Digital copyright confirmation method and system based on multi-chain collaboration and dynamic verification

Abstract

The invention relates to the field of digital copyright, and discloses a digital copyright confirmation method and system based on multi-chain collaboration and dynamic verification, and the method comprises the steps: carrying out the local region division of a digital image to be confirmed, and combining with a content segmentation rule to generate a fingerprint index set; carrying out fragmentation processing on the local copyright feature sequence, and combining node redundancy, access delay and a fragmentation integrity verification mechanism; based on a distributed storage structure, mapping the local fingerprint fragments to a plurality of block chains to form a multi-chain collaborative verification channel; performing multi-dimensional similarity calculation on the local fingerprint set subjected to multi-chain verification, and identifying copyright conflict candidate events of potential local elements; and defining a local area corresponding to the copyright conflict candidate event as a copyright detection range, constructing a local copyright identification map, and outputting an accurate digital copyright confirmation result. The method has the advantage of improving the precision of local copyright recognition.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of digital copyright, in particular to a digital copyright right confirmation method and system based on multi-chain cooperation and dynamic verification. BACKGROUND

[0002] With the wide application of digital images and multimedia content, the creation, dissemination and transaction of digital works have shown explosive growth. However, digital works are easy to be copied, modified and illegally disseminated, and copyright protection faces many challenges. In recent years, blockchain technology has been introduced into digital copyright management, through decentralized storage and tamper-proof ledger, realizing the reliable recording and automatic confirmation of copyright information; the existing blockchain digital copyright confirmation method mainly relies on single-chain storage or overall data hash verification, which can ensure the tamper-proof and automatic management of copyright information to a certain extent. However, there are still some technical problems in practical application: insufficient local element copyright identification the existing system is based on image hash or overall feature verification, which cannot effectively determine the copyright ownership of small areas, repeated elements or local materials contained in the image, and there is a potential risk of copyright dispute; limited expansion of single-chain storage with the rapid growth of digital content, single-chain storage faces the problems of low transaction throughput, large storage pressure and insufficient verification efficiency, which is difficult to support large-scale copyright confirmation and real-time verification demand; insufficient data security and cross-chain cooperation the existing technology lacks verification mechanism for local data integrity and cross-chain consistency when dealing with distributed storage and multi-chain cooperation, which is easy to cause data island or inconsistent copyright confirmation information. Therefore, it is necessary to design a digital copyright confirmation method and system based on multi-chain cooperation and dynamic verification to improve the accuracy of local copyright identification. SUMMARY

[0003] In view of the deficiencies of the prior art, the present application provides a digital copyright confirmation method and system based on multi-chain cooperation and dynamic verification, which has the advantage of improving the accuracy of local copyright identification and solving the problems in the above background art.

[0004] To achieve the above purpose of improving the accuracy of local copyright identification, the present application provides the following technical scheme: a digital copyright confirmation method and system based on multi-chain cooperation and dynamic verification, comprising the following steps:

[0005] Divide the local area of the digital image to be confirmed into a local fingerprint index set based on texture features, color distribution and depth embedding representation extraction, and obtain a local copyright feature sequence;

[0006] Fragment the local copyright feature sequence and store multiple copies of it on distributed nodes, and combine node redundancy, access delay and fragment integrity checking mechanism to construct a traceable distributed storage structure;

[0007] Based on the distributed storage structure, the local fingerprint fragments are mapped to multiple blockchains, and combined with cross-chain hash verification, on-chain timestamp and historical record backtracking, a multi-chain collaborative verification channel is formed;

[0008] The multi-dimensional similarity of the local fingerprint set verified by multiple chains is calculated, and combined with the historical copyright sample library, conflict detection rules and dynamic threshold adjustment strategy, the potential local element copyright conflict candidate event is identified;

[0009] The local area corresponding to the copyright conflict candidate event is delimited as the copyright detection range, and the local fingerprint nodes, cross-chain storage nodes and historical copyright nodes in the range are fused to construct a local copyright identification graph, and the digital copyright right confirmation result is output.

[0010] Preferably, the process of generating a fingerprint index set in combination with the content segmentation rule is:

[0011] By performing multi-layer semantic segmentation on the pixel distribution in the local area, the significant edges, texture patterns and color gradient changes in the image are extracted;

[0012] The segmentation results are matched with the image depth embedding representation to generate corresponding multi-modal fingerprint feature labels;

[0013] According to the semantic similarity and spatial adjacency relationship between the feature labels, a multi-level fingerprint index tree is established;

[0014] The features of the local area are archived and numbered through the index tree, and the fingerprint index set is output.

