry customization processing videos vary, the upload completion time for video files in different processing stages may differ. For example, the setting process requires continuous recording of gemstone positioning, prong pressing, and adjustment, resulting in longer video files; while the video files for pre-polishing confirmation or quality inspection are shorter. Shorter video files may be uploaded and obtain content identification earlier, while longer video files may obtain content identification later. Therefore, this embodiment does not use content identification. The time of acquisition is used as the basis for the processing sequence.

[0025] Obtaining content identification Then, the system will identify the content. Summary of process boundaries with the same processing step Bind the video and generate a summary of the process video binding. : in, This represents the video binding summary for the i-th processing step; due to The content identifier is generated before the video is uploaded to the distributed storage network, therefore, even if the content identifier is not generated. Even with a delayed return, the system can still pass. Delayed content identification Binding to process boundary summaries with a fixed actual acquisition order ; The system binds process videos to summaries. Write to the blockchain and record the on-chain confirmation time. On-chain confirmation time This indicates that the blockchain network confirms the inclusion of a process video binding summary. The time of the transaction or evidence record; the on-chain confirmation time. It is not used to directly determine the actual processing sequence, but rather to be used in step S4 to correlate with the locally acquired sequence number. By comparing, one can identify cases where the beginning and end are reversed; In this embodiment, a summary is bound to the process video. When writing to the blockchain, the system will include the order identifier ID and the processing stage identifier. Content Identifier Local collection serial number First frame acquisition time Tail frame acquisition time and on-chain confirmation time Summary of process video binding Related evidence storage; The associated evidence storage includes at least one of the following methods: linking the order ID and the processing segment ID... Content Identifier Local collection serial number First frame acquisition time Tail frame acquisition time and on-chain confirmation time Directly write a summary containing process video bindings Alternatively, the above fields can be written into the platform's evidence storage record, and the record identifier or record summary of the platform's evidence storage record can be bound to the process video summary. The data is written into the blockchain; the verification end recalculates the process boundary summary, process video binding summary, and adjacent process boundary relationship record based on the above-mentioned stored fields.

[0026] In this embodiment, the on-chain confirmation time Confirmation of process video binding summary for blockchain network The block timestamp; when the blockchain network does not provide a unified block timestamp, the on-chain confirmation time is the time when the central platform server receives the blockchain confirmation receipt. and the confirmation time on the chain. Summary of process video binding Related evidence storage; for each processing step within the same order, the method of obtaining the on-chain confirmation time remains consistent.

[0027] S4. Identify the inverted state of adjacent processes; The system targets any two adjacent processing steps. and Perform inverted head-and-tail identification; the system reads the previous processing step. Local collection sequence number The next processing stage Local collection sequence number Previous processing stage On-chain confirmation time and the next processing stage On-chain confirmation time ; In the actual data collection sequence, if the preceding processing step... Earlier than the next processing stage Once local sealing is complete, the following conditions must be met: If both are detected: This indicates the previous processing step. Although it is collected earlier than the next processing step in the local collection sequence. However, its process video is bound to a summary. Video binding summary of processes that occur later than the next processing stage Complete on-chain confirmation; at this point, the system determines that there is a head-to-tail inversion between the i-th processing segment and the (i+1)-th processing segment; In this embodiment, the criteria for identifying upside-down placement are: When the above conditions are met, the system generates a head-to-tail inversion recognition result indicating the existence of head-to-tail inversion: When the above conditions are not met, the system generates a result indicating that there is no inverted head-to-tail identification: This inverted order does not refer to a reversal of the actual processing flow, but rather to a reversal of the on-chain confirmation order and the local acquisition order. For custom jewelry processing, this phenomenon is usually caused by the fact that the video file of the previous processing stage is longer, and the uploading, segmentation, content identification generation, or on-chain confirmation takes longer. The video file of the subsequent processing stage is shorter, and the distributed storage and on-chain confirmation are completed first.

