Three-dimensional digital collection block chain evidence storage optimization method

Abstract

The invention belongs to the technical field of cross of block chains and digital collections, and particularly relates to a three-dimensional digital collection block chain evidence storage optimization method. The specific process is as follows: S1, setting a layered block chain adaptation architecture, wherein the architecture comprises an asset layer, a protocol layer and an application layer; s2, batch casting: establishing a double-array association mechanism, and associating a basic information array and a three-dimensional feature array of the collection through a unique mapping key; pre-casting verification is executed before batch casting, Merkle roots passing verification are generated, and actual casting is executed only after root Hash uplink is passed; s3, three-dimensional recording of metadata: recording technical metadata and process metadata of the casting process, and establishing three-dimensional data network association metadata for constructing dynamic association; and S4, multi-dimensional traceability verification: realizing the multi-dimensional traceability verification through on-chain and off-chain cooperative verification, timestamp hierarchical verification and an intelligent contract traceability engine.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of cross between blockchain and digital collection, and particularly relates to a three-dimensional digital collection blockchain storage optimization method. BACKGROUND

[0002] There are three core pain points in the existing three-dimensional digital collection storage: first, the general blockchain platform is not suitable for the characteristics of digital collection, and batch minting is prone to transaction congestion. For example, the patent with publication number CN202311011646.5 only realizes batch minting of two-dimensional digital collection, and does not solve the storage adaptation problem of three-dimensional model, and cannot cope with the characteristics of large capacity and high correlation of three-dimensional collection; second, the batch minting of ERC1155 standard lacks customized logic for three-dimensional models, and the minting efficiency and asset correlation are difficult to balance. For example, the patent with publication number CN201811004184.3 realizes the on-chain storage of three-dimensional models, but the minting logic does not consider the feature correlation of three-dimensional models, and does not design the associated graph record of metadata, which cannot support the correlation tracing of the same series of collection; third, the dimension of metadata record is limited and the traceability relies on single on-chain information, and the credibility verification level is shallow. Therefore, it is urgent to build an innovative technical architecture and method system. SUMMARY

[0003] The application provides a three-dimensional digital collection blockchain storage optimization method to solve the above problems, which builds a hierarchical blockchain adaptation architecture, designs ERC1155 customized minting, establishes a three-dimensional metadata system, and develops a multi-dimensional traceability mechanism, combined with on-chain and off-chain collaboration and artificial review to improve the credibility of digital collection blockchain storage.

[0004] The technical scheme of the application is as follows:

[0005] A three-dimensional digital collection blockchain storage optimization method, the specific process is as follows:

[0006] S1, a hierarchical blockchain adaptation architecture is set, the architecture includes an asset layer, a protocol layer and an application layer;

[0007] The asset layer preprocesses the three-dimensional digital collection model, the preprocessing includes lightweight processing of the three-dimensional digital collection model and generation of on-chain and off-chain mapping relationship;

[0008] The protocol layer dynamically schedules blockchain transaction processing resources and performs deep feature extraction, fusion and encryption processing on the preprocessed three-dimensional model data, stores the encrypted data in off-chain backup, and stores the data hash value and related index information on chain;

[0009] The application layer monitors the storage state, the storage state includes minting progress, metadata on-chain state, traceability link integrity and abnormal warning information;

[0010] S2, Batch Casting: Establish a dual-array association mechanism to associate the basic information array and the three-dimensional feature array of the collection through a unique mapping key; perform pre-casting verification before batch casting to generate a verified Merkle root, and only execute the actual casting after the root hash is successfully uploaded to the chain.

[0011] S3, Metadata 3D Recording: Records technical metadata and process metadata of the casting process, and establishes a dynamically related 3D data network to associate metadata;

[0012] S4, multi-dimensional traceability verification: achieved through on-chain and off-chain collaborative verification, timestamp-layered verification, and smart contract traceability engine.

[0013] Optionally, the lightweight processing described in this invention is as follows: First, wavelet decomposition is performed on the vertex coordinates and texture pixel values ​​of the 3D model, wherein the vertex coordinates are transformed using a 3D discrete wavelet transform and the texture pixels are transformed using a 2D wavelet transform; second, the 3D model is decomposed in the frequency domain into low-frequency approximation components and high-frequency detail components. The low-frequency approximation components represent the main shape and general outline of the model, while the high-frequency detail components reflect the subtle changes and texture characteristics of the model surface; the low-frequency components are retained and the high-frequency components are processed using a threshold quantization method.

[0014] When the model accuracy requirement is high, the threshold is set to 0.01-0.05; when the accuracy requirement is normal, the threshold is set to 0.05-0.1. The compression algorithm fault tolerance mechanism is set. When the ambient temperature exceeds 25℃±2℃, the temperature compensation coefficient is automatically activated, and the compression threshold correction amount is increased by 0.02 for every 1℃ deviation.

[0015] Optionally, this invention constructs an on-chain and off-chain mapping relationship between on-chain storage format and off-chain backup index through smart contracts. The on-chain storage format uses the hash value of the cultural relic digital asset as a unique identifier, while the corresponding off-chain backup index is stored in an off-chain database, and the index address and on-chain hash value are associated and recorded in the smart contract.

[0016] Optionally, the present invention deploys a 3D asset preprocessing module in the asset layer to clean, convert the format, and extract basic features of the data after lightweight processing; the cleaning operation aims to remove noise points and abnormal data generated during the data acquisition process; the format conversion unifies the 3D model formats from different sources into the glTF format; and the basic feature extraction performs preliminary analysis and extraction of the basic information of the model.