[0015] Preferably, the process of obtaining a local copyright feature sequence is:

[0016] The fingerprint index set is unfolded according to the spatial coordinate order, and the context environment generated based on the timestamp label of the image is generated;

[0017] Combined with the matching relationship between the depth embedding features and the index labels, a vectorized representation of the local features is generated;

[0018] Using a serialization encoding mechanism, the local feature vectors are sequentially mapped to a feature sequence space;

[0019] By an anomaly detection mechanism, redundant or noisy features are removed, and finally a structured local copyright feature sequence is obtained.

[0020] Preferably, the process of fragmenting the local copyright feature sequence is:

[0021] According to the length and complexity of the local copyright feature sequence, a dynamic sliding window segmentation mechanism is used to divide the sequence into multiple continuous fragments;

[0022] Generate a unique identifier for each shard, and embed the source index and generation timestamp;

[0023] Through the multi-copy redundancy mechanism, staggered replication of each shard between different distributed nodes;

[0024] Using the shard integrity checking algorithm, hash signature and consistency verification are performed on each shard, and the final storable shard set is output.

[0025] Preferably, the process of building a traceable distributed storage structure is:

[0026] Distribute the shard set according to the identifier, and perform dynamic scheduling combined with the storage load of the node;

[0027] Based on the access delay model, select the optimal storage node for each shard, and use the replica staggered strategy to improve query efficiency;

[0028] In the storage process, use the chain check mechanism to record the hash value and timestamp of each storage operation;

[0029] Through cross-node consistency comparison, a distributed storage structure with tamper resistance is formed.

[0030] Preferably, the process of combining cross-chain hash verification, on-chain timestamp and historical record backtracking is:

[0031] Calculate the hash digest of each shard in the storage structure and submit it to multiple independent blockchains;

[0032] Add a generation timestamp to each hash digest on the chain as a time reference for shard verification;

[0033] Through the cross-chain synchronization mechanism, the hash records in different blockchains are compared for consistency;

[0034] Combined with the historical on-chain record backtracking mechanism, detect whether the shard data has been abnormally changed, and form a verification report.

[0035] Preferably, the process of forming a multi-chain collaborative verification channel is:

[0036] Through the cross-chain communication protocol, establish a data interaction interface between multiple chains;

[0037] Embed cross-chain verification rules in the verification channel to synchronize and align the hash results and timestamps of different chains;

[0038] Use the consensus mechanism to rule on the multi-chain verification results, and output the final collaborative verification conclusion.

[0039] Preferably, the process of identifying potential local element copyright conflict candidate events is:

[0040] Multi-dimensional similarity calculation is performed on the local fingerprint set that passes the multi-chain verification, including texture similarity, color distribution similarity, and depth feature similarity.

[0041] The similarity calculation result is compared with the historical copyright sample library, and it is determined whether there is copyright overlap based on the conflict detection rule.

[0042] Combined with the dynamic threshold adjustment strategy, the determination threshold is adaptively corrected according to the scene complexity and the historical sample density.

[0043] If the similarity result exceeds the threshold, mark the corresponding local element as a copyright conflict candidate event, and generate an event list.

[0044] Preferably, the output digital copyright confirmation result process is:

[0045] According to the copyright conflict candidate event list, the local area involved is delimited as the copyright detection range;

[0046] In the range, local fingerprint nodes, cross-chain storage nodes, and historical copyright nodes are fused to construct a local copyright identification graph.

[0047] The identification graph is sorted by node importance and edge weight calculation to identify key copyright nodes and conflict paths.

[0048] Combined with the identification graph and the conflict determination result, an accurate digital copyright confirmation conclusion is generated and output to the confirmation database and the user terminal.

[0049] The digital copyright confirmation system based on multi-chain cooperation and dynamic verification includes:

[0050] Local fingerprint extraction module: divide the digital image into regions, and extract texture, color and depth representation to generate local copyright feature sequence;

[0051] Fragment storage management module: fragment processing of local copyright feature sequence, and multi-copy interleaved storage and integrity check on distributed nodes;

[0052] Multi-chain mapping verification module: map the fragmented data to multiple blockchains, and realize collaborative verification combined with cross-chain hash and timestamp mechanism;

[0053] Similarity detection module: multi-dimensional similarity calculation is performed on the local fingerprint set that passes the multi-chain verification, including texture similarity, color distribution similarity, and depth feature similarity;

[0054] Confirmation result generation module: fuse local fingerprint nodes, cross-chain storage nodes, and historical copyright nodes to construct a copyright identification graph and output the confirmation result.