[0028] S5. Generate records of boundary relationships between adjacent processes; The system targets each group of adjacent processing steps. and Generate boundary relationship records between adjacent processes Record of boundary relationships between adjacent processes Used to fix the boundary relationship between the previous processing step and the next processing step, and to record whether there is a head-to-tail reversal between the two; The system will summarize the process boundaries of the previous processing step. Summary of process boundaries for the next processing stage Summary of video binding for the previous processing stage Summary of video binding for the next processing stage and the results of the inverted head and tail recognition As input, a hash calculation is performed to obtain the boundary relationship record between adjacent processes. : in, This represents a record of the adjacent process boundary relationships between the i-th processing step and the (i+1)-th processing step; because It includes the first frame evidence, the last frame evidence, the acquisition time, the local acquisition sequence number, and the process boundary summary of the previous processing segment for the i-th processing step. It contains the first frame evidence, the last frame evidence, the acquisition time, the local acquisition sequence number, and the process boundary summary of the (i+1)th processing step. The adjacent boundary relationship between the last frame evidence of the i-th processing segment and the first frame evidence of the (i+1)-th processing segment can be jointly defined. when hour, This indicates that there is a head-to-tail reversal at the adjacent boundary, meaning that the actual data collection order is the i-th processing step first and the (i+1)-th processing step second, but the on-chain confirmation order is the (i+1)-th processing step first and the i-th processing step second; when hour, This indicates that there is no inverted beginning and end at the adjacent boundary; In this embodiment, the system does not delete or overwrite the process video binding summary already written to the blockchain, nor does it modify the on-chain confirmation time. Instead, it records the process video binding summary based on the boundary relationship between adjacent processes. The difference between the on-chain confirmation order and the actual collection order is recorded; this preserves the immutability of blockchain evidence and avoids mistaking the on-chain confirmation order caused by asynchronous uploads for the actual processing order.

[0029] S6. Construct a full-process Merkle root order based on the actual process boundaries and generate permanent traceability credentials; After the jewelry customization and processing is completed, the system uses the locally collected serial number. Determine the actual sequence of processes for each stage; the system does not follow the on-chain confirmation time. The processing steps are arranged in a specific order, and they are not arranged according to the content labels. The processing steps are arranged in chronological order of return time; For the i-th processing step, the system uses a process boundary summary for that processing step. Process video binding summary Record of the boundary relationship between adjacent processes corresponding to the previous processing step. And records of the boundary relationships between adjacent processes corresponding to the next processing step. Generate Merkel leaf nodes : When the i-th processing step is the first processing step, there is no record of the adjacent process boundary relationship corresponding to the previous processing step. The system will Set it to a preset fixed empty field, and add this preset fixed empty field to the Merkel leaf node. The generation of the process; when the i-th processing step is the last processing step, there is no record of the adjacent process boundary relationship corresponding to the next processing step. The system will Set it to a preset fixed empty field, and add this preset fixed empty field to the Merkel leaf node. The generation; In this embodiment, the preset fixed empty field is a fixed byte string registered by the system in the initial order storage, or a digest obtained by hash calculation of the fixed byte string; for the same order, the boundary relationship record of the adjacent process before the first missing processing process segment and the boundary relationship record of the adjacent process after the missing last processing process segment both use the same preset fixed empty field, and the preset fixed empty field remains consistent in the permanent traceability certificate generation stage and the verification stage. The system arranges Merkel leaf nodes according to the actual process sequence: Based on the aforementioned Merkel leaf nodes, a full-process Merkel tree is constructed to obtain the full-process Merkel root. : In this embodiment, when constructing the full-process Merkle tree, the system performs pairwise hash calculations on adjacent nodes according to the Merkle tree construction rules, generating parent nodes layer by layer upwards until a unique full-process Merkle root is obtained. During the calculation of each layer of nodes, if the number of nodes in that layer is even, then the adjacent two nodes are joined by fields according to the node sorting order and a hash calculation is performed to obtain the parent node of the previous layer. If the number of nodes in a certain layer is odd, the system copies the last node in the sorting order of that layer as a replacement node, and performs a hash calculation after concatenating the fields of the last node of that layer and the replacement node to obtain the corresponding parent node; the above odd-number node replacement rule is written into the initial order certificate and remains consistent in the permanent traceability certificate generation and verification stages.