[0017] Optionally, the protocol layer of the present invention dynamically schedules blockchain transaction processing resources, specifically by: monitoring the remaining computing power of nodes in real time; when a batch casting request is triggered, allocating processing resources according to the priority coefficient calculated based on the number of collectibles and the node load, where the priority coefficient = casting quantity × 0.6 + node idle computing power × 0.4; and simultaneously, setting up a node failure emergency plan, whereby the task is automatically transferred to the node with the second highest priority coefficient when the allocated node fails.

[0018] The protocol layer deploys a 3D model processing module, which is responsible for performing deep feature extraction, fusion, and encryption operations on the preprocessed 3D model data. Deep feature extraction uses machine learning algorithms to mine the deep texture and structural features of the model. The fusion operation integrates feature data from different channels and of different types. The encryption process uses an asymmetric encryption algorithm to encrypt key data. Through the above processing, a digital asset form that meets the requirements of blockchain storage and verification is generated, and the encrypted data is stored off-chain for backup, while the data hash value and related index information are stored on-chain.

[0019] Optionally, the present invention uses ERC1155 for batch casting. The unique mapping key is a globally unique code generated based on the key attributes of the cultural relic digital assets, and a hash algorithm is used to generate a fixed-length string as the mapping key value.

[0020] Optionally, the smart contract of this invention generates a URI and associates it with the on-chain verification information and off-chain storage location information of the 3D model; based on the URI, the complete on-chain and off-chain information of the 3D model can be obtained, and the effective connection between on-chain and off-chain information can be completed.

[0021] The smart contract has an authorized casting institution management mechanism. The list of authorized casting institution addresses is strictly controlled, and only authorized addresses on the list have the right to call the batch casting function.

[0022] The management mechanism for the list of authorized minting institution addresses is as follows: adding and removing institutions requires review and confirmation by the signer of the multisignature wallet. When adding an institution, a qualification certificate must be submitted, the hash of the qualification certificate is uploaded to the blockchain, and the original file is stored in IPFS and associated with the contract address. When removing an institution, the reason must be explained and the information must be published on the blockchain explorer.

[0023] Optionally, the technical metadata of the present invention includes: three-dimensional model topology parameters and texture mapping algorithm identifiers, the topology parameters are encoded using PER encoding, and structured data such as vertex connection methods are compressed into binary streams; the process metadata includes: recording the calibration log summary of the acquisition device and the encrypted fingerprint of the data transmission link, forming a full-link evidence chain of "acquisition-transmission-casting".

[0024] Optionally, the specific process of the associated metadata described in this invention is as follows: establishing and storing an association map of collections in the same series, and recording the mapping between collection ID and relative position coordinates in key-value pairs;

[0025] The smart contract has an association query function. When a user initiates a query request for a collection related to a certain cultural relic digital asset, the user enters the corresponding tokenId, and the query function is launched. First, the tokenId is validated. If the validation passes, the function searches for a list of tokenIds of other collections associated with the tokenId based on the predefined data structure in the smart contract. If there are associated collections, the list is returned to the user, enabling a quick query of associated collections; if there are no associated collections, the user is prompted with "No associated collections".

[0026] The smart contract also has an update association function, which is used to maintain and update the relationship between associated collectibles. When a new collectible is associated with an existing collectible, the tokenId of the existing collectible and the tokenId of the new collectible are input, and the update association function is called to add the tokenId of the new collectible to the association list corresponding to the existing collectible, thus completing the update of the association relationship.

[0027] Optionally, the on-chain and off-chain collaborative verification described in this invention involves storing metadata hashes and key parameters on-chain, backing up complete data off-chain through distributed storage nodes, and verifying the integrity of off-chain data through on-chain hashes during tracing. The timestamp layered verification identifies anomalies through three-level timestamp difference analysis, with the three timestamps being the hardware timestamp of the acquisition device, the on-chain minting timestamp, and the first transaction timestamp. The smart contract tracing engine is implemented by developing tracing functions that support targeted tracing based on metadata dimensions or relationships, and outputting a visual tracing chain report.

[0028] Beneficial effects:

[0029] This invention employs a layered adaptive architecture: a three-layer architecture of assets, protocols, and applications enables dynamic matching of blockchain resources with 3D digital artifacts, improving batch processing efficiency by over 30% compared to the traditional direct on-chain model. The technical challenge of this architecture lies in achieving dynamic collaboration across the three layers and precise resource scheduling. With the same hardware costs, batch processing efficiency is improved by over 30%, and the unit storage cost is reduced (by approximately 25%). This unit storage cost includes server hardware (40%), IPFS storage (30%), and manual maintenance (30%). After optimization, hardware utilization is improved by 30%, storage redundancy is reduced by 15%, and overall costs are reduced by 25%. In the batch storage of large-scale 3D digital artifacts of cultural heritage, this invention can quickly and efficiently complete the storage of a large number of artifacts. For example, when storing hundreds or thousands of digital cultural heritage artifacts, it can significantly shorten the overall storage time.

[0030] This invention employs a customized ERC1155 casting mechanism: a dual-array association and pre-casting verification mechanism, to solve the atomicity and correlation problems in the batch casting of 3D assets, reducing the casting failure rate to below 0.5%. The technical challenge of this mechanism lies in the precise association of the dual arrays and the efficient implementation of the pre-casting verification logic. Compared to traditional casting methods, it can cast approximately 40% more collectibles in the same amount of time, with a significantly reduced casting failure rate, minimizing resource waste caused by failures. In the digital collectibles issuance scenario, it can reduce resource waste and issuance delays due to casting failures.

