Modeling method, system and device based on dynamic 3D target and storage medium

Abstract

The invention relates to the technical field of three-dimensional modeling, in particular to a modeling method and system based on a dynamic 3D target, electronic equipment and a storage medium, and the method comprises the steps that a three-dimensional modeling operation instruction is generated, and the operation instruction at least comprises an operation type, a target object identifier and a space action range parameter; constructing an operation semantic tree based on the operation instruction; by comparing the local operation semantic tree with the received remote operation semantic tree, detecting a space action range overlapping region; when an overlapping region is detected, applying a space-time dimension priority merging rule; executing the combined operation instruction, and generating a binary log containing an operation semantic tree hash value; when an operation conflict is detected, local rollback is performed based on a differential CRC check value that computes only overlap region data blocks. According to the method, the problems of semantic understanding deficiency, logic inconsistency and low recovery efficiency during three-dimensional space operation are effectively solved.

Description

Technical Field

[0001] The present application relates to the technical field of three-dimensional modeling, and in particular to a modeling method, system, electronic device and storage medium based on dynamic 3D targets. Background Art

[0002] Dynamic 3D objects may involve real-time updates, multi-user collaboration, data synchronization, network optimization, etc. Improvements in communication logic may include data transmission protocols, state synchronization mechanisms, compression algorithms, distributed processing, etc.

[0003] In 3D modeling collaboration systems, data conflicts caused by concurrent multi-user operations have long been a problem. Traditional solutions, such as the Operation Transformation (OT) algorithm, rely on linear timestamp sequences for conflict resolution. However, these solutions suffer from inherent flaws when processing 3D spatial operations: a lack of semantic understanding (simplifying 3D spatial operations to text editing operations, failing to recognize the spatial semantic relevance of modeling operations like rotation and stretching); logical inconsistencies (merging data based solely on the timing of operations, resulting in spatial operations that violate physical laws, such as object penetration); and inefficient recovery (using a full log rollback mechanism, which generates up to GB of redundant data in complex modeling scenarios).

[0004] Therefore, there is an urgent need for a client that can effectively solve the problems of lack of semantic understanding, logical inconsistency and low recovery efficiency during three-dimensional space operations. Summary of the Invention

[0005] In order to solve at least one of the above technical problems, the present application provides a modeling method, system, electronic device and storage medium based on dynamic 3D targets.

[0006] In a first aspect, the present application provides a modeling method based on a dynamic 3D object, which adopts the following technical solutions: Generate a three-dimensional modeling operation instruction, wherein the operation instruction at least includes an operation type, a target object identifier, and a spatial scope parameter; An operation semantics tree is constructed based on the operation instruction, and the operation semantics tree node includes the following hierarchical structure: The three-dimensional spatial topological relationship between the parent node and the target object, Child node, the spatial scope weight value of the user operation; By comparing the local operation semantic tree with the received remote operation semantic tree, the spatial scope overlapping area is detected; When overlapping areas are detected, the spatiotemporal dimension priority merging rules are applied as follows: Time dimension priority condition: when the time difference between operations initiated is less than the threshold Δt, the spatial dimension judgment is activated; Spatial dimension priority condition, assigning priority coefficients according to the operation type, where the structure modification operation priority P1> geometric deformation operation priority P2> material editing operation priority P3; Execute the merged operation instructions and generate a binary log containing the hash value of the operation semantic tree; When an operation conflict is detected, a local rollback is performed based on a differential CRC check value calculated only for the overlapping area data blocks.