[0030] The system writes the entire Merkle root process into the blockchain, generating a permanent traceability certificate. The permanent traceability certificate includes at least the order ID, the entire Merkle root process, the certificate generation time, and the blockchain transaction identifier. Consumers, platforms, or regulators can use the permanent traceability certificate to verify whether the jewelry customization and processing file is complete and whether the process boundaries have been tampered with.

[0031] Through the above steps, this embodiment achieves the following effect: even if the video file of the previous processing step is long and causes its on-chain confirmation time to be later than that of the next processing step, the system still determines the actual processing sequence according to the local acquisition sequence number and the process boundary summary chain, and includes the inverted state of the beginning and end into the boundary relationship record of adjacent processes, and then permanently anchors it through the whole process Merkle root.

[0032] When consumers, platforms, or regulators need to verify the traceability records of customized jewelry in the future, they should follow the verification process below; The verification end reads the order identifier ID and the full-process Merkle root based on the permanent traceability certificate; the verification end reads the process video binding summary corresponding to each processing stage based on the on-chain record or the platform's evidence storage record associated with the on-chain record. Content Identifier Processing step identification First frame acquisition time Tail frame acquisition time Local collection serial number On-chain confirmation time In addition to the blockchain transaction identifier; the above fields are the process video binding summary. When writing to the blockchain, fields that have been associated with evidence storage must not be replaced by external fields that are not constrained by on-chain records or associated on-chain records. The verification end is based on the content identifier Read the corresponding video file from the distributed storage network. ; The verification end obtains the video file When extracting the first and last frame evidence, the same frame numbering rule as in the generation stage is used; that is, the verification end extracts the video file... The first decodeable video frame arranged by frame number is used as the first frame evidence. The last decodeable video frame, arranged by frame number, is used as the tail frame evidence. When the first video frame is undecodeable, search for the first decodeable video frame in ascending order of frame number; when the last video frame is undecodeable, search for the first decodeable video frame in descending order of frame number; if the verification end cannot extract the first frame evidence or the last frame evidence according to the above rules, the video file verification for this processing step is deemed to have failed.

[0033] The verification end obtains the video file The first frame evidence is extracted according to the video frame extraction rules consistent with those used in the generation stage. Evidence from the end frame Then, recalculate the hash of the first frame. End frame hash ; The verification end recalculates the process boundary summary according to the field order consistent with the generation stage. : For the first processing step, the verification stage uses the same fixed initial summary as the generation stage. The verification end then recalculates the process video binding summary. : If recalculated If the video recording of the process is not consistent with the one stored in the blockchain, it means that there is a discrepancy between the corresponding video file, content identifier, or process boundary summary, and the verification fails.

[0034] The verification end re-identifies the inverted beginning and end for adjacent processing steps; for adjacent processing steps and Verification end reads , , and When satisfied and At that time, regain Otherwise, get it again. The verification end is based on the re-obtained , , , and Recalculate the boundary relationship record between adjacent processes : The verification end is based on the recalculated , , and Regenerate Merkel leaf nodes The leaf nodes are arranged according to the local collection sequence number, and the full-process Merkle root for verification is reconstructed. : When verifying the entire Merkel process When the entire process Merkle root is consistent with the permanent traceability certificate, confirm that the video files, process boundary relationships, and actual process sequence have not been tampered with; when If there is a discrepancy with the Root, the verification will fail if at least one of the following is confirmed to be inconsistent: video file, process boundary summary, inverted head-to-tail recognition result, adjacent process boundary relationship record, or actual process order.