[0031] This invention employs a three-dimensional metadata system, adding dimensions such as topological parameters and relational graphs to upgrade from "information recording" to "knowledge construction." The technical challenge lies in the rational design of the metadata dimensions and the dynamic maintenance of relationships. Compared to traditional metadata systems, it provides more comprehensive and in-depth information about collectibles, offering strong support for the trading and research of digital collectibles. In digital collectible trading, it can provide buyers with more comprehensive information, such as the topological structure of the three-dimensional model and the relationships between collectibles in the same series, assisting them in making trading decisions.

[0032] This invention employs a multi-dimensional traceability mechanism: on-chain and off-chain collaboration + timestamp layering + intelligent engine, constructing a trusted verification closed loop, increasing traceability credibility to 99.9%. The technical challenge of this mechanism lies in the collaborative verification of multi-dimensional data and the efficient operation of the traceability engine. The improved traceability credibility makes the authentication of digital collectibles more accurate, reducing the risk of counterfeiting. In the scenario of authenticating digital collectibles, it can provide users with accurate traceability information, effectively preventing counterfeit digital collectibles.

[0033] The innovations are not isolated but rather form a technological synergy. The layered architecture provides resource support for ERC1155 casting, ensuring the efficiency of batch casting; the three-dimensional metadata provides the data foundation for multi-dimensional traceability, ensuring the comprehensiveness and accuracy of traceability information. The layered adaptation architecture, in conjunction with customized ERC1155 casting, improves batch casting efficiency by 15% compared to using either technology alone; the three-dimensional metadata system, in conjunction with the multi-dimensional traceability mechanism, improves the completeness of traceability information by 40%. Attached Figure Description

[0034] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0035] Figure 1 A flowchart illustrating the data interaction process in a layered architecture;

[0036] Figure 2 This is a schematic diagram of the multi-dimensional tracing logic. Detailed Implementation

[0037] The embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0038] It should be noted that, in the absence of conflict, the following embodiments and features can be combined with each other; and, based on the embodiments of this disclosure, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this disclosure.

[0039] It should be noted that various aspects of embodiments within the scope of the appended claims are described below. It will be apparent that the aspects described herein can be embodied in a wide variety of forms, and any particular structure and / or function described herein is merely illustrative. Based on this disclosure, those skilled in the art will understand that one aspect described herein can be implemented independently of any other aspect, and two or more of these aspects can be combined in various ways. For example, any number of aspects set forth herein can be used to implement the device and / or practice the method. Additionally, this device and / or method can be implemented using structures and / or functionalities other than one or more of the aspects set forth herein.

[0040] like Figure 1 As shown in the embodiment of this application, an optimized method for blockchain-based evidence storage of three-dimensional digital collectibles is described, and the specific process is as follows:

[0041] S1. Set up a layered blockchain adaptation architecture, which includes an asset layer, a protocol layer, and an application layer;

[0042] The asset layer preprocesses the 3D digital collectible model, including lightweighting the 3D digital collectible model and generating on-chain and off-chain mapping relationships.

[0043] The protocol layer dynamically schedules blockchain transaction processing resources and performs deep feature extraction, fusion, and encryption processing on the preprocessed 3D model data. The encrypted data is stored off-chain for backup, and the data hash value and related index information are stored on-chain.

[0044] The application layer monitors the evidence storage status, which includes casting progress, metadata on-chain status, traceability link integrity, and abnormal early warning information.

[0045] S2, Batch Casting: Establish a dual-array association mechanism to associate the basic information array and the three-dimensional feature array of the collection through a unique mapping key; perform pre-casting verification before batch casting to generate a verified Merkle root, and only execute the actual casting after the root hash is successfully uploaded to the chain.

[0046] S3, Metadata 3D Recording: Records technical metadata and process metadata of the casting process, and establishes a dynamically related 3D data network to associate metadata;

[0047] S4, multi-dimensional traceability verification: achieved through on-chain and off-chain collaborative verification, timestamp-layered verification, and smart contract traceability engine.

[0048] The process of the above steps will be explained in detail below:

[0049] S1, setting up a layered blockchain adaptation support architecture

[0050] Asset Layer: A 3D asset preprocessing module is deployed, primarily responsible for cleaning, format conversion, and basic feature extraction of the original 3D model data. Cleaning removes noise and outliers generated during data acquisition, ensuring data quality. Format conversion unifies 3D model formats from different sources into glTF format for subsequent processing. Basic feature extraction performs preliminary analysis and extraction of fundamental information such as the model's geometry and dimensions, providing a standardized and clean data foundation for later processing. For example, during the digitization of cultural relics, models in various formats such as OBJ and FBX may be acquired; the preprocessing module converts these to glTF format to improve data compatibility. When deploying the 3D asset preprocessing module, the digital collection model is first lightweighted (preserving accuracy parameters while compressing redundant data). The industry-standard glTF (GLTransmissionFormat) format is used to store the 3D model. This format, developed by the Khronos Group, features scalability, strong compression, and broad industry support. It uses JSON to describe the model structure and binary to store data such as geometry, materials, textures, and animations, which can effectively reduce file size and improve transmission and loading efficiency; it also supports PBR (physically based rendering) to achieve realistic rendering effects. For example, it can efficiently parse and display glTF format models in common 3D engines and applications, ensuring the stable transfer of digital cultural relics assets across different platforms.