[0055] Compared with the prior art, the present application provides a digital copyright right confirmation method and system based on multi-chain cooperation and dynamic verification, which has the following beneficial effects:

[0056] The present application can finely depict the local features of the image by dividing the local area of the digital image to be confirmed and extracting local fingerprints in combination with texture features, color distribution and depth embedding representation, which improves the accuracy and stability of feature extraction compared with the overall fingerprint method; through the fragmentation storage management mechanism, the inter-distributed node performs multiple copy interleaved storage and integrity checking, which guarantees the persistent storage and tamper resistance of the local copyright data, and has traceability, effectively avoiding single point failure and data loss problems; by mapping the local fingerprint fragments to multiple blockchains and combining cross-chain hash verification, on-chain timestamp and historical record backtracking, a multi-chain cooperative verification channel is established, which breaks through the performance bottleneck of single-chain storage and improves the security, credibility and consistency of the right confirmation process; by performing multi-dimensional similarity calculation on the local fingerprint set verified by multiple chains, and combining the historical copyright sample library, conflict detection rules and dynamic threshold adjustment strategy, it can adapt to different complexity of copyright judgment scene, which improves the accuracy and robustness of potential copyright conflict detection; by constructing a local copyright identification map, the local fingerprint node, cross-chain storage node and historical copyright node are organically integrated, which not only reveals the conflict path and key copyright node, but also provides support for the explainability and transparency of the right confirmation result; in summary, the present application can realize accurate right confirmation of digital images, has high security, traceability and intelligent processing capability, and has the advantage of significantly improving the accuracy of local copyright identification. BRIEF DESCRIPTION OF DRAWINGS

[0057] Figure 1 The present application is a schematic diagram of the method;

[0058] Figure 2 The present application is a schematic diagram of the structure. DETAILED DESCRIPTION

[0059] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0060] Embodiment 1: Please refer to Figure 1 The digital copyright right confirmation method and system based on multi-chain cooperation and dynamic verification described in the embodiments of the present application include the following steps:

[0061] S1: obtaining a local copyright feature sequence by dividing a digital image to be authenticated into local regions, extracting corresponding local fingerprints based on texture features, color distribution and depth embedding representation, and generating a fingerprint index set in combination with a content segmentation rule;

[0062] The process of generating a fingerprint index set in combination with a content segmentation rule in S1 is as follows:

[0063] By performing multi-layer semantic segmentation on the pixel distribution in the local region, the significant edges, texture patterns and color gradient changes in the image are extracted.

[0064] The digital image to be authenticated is divided into a plurality of local regions, and the pixel matrix of each region is represented in the form of a two-dimensional array. In each local region, multi-layer semantic segmentation is performed on the pixel distribution. The segmentation algorithm includes an image segmentation network based on a convolutional neural network. In the segmentation process, significant edge contours, texture pattern distributions and color gradient change values are extracted respectively. The results are stored in the form of edge binary masks, texture feature matrices and color gradient vectors.

[0065] The segmentation results are matched with the image depth embedding representation to generate corresponding multi-modal fingerprint feature labels.

[0066] The obtained segmentation results are matched with the depth embedding representation of the image. The depth embedding representation is generated by a pre-trained feature extraction model and includes high-dimensional vector features. In the matching process, the edge mask, texture matrix and color gradient vector are similarity calculated with the corresponding embedding vector, and the cosine similarity or Euclidean distance is used as the measurement method. According to the matching results, multi-modal fingerprint feature labels are generated for each local region. The label content includes feature categories, coordinate ranges and corresponding embedding index numbers.

[0067] A multi-level fingerprint index tree is established according to the semantic similarity and spatial adjacency relationship between the feature labels.

[0068] A multi-level index tree structure is constructed according to the semantic similarity and spatial adjacency relationship between the feature labels. In the index tree, the first layer is classified according to the feature categories, the second layer is grouped according to the spatial adjacency relationship, and the third layer is sorted according to the semantic similarity. Each node of the index tree records the unique number of the feature label, the feature vector abstract and the adjacency pointer.

[0069] The features of the local regions are archived and numbered through the index tree, and a fingerprint index set is output.

[0070] The features of local regions are archived using an index tree, with each feature node corresponding to a unique number, and an archive path is generated according to the hierarchical relationship. All numbers and path information are organized into a structured index set, which is stored in the form of key-value pairs, where the key is the number and the value is the archive path and the corresponding feature vector summary. The resulting fingerprint index set is output as input data for the subsequent generation of local copyright feature sequences.