[0035] The following example, a custom-made diamond ring, illustrates the specific operation process of this embodiment. Consumers confirm their diamond ring orders on a jewelry customization platform. The order details include the ring setting material, main diamond specifications, setting method, ring size, and delivery requirements. The central platform server generates an order ID and records the order's processing flow. The processing flow includes five stages: raw material confirmation, setting, polishing, and quality inspection, denoted as follows: , , , and ; In the raw material confirmation process In the process, the acquisition terminal captures images of the ring holder blank and the main drill material, generating video files. Extract the first frame of evidence Evidence from the end frame Record the first frame acquisition time Tail frame acquisition time and local collection sequence number The system calculates the hash of the first frame. End frame hash and with a fixed initial summary As a process boundary summary of the previous processing step, a process boundary summary for the raw material verification process is generated. ; In the inlay process In the process, the craftsman positions, aligns, and presses the prongs onto the main diamond. Because this step requires complete recording of the gem placement, prong adjustment, and fixing, the recording device continuously captures footage for a considerable period, generating a large video file. The data acquisition terminal is stored locally. Then, extract the first frame of evidence. Evidence from the end frame Record the first frame acquisition time Tail frame acquisition time and local collection sequence number The system is based on , , , , Summary of the boundary with the previous process Generate process boundary summary for tessellation process ; During the polishing process In the middle, the craftsman polishes a portion of the set ring; because the initial polishing video is relatively short, the video file generated by the acquisition terminal... Video files smaller than the inlay process Local storage at the data acquisition terminal Then, extract the first frame of evidence. Evidence from the end frame Record the first frame acquisition time Tail frame acquisition time and local collection sequence number The system is based on , , , , Summary of the boundary with the previous process Generate process boundary summary for the grinding process ; because The file is short. Possibly before Complete the distributed storage network upload and obtain the content identifier. The system generates a video summary of the polishing process. and will Write to the blockchain and record the on-chain confirmation time. Subsequently, a long video file of the inlay process. Only after uploading and obtaining content identification. The system generates a video summary of the inlay process. and will Write to the blockchain and record the on-chain confirmation time. ; At this point, the actual collection order satisfies However, the on-chain confirmation time meets the requirements. The system uses this information to identify the inlay process. and polishing process There is a head-to-tail inversion, generate a head-to-tail inversion recognition result. The system then generates records of the boundary relationships between adjacent processes. The record shows that the last frame evidence of the inlay process and the first frame evidence of the polishing process form a real adjacent boundary, and the reversed on-chain confirmation order is caused by the delayed upload of the long inlay video, rather than the reversed processing order. During the polishing process and quality inspection process In the process, the system generates process boundary summaries in the same manner. and The corresponding video file is uploaded to the distributed storage network to obtain the content identifier. and Generate process video binding summary and It also generates boundary relationship records for adjacent processes based on the local acquisition sequence number and on-chain confirmation time of adjacent processes. and ; Once all orders are processed, the system will process them according to the locally collected sequence number. Arrange five processing steps and generate Merkel leaf nodes for each. , , , and ;for If there is no record of adjacent process boundary relationship corresponding to the previous processing step, the system sets this field to a preset fixed empty field; for If there is no record of the boundary relationship between adjacent processes corresponding to the next processing step, the system will also set this field as a preset fixed empty field; the system will then... to Construct a full-process Merkle tree, obtain the full-process Merkle root, and write the root into the blockchain to generate a permanent traceability certificate; After receiving the goods, consumers can view a permanent traceability certificate through the platform. The platform's verification end reads the Root based on the certificate, reads video files from the distributed storage network based on each content identifier, and recalculates the boundary summaries of each process, the process video binding summaries, the boundary relationship records of adjacent processes, and the full-process Merkle root verification. ;when When the Root is consistent with the permanent traceability certificate, the consumer can confirm that the video file in the diamond ring customization file has not been replaced, the boundary relationship between adjacent processes has not been tampered with, and the actual process sequence has not been incorrectly expressed by the reversed on-chain confirmation sequence caused by asynchronous upload.