[0051] A lightweight 3D model compression module is deployed for processing. Specifically, a 3D discrete wavelet transform algorithm is used, employing the db4 wavelet basis function and a decomposition level of 3 to balance compression efficiency and accuracy. Wavelet decomposition is performed on the vertex coordinates (x, y, z) and texture pixel values ​​(R, G, B) of the 3D model. Vertex coordinates are decomposed using 3D discrete wavelet transform, while texture pixels are decomposed using 2D wavelet transform, ensuring the compressibility of geometric and texture information. The 3D model is then further decomposed to obtain low-frequency approximation components and high-frequency detail components. The low-frequency approximation components represent the main shape and general outline of the model, including basic information such as the overall structure and main geometric features, reflecting the macroscopic form of the artifact model, such as the overall shape of a Buddha statue or the main shape of a bronze vessel. The high-frequency detail components reflect subtle changes and texture features on the model's surface, such as the subtle facial expressions of a Buddha statue or the decorative texture details on the surface of a bronze vessel. This frequency domain decomposition of the model enables the separation of features at different levels, providing multi-dimensional data support for subsequent fusion and processing. Low-frequency components are retained, while high-frequency components are processed using a threshold quantization method. The threshold is dynamically adjusted according to the model accuracy requirements (range: 0.01-0.1).

[0052] Compared to the JPEG2000 3D extension algorithm, the wavelet transform compression algorithm used in this invention achieves a 20% increase in compression speed and a 15% reduction in file size while maintaining the same accuracy loss. The comparative data is based on JPEG2000 3D extension algorithm version 3.2, with the test model being a grotto Buddha statue model with 2 million triangle faces and a 4K texture resolution. The aforementioned advantages in compression speed and file size are based on tests conducted on 100 3D models of varying complexity (500,000-5 million triangle faces, 2K-8K texture resolution) under the same hardware (CPU: Intel Core i7-10700K and above, memory: 32GB and above) and software (WaveletToolbox v9.1 wavelet transform library) and ambient temperature (25℃±2℃). The average compression time of this invention's algorithm is 12 seconds, while the average compression time of the JPEG2000 algorithm is 15 seconds. When the ambient temperature is >40℃ or <0℃, the backup compression algorithm (lossless compression based on LZ77) is automatically activated to ensure an accuracy loss rate of ≤2%, while triggering a hardware temperature control warning. Of the 100 samples, 30 are high-precision models and 70 are ordinary-precision models; the compression speed is improved by 18%-22% for models with 500,000-2 million triangle faces, and by 19%-21% for models with 2 million-5 million triangle faces; the compression speed is improved by 18%-20% for models with 2K texture resolution and by 20%-22% for models with 8K texture resolution. The accuracy loss rate is ≤1%. The verification method is to compare the triangle facet error and texture pixel deviation of the model before and after compression. The triangle facet error value is calculated as (triangle facet error after compression - original facet error) / original facet error × 100%. Through testing on 100 sets of samples, this value is ≤5%. The texture pixel deviation is ≤2%. The accuracy loss rate is calculated by combining the results. The specific formula is "Accuracy loss rate = (triangle facet error × 0.6 + texture pixel deviation × 0.4) × 100%". The compression performance data under different precision requirements are as follows: When the model precision requirement is high (for cultural relic display), with a threshold of [0.01-0.05], the compressed file size is reduced by 10%-12%, and the precision loss rate is ≤0.5%; when the precision requirement is normal (for circulation of ordinary digital collections), with a threshold of (0.05-0.1], the compressed file size is reduced by 15%-18%, and the precision loss rate is ≤1%. Simultaneously, a fault tolerance mechanism for the compression algorithm is set. When the ambient temperature exceeds 25℃±2℃, the temperature compensation coefficient is automatically activated (increasing the compression threshold correction by 0.02 for every 1℃ deviation) to ensure that the precision loss rate remains ≤1.5%.

[0053] The deployment mapping relationship generation module constructs a mapping relationship between on-chain storage formats and off-chain backup indexes through smart contracts. Within the smart contract, a specific data structure (such as a mapping table) is created. The on-chain storage format uses the hash value of the cultural relic digital asset as a unique identifier, while the corresponding off-chain backup index is stored in an off-chain database (such as the index database associated with the IPFS distributed file system). The index address and the on-chain hash value are linked and recorded in the smart contract. For example, when an on-chain query request for a certain cultural relic digital asset is initiated, the smart contract retrieves the corresponding off-chain index address from the mapping table based on the hash value, thereby locating the off-chain backup data. This achieves accurate matching and rapid querying of on-chain and off-chain information, ensuring data integrity and traceability, while also resolving the contradiction between large-capacity storage of 3D models and blockchain efficiency.

[0054] Protocol Layer: Develop smart contract adaptation interfaces to support dynamic invocation of transaction processing resources from the underlying blockchain. A resource scheduling module is deployed at the protocol layer. Utilizing a resource scheduling algorithm, it monitors the remaining computing power of nodes in real time. When a batch minting request is triggered, processing resources are allocated according to a priority coefficient between the number of collectibles and the node's load. The priority coefficient = minting quantity × 0.6 + node idle computing power × 0.4. This algorithm does not modify the consensus mechanism; it only optimizes local node resource allocation to ensure efficient transaction processing during batch minting.