[0071] The process of obtaining the local copyright feature sequence in S1 is as follows:

[0072] The fingerprint index set is expanded in spatial coordinate order, and the context of image generation is marked based on timestamps;

[0073] Each fingerprint index in the set is expanded in the order of spatial coordinates of the corresponding local region. The spatial coordinates are represented by two-dimensional pixel coordinates (x, y) of the image. During the expansion process, a timestamp of the image generation time is added to each index entry. The timestamp adopts a standard time format and is used to record the correspondence between the index data and the generation environment.

[0074] By combining the matching relationship between deep embedding features and index labels, a vectorized representation of local features is generated;

[0075] The expanded index entries are matched with the deep embedding features of the image, which are stored in the form of multi-dimensional floating-point vectors. For each index entry, the corresponding embedding vector is retrieved according to its feature label number, and a mapping relationship is established. The index entries and their embedding vectors are combined to generate a vectorized representation of the local features. Each record contains: spatial coordinates, timestamp, index number, and embedding feature vector.

[0076] A serialization encoding mechanism is used to map local feature vectors sequentially to the feature sequence space;

[0077] A serialization encoding mechanism is adopted to arrange local feature vectors in order of spatial coordinates and timestamps. For the arranged feature vectors, a fixed-length or variable-length encoding method is used to map the high-dimensional vectors to the feature sequence space, and the output is an ordered sequence structure. Each sequence element contains a unique sequence number, original vector data and its context label.

[0078] Redundant or noisy features are removed by an anomaly detection mechanism, ultimately resulting in a structured local copyright feature sequence.

[0079] Anomaly detection is performed on the serialized feature sequence using methods such as statistical distribution analysis, cluster distance thresholding, or noise filtering algorithms. If redundant features are found, they are merged or deleted. If noise features are detected, they are marked and removed. The filtered structured local copyright feature sequence is output as input for subsequent segmentation processing.

[0080] S2: The local copyright feature sequence is fragmented and stored in multiple copies on distributed nodes. Combined with node redundancy, access latency and fragment integrity verification mechanism, a traceable distributed storage structure is constructed.

[0081] The process of segmenting the local copyright feature sequence in S2 is as follows:

[0082] Based on the length and complexity of the local copyright feature sequence, a dynamic sliding window segmentation mechanism is adopted to divide the sequence into multiple continuous segments;

[0083] Input the structured local copyright feature sequence obtained in the previous step. Each sequence element contains a feature vector, spatial coordinates, timestamp, and index number. Based on the length of the sequence and the feature complexity, set the sliding window length and sliding window step size. The sliding window length can be a fixed number of elements or dynamically adjusted based on the feature entropy. Use the sliding window to sequentially divide the feature sequence to generate multiple continuous segments. Each segment contains a continuous subset of feature vectors.

[0084] Generate a unique identifier for each shard and embed the source index and generation timestamp;

[0085] Generate a unique identifier for each fragment, which can be a globally unique identifier or a hash-based generation method; embed source index information, including the original sequence number and the corresponding image region identifier, into the fragment metadata; attach a generation timestamp to each fragment to record the time information of fragment generation for subsequent version management and traceability.

[0086] Through a multi-replica redundancy mechanism, each shard is replicated interleaved among different distributed nodes;

[0087] Each shard is replicated across different distributed storage nodes according to the required number of replicas, ensuring that each node stores a different combination of shards. During the replication process, the node location, storage time, and shard UID of each replica are recorded for shard tracking and access management. Node load monitoring and distribution strategies ensure that redundant replicas are evenly distributed throughout the system, meeting access efficiency and fault tolerance requirements.

[0088] A fragment integrity verification algorithm is adopted to perform hash signature and consistency verification on each fragment, and output the final storable fragment set.

[0089] For each shard, a hash value is calculated to generate a shard signature, which serves as the basis for uniqueness and integrity verification. Consistency verification is performed on the shard replicas by comparing the hash signatures on each node with the original shard hash value. The shards that pass verification are collected into a final storable shard set. The shard set that has undergone verification and redundancy processing is formed into a unified storage list, with each record containing the shard UID, original sequence information, storage node location, and verification status. This shard set is then output as input data for subsequent mapping to multi-chain blockchain verification and copyright conflict identification.

[0090] The process of constructing a traceable distributed storage structure in S2 is as follows:

[0091] The shard sets are distributed according to identifiers, and dynamic scheduling is performed based on the storage load of the nodes.