[0036] This embodiment also provides a jewelry customization traceability system based on blockchain and distributed storage, which is used to execute the above method; the system includes a data acquisition terminal, a central platform server, a distributed storage network, and a blockchain network; The acquisition terminal includes a video acquisition unit, a boundary frame extraction unit, a local acquisition sequence number generation unit, and a data transmission unit; the video acquisition unit acquires video files corresponding to the processing steps. The boundary frame extraction unit extracts data from the video file. Extracting the first frame of evidence Evidence from the end frame The local acquisition sequence number generation unit generates a local acquisition sequence number before the video file is uploaded to the distributed storage network. And record the first frame acquisition time. Acquisition time of the last frame The data sending unit will transmit the video file. First frame evidence End frame evidence First frame acquisition time Tail frame acquisition time and local collection sequence number Send to the central platform server; The central platform server includes a process boundary summary generation module, a video upload module, an on-chain anchoring module, a first-to-last inverted recognition module, an adjacent process boundary relationship generation module, a Merkle root generation module, and a traceability verification module; The process boundary summary generation module is used to process the first frame of evidence. Evidence from the end frame Perform hash calculation to obtain the hash of the first frame. End frame hash And based on the order identifier ID and processing procedure segment First frame hash Tail frame hash First frame acquisition time Tail frame acquisition time Local collection serial number Generate the process boundary summary of the current processing segment from the process boundary summary of the previous processing segment or a fixed initial summary. ; The video upload module is used to upload video files. Asynchronously upload to the distributed storage network and receive the content identifier returned by the distributed storage network. The video upload module does not use the content identifier return time as a basis for sorting processing steps. The on-chain anchoring module is used to summarize the process boundaries. With content identifiers Binding generation process video binding summary And bind the process video to a summary Write to the blockchain and record the on-chain confirmation time. The on-chain anchoring module is also used to link order ID and processing segment ID. Content Identifier Local collection serial number First frame acquisition time Tail frame acquisition time and on-chain confirmation time Summary of process video binding Related evidence storage; The inverted head-and-tail recognition module is used to read the locally acquired sequence number of adjacent processing steps. and and on-chain confirmation time and ,when and The system generates an inverted head-to-tail recognition result indicating the presence of an inverted head-to-tail structure. Otherwise, generate a result indicating that there is no inverted head-to-tail recognition. ; The adjacent process boundary relationship generation module is used to generate a summary of the process boundaries between two adjacent processing process segments. and Process video binding summary and and the results of the inverted head and tail recognition Generate boundary relationship records between adjacent processes ; The Merklegen generation module is used to arrange each processing step segment according to the local acquisition sequence number, and to summarize the process boundary of each processing step segment. Process video binding summary Record of the boundary relationship between adjacent processes corresponding to the previous processing step. And records of the boundary relationships between adjacent processes corresponding to the next processing step. Generate Merkel leaf nodes The Merkel root is constructed based on all Merkel leaf nodes. The Merkel root generation module is also used to write the full-process Merkel root into the blockchain to generate a permanent traceability certificate. The source tracing and verification module is used to verify content identifiers. Reading video files from a distributed storage network Based on the stored order identifier, processing segment identifier, first frame acquisition time, last frame acquisition time, local acquisition sequence number, and on-chain confirmation time, the module recalculates the process boundary summary, process video binding summary, adjacent process boundary relationship record, and verification full-process Merkle root. When the verification full-process Merkle root matches the full-process Merkle root in the permanent traceability certificate, the traceability verification module confirms that the video file, process boundary relationship, and actual process sequence have not been tampered with. When the two do not match, the traceability verification module outputs a verification failure result. Distributed storage networks are used to receive and save video files. Return content identifier Blockchain networks are used to store process video binding summaries. On-chain confirmation time Full-process Merkle root and permanent traceability certificate.