[0055] In addition, a node failure emergency plan is set up so that when the assigned node fails (does not respond within 3 seconds), the task will be automatically transferred to the node with the second highest priority coefficient. The transfer time is ≤1 second, ensuring that the total casting time deviation is ≤2 seconds.

[0056] The protocol layer deploys a 3D model processing module responsible for deep feature extraction, fusion, and encryption of preprocessed 3D model data. Deep feature extraction utilizes machine learning algorithms to uncover deep-level texture and structural features of the model; the fusion operation integrates feature data from different channels and of different types to enrich the model's feature dimensions; and encryption employs asymmetric encryption algorithms to encrypt critical data, ensuring data security. Through these processes, a digital asset format that meets blockchain storage and verification requirements is generated. The on-chain / off-chain collaboration mechanism involves storing encrypted data off-chain after data processing and storing the data hash value and related index information on-chain, leveraging the immutability of blockchain to ensure data integrity verification. Manual review intervenes at key nodes (such as feature extraction result review and encrypted data verification) to manually confirm the automated processing results, correcting any potential deviations and thus comprehensively enhancing the credibility of digital cultural relic assets.

[0057] Application layer: Construct a visual interface and status monitoring module to monitor casting progress, metadata on-chain status, traceability chain integrity, and anomaly warning information in real time, providing users with operation feedback and anomaly warnings.

[0058] This architecture breaks through the traditional "direct on-chain" model, and achieves dynamic matching between blockchain resources and the characteristics of 3D digital collectibles through layered adaptation, improving batch processing efficiency by more than 30% compared with existing technologies.

[0059] S2, ERC1155 Customized Batch Casting Method

[0060] A dual-array association mechanism is employed: a basic information array (storing core attributes such as collection ID and name) and a three-dimensional feature array (storing model hash values, precision levels, etc.) are designed. Association is achieved through a unique mapping key. This unique mapping key is a globally unique code generated based on the key attributes of the digital cultural relics assets (such as category, age, region, and digitization acquisition time). A fixed-length string is generated using a hash algorithm (such as SHA-256) as the mapping key value. Taking the association between on-chain storage records and off-chain backup data of digital cultural relics assets as an example, this mapping key value is embedded in the metadata of the digital cultural relics assets stored on-chain. Simultaneously, the same mapping key value is recorded in the index file of the off-chain backup data. When data needs to be queried or retrieved, the corresponding relevant information of the digital cultural relics assets is accurately located in different storage locations on and off-chain based on this unique mapping key, ensuring data consistency and integrity and achieving efficient data association and management.

[0061] The smart contract has the ability to associate generated URIs with corresponding 3D models, implemented through specific functional logic. Once the URI (Uniform Resource Identifier) ​​of the 3D model is generated, the association function in the smart contract is activated. This function binds and records the URI with the off-chain storage location information of the 3D model (such as an IPFS address) and on-chain verification information (such as the model hash value and digital signature). This allows for subsequent queries to retrieve complete on-chain and off-chain information about the 3D model using this URI, achieving effective integration of on-chain and off-chain information. Simultaneously, the smart contract includes an authorized minting mechanism. The list of authorized minting addresses (authorizedMinters) is strictly controlled, and only authorized addresses on the list are authorized to call the batch minting function. The addition and removal of authorized minting institutions follow a strict process. When adding an institution, proof of its qualifications (such as a certificate of qualification for digital acquisition of cultural heritage or a license for the issuance of digital collectibles) must be submitted. This certificate is first converted to PDF format, then its hash value is calculated and stored on the blockchain. The original file is stored in IPFS and associated with a smart contract address to ensure the certificate is traceable and tamper-proof. When removing an institution, the reason for removal must be stated in the smart contract, and the removal instruction must be publicly displayed on the blockchain explorer for 72 hours. The start time for this display is the on-chain timestamp of the multi-signature wallet confirming the removal instruction, subject to network-wide supervision to ensure transparency and fairness in authorization management. The specific `setTokenURI` function associates the generated URI with the corresponding 3D model, allowing the user to query the off-chain storage location of the 3D model and its on-chain verification information, achieving effective integration of on-chain and off-chain information. `AuthorizedMinters` is a list of authorized minting institution addresses; only authorized addresses can call the batch minting function. The management mechanism of AuthorizedMinters is as follows: adding and removing minters requires review and confirmation by the signers of the multisignature wallet (at least 3 / 5). When adding minters, the institution's qualification certificate (such as a certificate of qualification for digital collection of cultural heritage or a license for the issuance of digital collections) must be submitted. The qualification certificate is uploaded to the blockchain in PDF format and the original file is stored in IPFS and associated with the contract address. When removing minters, the reason must be explained and the information must be published on the blockchain explorer for 72 hours. The 72-hour publication period starts from the on-chain timestamp of the multisignature wallet confirming the removal instruction.

[0062] Atomicity Guarantee in Casting: Atomicity in casting is guaranteed through a pre-casting verification mechanism. Before mass casting of collectibles, the system performs a comprehensive pre-verification of the model hashes and metadata of all collectibles off-chain. Data that passes verification generates Merkle roots. Only when the Merkle root hash is successfully uploaded to the blockchain and verified by the blockchain network will the actual casting operation be triggered. This mechanism avoids asset chaos caused by failures in part of the casting process, ensuring the atomicity of the entire casting process—either all succeed or all fail. Furthermore, smart contracts are regularly scanned monthly using formal verification tools (such as Certora) to detect common smart contract vulnerabilities, including integer overflows, reentrancy attacks, access control issues, and logical vulnerabilities. Scan results are stored synchronously on-chain for easy access by relevant personnel, ensuring the security and stability of the smart contracts.