[0092] The input is the set of shards generated in the previous step. Each shard contains a unique identifier (UID), source sequence information, and verification status. Based on the shard UID, the set of shards is mapped to multiple distributed storage nodes, and dynamic scheduling is performed in combination with the current storage load and capacity limits of each node. The scheduling algorithm adopts a load balancing strategy to ensure that the number of shards allocated to each node matches the node's storage capacity, while recording the allocation mapping table.

[0093] The optimal storage node is selected for each shard based on the access latency model, and a replica interleaving strategy is adopted to improve query efficiency.

[0094] The access latency of each shard at the candidate nodes is calculated based on the access latency model, including network latency, read / write throughput, and node response time. The node with the lowest latency is selected as the primary storage node, and replicas are generated on other candidate nodes. An interleaving strategy is used to ensure that redundant replicas are evenly distributed. The storage node, generation timestamp, and shard UID of each replica are recorded to form a replica distribution table for subsequent access and tracking.

[0095] During the storage process, a chained verification mechanism is used to record the hash value and timestamp of each storage operation;

[0096] For each shard storage operation, a hash value is calculated and a chain of verification records is generated. The current shard hash is linked with the hash value of the previous operation to form a chain structure. A timestamp is attached to each chain record to mark the exact time point of shard storage. The chain of verification records are stored on distributed nodes or independent audit nodes for operation traceability and data verification.

[0097] By comparing consistency across nodes, a tamper-resistant distributed storage structure is formed;

[0098] After sharding is completed, a consistency comparison is performed on the shard replicas stored on each node, comparing hash values, UIDs, and timestamps. If the comparison results are consistent, the sharding is confirmed to be valid. If there are differences, anomaly recording and redundant replica repair operations are triggered. The comparison results are recorded in the distributed log to form a complete sharding traceability chain. The shard set after dynamic allocation, interleaved storage, and consistency verification is formed into a unified storage structure description, including a node distribution table, a chained verification chain, and replica information. This storage structure is output as input for subsequent multi-chain mapping verification and copyright confirmation modules.

[0099] S3: Based on a distributed storage structure, local fingerprint shards are mapped to multiple blockchains, and cross-chain hash verification, on-chain timestamps and historical record backtracking are combined to form a multi-chain collaborative verification channel;

[0100] The process in S3 that combines cross-chain hash verification, on-chain timestamps, and historical record backtracking is as follows:

[0101] Calculate a hash digest for each piece in the storage structure and submit it to multiple independent blockchains;

[0102] Input the distributed storage structure generated in the previous step. Each shard contains a unique identifier, storage node information, and chained verification records. Calculate a hash digest for each shard to obtain a unique hash value used to identify the shard content. Submit the hash digest and the corresponding shard UID to multiple independent blockchains through the blockchain node interface. Each chain processes the hash data independently.

[0103] Attach a generated timestamp to each hash digest on the chain as a time reference for shard verification;

[0104] While submitting the hash digest to the blockchain, an on-chain timestamp is generated to record the precise time of the submission operation; the timestamp, along with the shard UID and hash digest, is stored in the blockchain transaction record to form a time reference; and each blockchain transaction is ensured to contain complete verification metadata to support subsequent cross-chain comparison and traceability operations.

[0105] A cross-chain synchronization mechanism is used to perform consistency comparisons of hash records across different blockchains.

[0106] The hash records and timestamps of the same shard UID on different blockchains are obtained through a cross-chain synchronization mechanism; the records of each chain are compared to determine whether the hash digests are consistent and whether the timestamps match; the consistency comparison results are used to generate a comparison matrix to record the shard verification status between each blockchain.

[0107] By combining historical on-chain record backtracking mechanisms, we can detect whether there are abnormal changes in sharded data and generate a verification report.

[0108] For each shard UID, previous commit records, including hash values ​​and timestamps, are retrieved from the blockchain history. The hash digest of the current commit is compared with the historical records to determine if there are any abnormal changes or duplicate commits. If an abnormal change is found, the shard UID, chain number, change type, and time point are recorded and marked as pending. The cross-chain consistency comparison results and historical record backtracking detection results are integrated to generate a verification report. The report includes the hash digest, on-chain timestamp, comparison status, abnormal change records, and shard UID index table for each shard. This verification report is output as input data for multi-chain collaborative verification channels and copyright conflict identification.