[0037] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for jewelry customization traceability based on blockchain and distributed storage, characterized in that, Includes the following steps: S1. The acquisition terminal divides the jewelry customization and processing video into multiple processing steps according to the processing flow. It acquires video files for each processing step, extracts the first frame evidence and the last frame evidence, and records the first frame acquisition time, the last frame acquisition time, and the local acquisition sequence number. S2. Based on the first frame evidence, the last frame evidence, the first frame acquisition time, the last frame acquisition time, the local acquisition sequence number, and the process boundary summary or fixed initial summary of the previous processing process segment, generate the process boundary summary of the current processing process segment. S3. Asynchronously upload the video file to the distributed storage network to obtain the content identifier, bind the process boundary summary with the content identifier to generate the process video binding summary, write the process video binding summary to the blockchain and record the on-chain confirmation time; S4. For adjacent processing steps, identify the inverted beginning and end based on the local collection sequence number and the on-chain confirmation time, and generate the inverted beginning and end identification result. S5. Generate records of adjacent process boundary relationships based on process boundary summaries, process video binding summaries, and first-to-last-end inversion recognition results of adjacent processing process segments; S6. Determine the actual process sequence based on the local collection sequence number, construct the full-process Merkle root based on the process boundary summary, process video binding summary and adjacent process boundary relationship record, and write the full-process Merkle root into the blockchain to generate a permanent traceability certificate. 2.The jewelry customization traceability method based on blockchain and distributed storage according to claim 1, characterized in that, The acquisition terminal divides the jewelry customization processing video into multiple processing steps according to the processing flow, including: taking the process steps registered according to the processing flow of the jewelry customization order as processing steps, the process steps include at least two of the following: design confirmation, raw material confirmation, setting, grinding, polishing, quality inspection and delivery confirmation, and generating a local acquisition sequence number before uploading the video file of each processing step to the distributed storage network. 3.The jewelry customization traceability method based on blockchain and distributed storage according to claim 1, characterized in that, The extraction of first-frame evidence and last-frame evidence includes: extracting first-frame evidence representing the state of the jewelry at the beginning of the processing segment from the video file corresponding to the processing segment, and extracting last-frame evidence representing the state of the jewelry at the end of the processing segment, so that the beginning and end states of the same processing segment form boundary evidence.

4. The jewelry customization traceability method based on blockchain and distributed storage according to claim 1, characterized in that, The process boundary summary of the current processing segment includes: performing hash calculations on the first frame evidence and the last frame evidence respectively to obtain the first frame hash and the last frame hash; taking the order identifier, processing segment identifier, first frame hash, last frame hash, first frame acquisition time, last frame acquisition time, local acquisition sequence number and the process boundary summary of the previous processing segment as input, performing hash calculations to obtain the process boundary summary of the current processing segment; when the current processing segment is the first processing segment, the fixed initial summary is used as the process boundary summary of the previous processing segment to participate in the hash calculation.

5. The jewelry customization traceability method based on blockchain and distributed storage according to claim 1, characterized in that, The step of binding the process boundary summary with the content identifier to generate the process video binding summary includes: after the content identifier of the video file is returned from the distributed storage network, the content identifier and the process boundary summary of the same processing process segment are used as input to perform hash calculation to obtain the process video binding summary, so that the delayed content identifier is mapped to the generated process boundary summary.

6. The jewelry customization traceability method based on blockchain and distributed storage according to claim 1, characterized in that, The method of identifying the inversion of the first and last processes based on the local acquisition sequence number and the on-chain confirmation time includes: when the local acquisition sequence number of the preceding processing process is earlier than that of the following processing process, and the on-chain confirmation time of the preceding processing process is later than that of the following processing process, an inversion identification result is generated indicating the existence of the inversion; otherwise, an inversion identification result is generated indicating that the inversion does not exist.