[0063] S3, Metadata Three-Dimensional Recording System

[0064] Technical metadata: New professional parameters have been added, including 3D model topology parameters (such as vertex connection methods) and texture mapping algorithm identifiers. Topology parameters are encoded using PER (Packed Encoding Rules), compressing structured data such as vertex connection methods into a binary stream, reducing the storage size of a single metadata field by 60%. PER encoding improves topology parameter storage efficiency by 60% by converting vertex connection relationships (such as triangle face indices) into a variable-length binary code stream and using differential encoding for repetition patterns (e.g., using 1 bit to identify consecutive vertex indices with a difference of 1).

[0065] Process metadata: Records the calibration log summary of the acquisition device (such as the last calibration time and deviation value) and the encrypted fingerprint of the data transmission link, forming a complete evidence chain of "acquisition-transmission-casting".

[0066] Related metadata: Establish a relational map of collections within the same series, such as the spatial relationships and historical connections between collections in the "Five Caves of Tan Yao". The relational map is stored in JSON format, recording the mapping between collection IDs and relative location coordinates as key-value pairs, such as...

[0067] {"tokenId1":{"x":10,"y":20,"z":5},"tokenId2":{"x":15,"y":25,"z":8}}. The smart contract has a related query function, implemented through a specific query function. When a user initiates a query request for a collection related to a certain cultural relic digital asset, they enter the corresponding tokenId (a unique token number identifying the cultural relic digital asset). The query function is activated, first validating the tokenId to ensure its existence and conformity to the format specifications. If the validation passes, the function uses a predefined data structure (such as a mapping table) within the smart contract to search for a list of tokenIds of other collections associated with that tokenId. If associated collections exist, the list is returned to the user, enabling quick querying of associated collections; if no associated collections exist, the user is prompted with "No associated collections". The smart contract also has an update function for maintaining and updating the relationships between associated collections. When a new collectible is associated with an existing one, the system inputs the tokenId of the original collectible (originalTokenId) and the tokenId of the new collectible (newTokenId) to invoke the update association function. This function first verifies the caller's identity; only entities with administrator privileges (through the "onlyOwner" permission control mechanism) can perform this operation. After successful verification, the function adds the tokenId of the new collectible to the associated list of the original collectible, completing the update of the association relationship and ensuring the real-time nature and accuracy of the associated collectible data in the smart contract.

[0068] `relatedMap` (a data structure storing the relationships between items in the same series, essentially a mapping table defined within a smart contract, used to record the association between unique token IDs of different items, supporting quick querying and updating of the list of related items for a specific item. Through `relatedMap`, smart contracts can efficiently query and manage related items, providing underlying data support for maintaining the association relationships of digital assets such as the "Tan Yao Five Caves" series of cultural relics) is initialized when items in the same series are minted, writing the association data; when a new related item is generated, a specific smart contract function updates `relatedMap`. The core logic of this update function is: only entities with contract owner permissions are allowed to call this function, and two key parameters must be input when calling: the unique identifier of the original item (`originalTokenId`) and the unique identifier of the new related item (`newTokenId`). When the function is executed, the caller's permissions are strictly verified first. After confirming that the caller is the contract owner, the unique identifier of the new related item (`newTokenId`) is added to the corresponding association list of the original item (`originalTokenId`) in `relatedMap`, thus completing the association update operation. The identity of the contract owner is verified through a multi-signature wallet mechanism, requiring confirmation from at least 3 / 5 of the signers to obtain contract owner privileges. Simultaneously, function execution must also meet the conditions that the new associated collectible has been minted and its metadata verification has passed. Specific rules for verifying the metadata of the new associated collectible include ensuring that the metadata hash matches the on-chain record and that the association relationship conforms to preset logic (e.g., spatial location deviation ≤ 5 units), to ensure the integrity and security of the relatedMap update logic.

[0069] This function can only be called by the contract owner, who verifies their identity through a multi-signature wallet mechanism (requiring confirmation from at least 3 / 5 of the signers). The signers in the multi-signature wallet can be representatives of cultural heritage institutions, technical operations teams, or third-party auditing agencies. Execution is conditional upon the completion of the minting of the new associated collectible and the successful verification of its metadata. Specific rules for verifying the metadata of the new associated collectible include ensuring that the metadata hash matches the on-chain record and that the association relationship conforms to preset logic (e.g., spatial location deviation ≤ 5 units), ensuring the integrity and security of the relatedMap update logic. The qualification review standard for authorized minting institutions is that they possess relevant cultural heritage digital acquisition qualifications or digital collectible issuance licenses, which are jointly reviewed and confirmed by the multi-signature wallet signers.

[0070] The innovation lies in breaking through the traditional static recording mode of metadata and constructing a dynamically linked three-dimensional data network, so that metadata can not only be used for tracing the source, but also support the construction of knowledge graphs for digital collections.