[0109] The process of forming a multi-chain collaborative verification channel in S3 is as follows:

[0110] Establish data interaction interfaces between multiple chains through cross-chain communication protocols;

[0111] A cross-chain communication protocol is used to create a data exchange channel between multiple blockchains. Interface parameters are configured, including chain ID, node list, transmission protocol type and data format specifications. After the interface is established, bidirectional transmission tests are performed on key data such as shard UID and hash digest to ensure that the data can be recognized and received by different chain nodes.

[0112] Cross-chain verification rules are embedded in the verification channel to synchronize and align the hash results and timestamps of different chains.

[0113] Cross-chain verification rules are defined in the verification channel, including hash digest matching, timestamp synchronization, and on-chain metadata consistency checks. These rules are applied to the transmitted data, automatically comparing and synchronizing the hash values ​​and timestamps of the same shard across different chains. The comparison results are recorded in the cross-chain verification log, including the shard UID, hash values ​​of each chain, timestamps, and verification status.

[0114] A consensus mechanism is used to adjudicate the multi-chain verification results and output the final collaborative verification conclusion.

[0115] The system adjudicates the hash results of the same shard submitted by multiple blockchains, selecting the hash value with the best consistency as the final verification result. The consensus mechanism can adopt Byzantine fault tolerance, proof-of-stake, or other applicable algorithms to ensure that the verification results of multiple chains are consistent and traceable. The adjudication result is recorded in the verification channel log, and the final collaborative verification conclusion is output, including the shard UID, adjudication hash value, related chain ID, and timestamp. The final collaborative verification conclusion is compiled into a report for use by the subsequent copyright conflict identification module. The report includes the verification status of each shard, adjudication hash value, multi-chain source, and timestamp information to form a traceable record of multi-chain verification.

[0116] S4: Perform multi-dimensional similarity calculation on the local fingerprint set that has been verified by multiple chains, and combine it with the historical copyright sample library, conflict detection rules and dynamic threshold adjustment strategy to identify potential copyright conflict candidate events of local elements;

[0117] The process for identifying potential copyright conflict candidate events for local elements in S4 is as follows:

[0118] Multidimensional similarity calculations are performed on the local fingerprint set that has passed multi-chain verification, including texture similarity, color distribution similarity, and depth feature similarity.

[0119] Input a set of local fingerprints that have passed multi-chain collaborative verification. Each fingerprint contains a texture feature vector, a color distribution vector, and a depth feature vector. Calculate the texture similarity, color distribution similarity, and depth feature similarity between local fingerprints respectively. Combine the various similarities according to preset weights to form a comprehensive multidimensional similarity value matrix for each pair of local fingerprints.

[0120] The similarity calculation results are compared with the historical copyright sample database, and the existence of copyright overlap is determined based on the conflict detection rules.

[0121] Retrieve samples from the historical copyright sample database that are similar to the current local fingerprint features and extract the corresponding historical copyright information; compare the multidimensional similarity value with the feature vector of the historical samples to calculate the similarity matching degree; and preliminarily determine whether there is copyright overlap based on the conflict detection rules.

[0122] By combining a dynamic threshold adjustment strategy, the judgment threshold is adaptively corrected based on the scene complexity and historical sample density.

[0123] Based on the complexity of the current scene and the density of the historical sample library, an adaptive threshold correction coefficient is calculated to adjust the original similarity threshold by weight, resulting in a dynamic threshold that adapts to different image scenes and sample distributions. The adjusted threshold is then used for subsequent copyright conflict determination to ensure that the determination criteria match the actual feature distribution.

[0124] If the similarity result exceeds the threshold, the corresponding local element is marked as a candidate event for copyright conflict, and an event list is generated;

[0125] For local elements whose overall similarity value exceeds the dynamic threshold, they are marked as potential copyright conflict candidate events. The candidate event information is recorded, including the fragment UID, local fingerprint index, matching historical sample ID and similarity value. All marked events are organized into an event list and output for subsequent use in building a local copyright identification map.

[0126] S5: Define the local area corresponding to the copyright conflict candidate event as the copyright detection range, and integrate the local fingerprint nodes, cross-chain storage nodes and historical copyright nodes within this range to construct a local copyright identification map and output the digital copyright confirmation result.

[0127] The process of outputting the digital copyright confirmation result in S5 is as follows:

[0128] Based on the list of potential copyright conflict events, the relevant local areas are defined as the scope of copyright detection;

[0129] Input a list of copyright conflict candidate events. Each event includes a segment UID, a local fingerprint index, and a similarity value. Based on the spatial coordinates and image region information of the candidate events, mark the relevant local regions in the original image as the copyright detection range. Create an index table for the marked regions to record the corresponding segments, fingerprint nodes, and historical sample mapping relationships, providing basic data for map construction.