7. The jewelry customization traceability method based on blockchain and distributed storage according to claim 1, characterized in that, The process of generating adjacent process boundary relationship records includes: performing hash calculations on the process boundary summaries of two adjacent processing process segments, the process video binding summaries of two adjacent processing process segments, and the corresponding first-to-last inverted recognition results to obtain adjacent process boundary relationship records; the adjacent process boundary relationship records contain adjacent boundary relationships jointly defined by the process boundary summaries of the previous processing process segment and the process boundary summaries of the next processing process segment.

8. The jewelry customization traceability method based on blockchain and distributed storage according to claim 1, characterized in that, The construction of the full-process Merkle root includes: arranging each processing segment according to the local acquisition sequence number; for each processing segment, generating Merkle leaf nodes based on the process boundary summary, process video binding summary, adjacent process boundary relationship record corresponding to the previous processing segment, and adjacent process boundary relationship record corresponding to the next processing segment; constructing the full-process Merkle root based on the Merkle leaf nodes arranged according to the actual process sequence; when the first processing segment does not have an adjacent process boundary relationship record corresponding to the previous processing segment, setting the adjacent process boundary relationship record as a preset fixed empty field; when the last processing segment does not have an adjacent process boundary relationship record corresponding to the next processing segment, setting the adjacent process boundary relationship record as a preset fixed empty field, and participating the preset fixed empty field in the generation of the corresponding Merkle leaf node.

9. The jewelry customization traceability method based on blockchain and distributed storage according to claim 1, characterized in that, The verification of the permanent traceability certificate includes: reading the video file in the distributed storage network according to the content identifier, and recalculating the corresponding process boundary summary, process video binding summary and adjacent process boundary relationship record according to the stored order identifier, processing process segment identifier, first frame acquisition time, last frame acquisition time, local acquisition sequence number and on-chain confirmation time; constructing the full-process Merkle root for verification based on the recalculated data; when the full-process Merkle root for verification is consistent with the full-process Merkle root in the permanent traceability certificate, it is confirmed that the video file, process boundary relationship and real process sequence have not been tampered with.

10. A jewelry customization traceability system based on blockchain and distributed storage, used to execute the jewelry customization traceability method based on blockchain and distributed storage as described in any one of claims 1-9, characterized in that, include: Data acquisition terminals, central platform servers, distributed storage networks, and blockchain networks; The acquisition terminal is used to divide the jewelry customization and processing video into multiple processing steps according to the processing flow, acquire video files for each processing step, extract the first frame evidence and the last frame evidence, and record the first frame acquisition time, the last frame acquisition time and the local acquisition sequence number. The central platform server includes a process boundary summary generation module, a video upload module, an on-chain anchoring module, a first-and-last inverted recognition module, an adjacent process boundary relationship generation module, and a Merkle root generation module; The process boundary summary generation module is used to generate the process boundary summary of the current processing process segment based on the first frame evidence, the last frame evidence, the first frame acquisition time, the last frame acquisition time, the local acquisition sequence number, and the process boundary summary or fixed initial summary of the previous processing process segment. The video upload module is used to asynchronously upload video files to the distributed storage network and obtain the content identifier returned by the distributed storage network; The on-chain anchoring module is used to bind the process boundary summary with the content identifier to generate a process video binding summary, write the process video binding summary into the blockchain network and record the on-chain confirmation time; The inverted head and tail identification module is used to identify the inverted head and tail of adjacent processing segments based on the local acquisition sequence number and the on-chain confirmation time, and generate the inverted head and tail identification result. The adjacent process boundary relationship generation module is used to generate adjacent process boundary relationship records based on the process boundary summary, process video binding summary and the first and last inverted recognition results of adjacent processing process segments; The Merkle root generation module is used to determine the actual process sequence according to the local collection sequence number, construct the full-process Merkle root based on the process boundary summary, process video binding summary and adjacent process boundary relationship record, and write the full-process Merkle root into the blockchain network to generate a permanent traceability certificate.

Images