[0071] S4, Dimensional Source Tracing Trust Mechanism

[0072] On-chain and off-chain collaborative verification: Metadata hashes and key parameters are stored on-chain, while complete original data is stored off-chain (backed up via distributed storage nodes). The distributed storage nodes employ the 3-replica backup strategy described in claim 6, with nodes distributed across different regions (e.g., one replica node each in North China, East China, and South China). Specific node configuration parameters include an 8-core CPU, 16GB of memory, 1TB of SSD storage, and a network bandwidth ≥100Mbps, ensuring data read / write speeds ≥100MB / s and guaranteeing off-chain data security. During tracing, on-chain hashes are used to verify the integrity of off-chain data, resolving the conflict between on-chain storage capacity and tracing depth.

[0073] Timestamp Layered Verification: A three-tiered timestamp system is designed—the hardware timestamp of the data collection device (immutable), the on-chain minting timestamp, and the first transaction timestamp. Abnormal assets are identified through time difference analysis. Assets with abnormal timestamps are automatically marked as "pending verification," suspending their trading function until manual review (2 / 3 signature confirmation from a multi-signature wallet). The abnormal record is permanently stored on the blockchain. Anomalies are marked if the difference between the hardware timestamp and the minting timestamp exceeds one hour. This threshold is based on statistics from 1000 normal data collections, with normal time differences ≤ 30 minutes. The criteria for judging abnormal first transaction timestamps are: if the first transaction timestamp is earlier than the minting timestamp, or if the difference exceeds 72 hours (set according to the normal circulation cycle of digital collectibles), it is marked as abnormal.

[0074] The smart contract traceability engine develops dedicated traceability functions to support targeted traceability based on metadata dimensions (e.g., "filtering collections with an accuracy level ≥ 0.9") or relationships (e.g., "querying derivative collections related to Cave 3"), outputting a visual traceability chain report. When tracing by metadata, it iterates through on-chain metadata, filters collection IDs that meet the accuracy level ≥ 0.9 condition, and returns relevant information. When tracing by relationships, it queries derivative collection IDs and corresponding information related to the target collection based on the relationships stored in the relatedMap. The visual traceability chain report is presented in a web-based interactive interface. Core functional modules include a timestamp comparison chart (displaying three levels of timestamp values ​​and differences), a hash verification result pop-up (displaying on-chain and off-chain hash comparisons), and a related collection graph (clickable and interactive). When a user clicks on a node in the related collection graph, the system automatically loads the corresponding collection's lightweight 3D model preview file (called via an IPFS link). The technical parameters of the lightweight 3D model preview file are: glTF format, file size ≤ 10MB, face count ≤ 100,000, ensuring a loading speed ≤ 2 seconds. The report includes three levels of timestamp comparison results, on-chain and off-chain hash verification status, a related collection graph, and a credibility score (0-10 points). The specific calculation rules for the credibility score are: Credibility Score = Normal Timestamp (4 points) + Passed Hash Verification (5 points) + Complete Related Graph (1 point). Abnormal timestamps deduct 4 points, failed hash verification deducts 5 points, and missing related graphs deduct 1 point. Simultaneously, a manual review mechanism for abnormal situations is set up. When the score is ≤3 points, a review by the multi-signature wallet signers (at least 2 / 3) is automatically triggered. The review result serves as the final traceability conclusion and is synchronously recorded on the blockchain. For example... Figure 2 As shown.

[0075] Table 1 Comparison between the present invention and prior art

[0076]

[0077]

[0078] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. An optimized method for blockchain-based evidence storage of three-dimensional digital collectibles, characterized in that, The specific process is as follows: S1, Set up a layered blockchain adaptation architecture, which includes an asset layer, a protocol layer, and an application layer; The asset layer preprocesses the 3D digital collectible model, including lightweighting the 3D digital collectible model and generating on-chain and off-chain mapping relationships. The protocol layer dynamically schedules blockchain transaction processing resources and performs deep feature extraction, fusion, and encryption processing on the preprocessed 3D model data. The encrypted data is stored off-chain for backup, and the data hash value and related index information are stored on-chain. The application layer monitors the evidence storage status, which includes casting progress, metadata on-chain status, traceability link integrity, and abnormal early warning information. S2, Batch Casting: Establish a dual-array association mechanism to associate the basic information array and the three-dimensional feature array of the collection through a unique mapping key; perform pre-casting verification before batch casting to generate a verified Merkle root, and only execute the actual casting after the root hash is successfully uploaded to the chain. S3, Metadata 3D Recording: Records technical metadata and process metadata of the casting process, and establishes a dynamically related 3D data network to associate metadata; S4, multi-dimensional traceability verification: achieved through on-chain and off-chain collaborative verification, timestamp-layered verification, and smart contract traceability engine.

2. The optimized method for blockchain-based evidence storage of three-dimensional digital collectibles according to claim 1, characterized in that, The lightweighting process is as follows: First, wavelet decomposition is performed on the vertex coordinates and texture pixel values ​​of the 3D model, where the vertex coordinates are transformed using a 3D discrete wavelet transform and the texture pixels are transformed using a 2D wavelet transform. Second, the 3D model is decomposed in the frequency domain into low-frequency approximation components and high-frequency detail components. The low-frequency approximation components represent the main shape and general outline of the model, while the high-frequency detail components reflect the subtle changes and texture characteristics of the model surface. The low-frequency components are retained, and the high-frequency components are processed using a threshold quantization method. When the model accuracy requirement is high, the threshold is set to [0.01-0.05]; when the accuracy requirement is normal, the threshold is set to (0.05-0.1); the compression algorithm fault tolerance mechanism is set, and when the ambient temperature exceeds the range of 25℃±2℃, the temperature compensation coefficient is automatically enabled, and the compression threshold correction amount is increased by 0.02 for every 1℃ deviation.