[0130] Within this scope, local fingerprint nodes, cross-chain storage nodes, and historical copyright nodes are integrated to construct a local copyright identification graph;

[0131] Within the scope of copyright detection, local fingerprint nodes, cross-chain storage nodes, and historical copyright nodes are mapped to a set of nodes in a graph. Edge connections are established between nodes based on feature similarity, historical copyright relationships, and storage node associations, with each edge containing weight information. The output is a local copyright identification graph, including a set of nodes, a set of edges, and its weight matrix, for subsequent analysis.

[0132] The node importance of the identification graph is ranked and the edge weights are calculated to identify key copyright nodes and conflict paths.

[0133] The nodes in the graph are ranked by importance, and the node centrality index is calculated based on the number of node connections, the cumulative value of edge weights, and the historical copyright coverage. The weights of the edges are analyzed to identify high-weight paths connecting important nodes and to mark paths that may have copyright conflicts. The information on key nodes and conflicting paths is compiled into an index table to provide a reference for generating the rights confirmation conclusion.

[0134] By combining the identification map and conflict determination results, an accurate digital copyright confirmation conclusion is generated and output to the confirmation database and user terminal;

[0135] By combining the graph structure and the candidate event list, the copyright ownership of each local element is determined, generating an accurate digital copyright confirmation conclusion. The confirmation conclusion is stored in the confirmation database according to the fragment UID, node index and copyright information, and is synchronously output to the user terminal interface. The generated conclusion record includes graph nodes, conflict paths, confirmation timestamps and historical reference information to ensure traceability and data integrity.

[0136] Example 2: Please refer toFigure 2 As shown, the digital copyright confirmation system based on multi-chain collaboration and dynamic verification includes:

[0137] Local fingerprint extraction module: Divides digital images into regions and extracts texture, color, and depth representations to generate local copyright feature sequences;

[0138] The fragmented storage management module: fragments the local copyright feature sequence and performs multi-replica interleaved storage and integrity verification on distributed nodes;

[0139] Multi-chain mapping verification module: Maps sharded data to multiple blockchains and combines cross-chain hashing and timestamp mechanisms to achieve collaborative verification;

[0140] Similarity detection module: performs multi-dimensional similarity calculation on the verified local fingerprint set, and identifies copyright conflict candidate events by combining rules and thresholds;

[0141] The copyright confirmation result generation module integrates local fingerprint nodes, cross-chain storage nodes, and historical copyright nodes to construct a copyright identification graph and output the copyright confirmation result.

[0142] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0143] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A digital copyright right confirmation method and system based on multi-chain cooperation and dynamic verification, characterized in that, The method comprises the following steps: The local area is divided by the digital image to be authenticated, and the corresponding local fingerprint is extracted based on the texture features, color distribution and depth embedding representation, and a fingerprint index set is generated by combining the content segmentation rules to obtain a local copyright feature sequence; The local copyright feature sequence is processed by fragmentation, and multiple copies are stored on distributed nodes, combined with node redundancy, access delay and fragmentation integrity checking mechanism, to construct a traceable distributed storage structure; Based on the distributed storage structure, the local fingerprint fragments are mapped to multiple blockchains, combined with cross-chain hash verification, on-chain timestamp and historical record backtracking, to form a multi-chain collaborative verification channel; The multi-chain verified local fingerprint set is calculated for multi-dimensional similarity, combined with the historical copyright sample library, conflict detection rules and dynamic threshold adjustment strategy, to identify potential local element copyright conflict candidate events; The local area corresponding to the copyright conflict candidate event is delimited as the copyright detection range, and the local fingerprint nodes, cross-chain storage nodes and historical copyright nodes in the range are fused to construct a local copyright identification map, and the digital copyright authentication result is output.

2. The method and system for digital copyright identification based on multi-chain coordination and dynamic verification according to claim 1, characterized in that, The process of generating a fingerprint index set combined with content segmentation rules is: Perform multi-layer semantic segmentation on the pixel distribution in the local area to extract significant edges, texture patterns and color gradient changes in the image; Match the segmentation results with the image depth embedding representation to generate corresponding multi-modal fingerprint feature labels; According to the semantic similarity and spatial adjacency relationship between the feature labels, a multi-level fingerprint index tree is established; The features of the local area are archived and numbered through the index tree to output the fingerprint index set. 3.The digital copyright right confirmation method and system based on multi-chain coordination and dynamic verification according to claim 2, characterized in that, The process of obtaining a local copyright feature sequence is: Expand the fingerprint index set according to the spatial coordinate order, and generate the context environment based on the timestamp label of the image; Generate a vectorized representation of the local features based on the matching relationship between the depth embedding features and the index labels; Use a serialization encoding mechanism to map the local feature vectors to a feature sequence space in order; Eliminate redundant or noisy features through an anomaly detection mechanism to finally obtain a structured local copyright feature sequence.

4. The method and system for digital copyright identification based on multi-chain coordination and dynamic verification according to claim 3, characterized in that, The process of processing the local copyright feature sequence by fragmentation is: According to the length and complexity of the local copyright feature sequence, use a dynamic sliding window segmentation mechanism to divide the sequence into multiple continuous fragments; Generate a unique identifier for each fragment and embed the source index and generation timestamp; Interleave and replicate each fragment between different distributed nodes through a multi-copy redundancy mechanism; Use a fragmentation integrity checking algorithm to perform hash signature and consistency verification on each fragment to output the final storable fragment set.

5. The method and system for digital copyright identification based on multi-chain coordination and dynamic verification according to claim 4, characterized in that, The process of constructing a traceable distributed storage structure is: Distribute the fragment set according to the identifier, and perform dynamic scheduling combined with the storage load of the nodes; Select the optimal storage node for each fragment based on the access delay model, and use a replica interleaving strategy to improve query efficiency; During storage, use a chain checking mechanism to record the hash value and timestamp of each storage operation; Through cross-node consistency comparison, a distributed storage structure with tamper resistance is formed.

6. The method and system for digital copyright protection based on multi-chain coordination and dynamic verification according to claim 5, characterized in that, The cross-chain hash verification, on-chain timestamp, and historical record backtracking process is: Calculate the hash digest of each shard in the storage structure and submit it to multiple independent blockchains; Add a timestamp to each hash digest on the chain as a time reference for shard verification; Conduct consistency comparison of hash records in different blockchains through cross-chain synchronization mechanism; Detect whether there are abnormal changes in shard data through historical on-chain record backtracking mechanism and form a verification report.

7. The method and system for digital copyright protection based on multi-chain coordination and dynamic verification according to claim 6, characterized in that, The multi-chain collaborative verification channel process is: Establish data interaction interfaces between multiple chains through cross-chain communication protocols; Embed cross-chain verification rules in the verification channel to synchronize and align hash results and timestamps from different chains; Use consensus mechanisms to rule on multi-chain verification results and output the final collaborative verification conclusion. 8.The digital copyright right confirmation method and system based on multi-chain coordination and dynamic verification according to claim 7, wherein, The process of identifying potential local element copyright conflict candidate events is: Perform multi-dimensional similarity calculation on the local fingerprint set that passes the multi-chain verification, including texture similarity, color distribution similarity, and depth feature similarity; Compare the similarity calculation results with the historical copyright sample library and determine whether there is copyright overlap based on conflict detection rules; Adjust the threshold value adaptively based on scene complexity and historical sample density using a dynamic threshold adjustment strategy; If the similarity result exceeds the threshold value, mark the corresponding local element as a copyright conflict candidate event and generate an event list.

9. The method and system for digital copyright protection based on multi-chain coordination and dynamic verification according to claim 8, characterized in that, The process of outputting digital copyright identification results is: According to the copyright conflict candidate event list, define the local area involved as the copyright detection range; Fuse local fingerprint nodes, cross-chain storage nodes, and historical copyright nodes within the range to construct a local copyright identification graph; Sort the nodes and calculate the edge weights of the identification graph to identify key copyright nodes and conflict paths; Combine the identification graph and conflict judgment results to generate accurate digital copyright identification conclusions and output them to the identification database and user terminal.

10. A digital copyright right-asserting system based on multi-chain coordination and dynamic verification, applied to the method of any one of claims 1-9, characterized in that, It includes: Local fingerprint extraction module: divide the digital image into regions and extract texture, color, and depth features to generate local copyright feature sequences; Shard storage management module: shard the local copyright feature sequences and perform multi-copy interleaved storage and integrity check on distributed nodes; Multi-chain mapping verification module: map the shard data to multiple blockchains and implement collaborative verification using cross-chain hash and timestamp mechanisms; Similarity detection module: calculate the multi-dimensional similarity of the verified local fingerprint set and identify copyright conflict candidate events based on rules and thresholds; Identification result generation module: fuse local fingerprint nodes, cross-chain storage nodes, and historical copyright nodes to construct a copyright identification graph and output the identification result.

Images