3. The optimized method for blockchain-based evidence storage of three-dimensional digital collectibles according to claim 2, characterized in that, The on-chain and off-chain mapping relationship between the on-chain storage format and the off-chain backup index is constructed through smart contracts. The on-chain storage format uses the hash value of the cultural relic digital asset as a unique identifier, while the corresponding off-chain backup index is stored in the off-chain database, and the index address and the on-chain hash value are associated and recorded in the smart contract.

4. The optimized method for blockchain-based evidence storage of three-dimensional digital collectibles according to claim 3, characterized in that, The asset layer deploys a 3D asset preprocessing module to clean, convert formats, and extract basic features from the lightweight processed data. The cleaning operation aims to remove noise points and abnormal data generated during the data acquisition process. Format conversion unifies the 3D model formats from different sources into the glTF format; basic feature extraction performs preliminary analysis and extraction of the model's basic information.

5. The optimized method for blockchain-based evidence storage of three-dimensional digital collectibles according to claim 1, characterized in that, The protocol layer dynamically schedules blockchain transaction processing resources, specifically by: monitoring the remaining computing power of nodes in real time; when a batch casting request is triggered, allocating processing resources based on the priority coefficient calculated according to the number of collectibles and the node load, where the priority coefficient = number of castings × 0.6 + node idle computing power × 0.4; and simultaneously setting up a node failure emergency plan, automatically transferring the task to the node with the second highest priority coefficient when the allocated node fails. The protocol layer deploys a 3D model processing module, which is responsible for performing deep feature extraction, fusion, and encryption operations on the preprocessed 3D model data. Deep feature extraction uses machine learning algorithms to mine the deep texture and structural features of the model. The fusion operation integrates feature data from different channels and of different types. The encryption process uses an asymmetric encryption algorithm to encrypt key data. Through the above processing, a digital asset form that meets the requirements of blockchain storage and verification is generated, and the encrypted data is stored off-chain for backup, while the data hash value and related index information are stored on-chain.

6. The optimized method for blockchain-based evidence storage of three-dimensional digital collectibles according to claim 5, characterized in that, The ERC1155 is used for batch casting. The unique mapping key is a globally unique code generated based on the key attributes of the cultural relic digital assets. A hash algorithm is used to generate a fixed-length string as the mapping key value.

7. The optimized method for blockchain-based evidence storage of three-dimensional digital collectibles according to claim 1, characterized in that, During the batch casting stage, a URI is generated through a smart contract and associated with the on-chain verification information and off-chain storage location information of the 3D model; based on the URI, the complete on-chain and off-chain information of the 3D model can be obtained, and the effective connection between on-chain and off-chain information is completed. The smart contract has an authorized casting institution management mechanism. The list of authorized casting institution addresses is strictly controlled, and only authorized addresses on the list have the right to call the batch casting function. The management mechanism for the list of authorized minting institution addresses is as follows: adding and removing institutions requires review and confirmation by the signer of the multisignature wallet. When adding an institution, a qualification certificate must be submitted, the hash of the qualification certificate is uploaded to the blockchain, and the original file is stored in IPFS and associated with the contract address. When removing an institution, the reason must be explained and the information must be published on the blockchain explorer.

8. The optimized method for blockchain-based evidence storage of three-dimensional digital collectibles according to claim 1, characterized in that, The technical metadata includes: 3D model topology parameters and texture mapping algorithm identifiers. The topology parameters are encoded using PER encoding, and structured data such as vertex connection methods are compressed into a binary stream. The process metadata includes: recording the calibration log summary of the acquisition device and the encrypted fingerprint of the data transmission link, forming a complete evidence chain of "acquisition-transmission-casting".

9. The optimized method for blockchain-based evidence storage of three-dimensional digital collectibles according to claim 8, characterized in that, The specific process of the associated metadata is as follows: establish and store the association map of the same series of collections, and record the mapping between collection ID and relative position coordinates in key-value pairs; The smart contract has an association query function. When a user initiates a query request for a collection related to a certain cultural relic digital asset, the user enters the corresponding tokenId, and the query function is launched. First, the tokenId is validated. If the validation passes, the function searches for a list of tokenIds of other collections associated with the tokenId based on the predefined data structure in the smart contract. If there are associated collections, the list is returned to the user, enabling a quick query of associated collections; if there are no associated collections, the user is prompted with "No associated collections". The smart contract also has an update association function, which is used to maintain and update the relationship between associated collectibles. When a new collectible is associated with an existing collectible, the tokenId of the existing collectible and the tokenId of the new collectible are input, and the update association function is called to add the tokenId of the new collectible to the association list corresponding to the existing collectible, thus completing the update of the association relationship.

10. The optimized method for blockchain-based evidence storage of three-dimensional digital collectibles according to claim 1, characterized in that, The on-chain and off-chain collaborative verification involves storing metadata hashes and key parameters on-chain, while backing up complete data off-chain through distributed storage nodes. During tracing, the integrity of off-chain data is verified using on-chain hashes. The timestamp hierarchical verification identifies anomalies through three-level timestamp difference analysis. The three levels of timestamps are the hardware timestamp of the data acquisition device, the on-chain minting timestamp, and the first transaction timestamp. The smart contract tracing engine is implemented by developing tracing functions that support targeted tracing based on metadata dimensions or relationships, and outputting a visual tracing chain report.

Images