[0007] Through the above scheme, the operation semantic tree is constructed, and the spatial context relationship map of three-dimensional operations is constructed by utilizing parent node topology encoding, child node scope quantification, and semantic conflict detection. This enables the system to understand the semantic relevance of compound operations, and the semantic recognition accuracy is improved compared to traditional methods. At the same time, the spatiotemporal dimension priority merging rules are adopted, and the use of time condition triggering, physical constraint injection, and conflict resolution algorithms has greatly reduced the incidence of physical anomalies in specific scenarios compared to traditional OT algorithms, thereby improving the probability of correct merging in accordance with rigid body dynamics. In addition, through differential CRC local rollback, enhanced spatial locality utilization, data compression optimization, and the use of fast recovery protocols, the amount of data for a single rollback is reduced and the recovery time is shortened. In short, this scheme forms a real-time collaborative framework with topology perception capabilities by organically integrating three technical means: structuring three-dimensional spatial semantics, quantifying operation priorities, and localizing recovery mechanisms. This effectively solves the problems of lack of semantic understanding, logical inconsistency, and low recovery efficiency during three-dimensional spatial operations.

[0008] In one possible implementation, the step of constructing an operational semantics tree based on the operation instruction includes: The three-dimensional spatial topology of the parent node uses the half-edge data structure to record the connection relationship between vertices, edges and faces; The spatial scope weight value of the child node is calculated by the Gaussian decay function, and the action radius σ is positively correlated with the preset influence range of the operation type, where Structural modification operation σ = 20% of the diagonal length of the model bounding box; Geometric deformation operation σ = 150% of the radius of the operating tool; Material editing operation σ = 30% of the texture map UV unfolding area.

[0009] Through the above solution, the accuracy of scope recognition is improved in the building model test, effectively improving the geometric accuracy of semantic understanding.

[0010] In one possible implementation, when an overlapping area is detected, in the step of applying the spatiotemporal dimension priority merging rule, the priority coefficient is dynamically adjusted according to the model type, including: In the mechanical assembly model, P1:P2:P3=4:3:1; The P2 weight in the organic biological model increased by 30%; Added lighting operation priority P4=0.8×P1 in the building scene model; The adjustment is based on the classification label in the model file metadata.

[0011] Through the above solution, the correct merging rate in the parts assembly scenario is improved, and the logical consistency adaptation capability is enhanced.

[0012] In one possible implementation, when an operation conflict is detected, performing a local rollback based on a differential CRC check value includes: Locate the subtree nodes that need to be rolled back based on the operational semantics tree; Generate an index list V_list of the affected vertices, sorted by the spatial Morton code; Perform coordinate fallback calculations only for vertices in V_list.

[0013] Through the above solution, the amount of rollback data is further reduced, reducing the amount of recovery data.

[0014] In a possible implementation, before the step of detecting the spatial scope overlap region by comparing the local operational semantics tree with the received remote operational semantics tree is performed, a preloading step is further included: Predict the model area that may be modified based on the type of tool currently operated by the user; Pull the spatial topology data of the area from the server in advance; The cache data is organized in a quadtree structure, and the cache hit rate is over 85%.

[0015] Through the above solution, the standard deviation of operation response time is reduced, thereby reducing the impact of network delay.

[0016] In one possible implementation, when an overlapping area is detected, applying the spatiotemporal dimension priority merging rule step further includes: Collect historical operation conflict data to train the LSTM prediction model; Proactively restrict non-critical operations when a high probability of conflict (>70%) is detected; Dynamically adjust the Δt threshold: Δt = base value 50ms × (1-collision probability).

[0017] Through the above-mentioned plan, the proportion of proactive conflict avoidance is increased, the pressure of post-processing is reduced, and preventive conflict avoidance is achieved.

[0018] In one possible implementation, in the step of executing the merged operation instructions and generating a binary log including a hash value of the operation semantics tree, the binary log is compressed using differential encoding, including: Create a dictionary encoding for the operation type field, with 1 byte replacing the original 4-byte enumeration value; The spatial coordinates are stored using the Δ value relative to the previous operation; Use DEFLATE algorithm for streaming compression; The compressed log format is: |Header identifier (0xAE)|Compression flag|Dictionary encoding|Δ coordinate value|CRC check|.

[0019] Through the above solution, the log volume is further reduced, reducing storage and transmission overhead.

[0020] In a second aspect, the present application provides a modeling system based on a dynamic 3D target, comprising: An operation instruction generation module: configured to receive user input and generate a three-dimensional modeling operation instruction, wherein the instruction at least includes an operation type field, a target object unique identifier, and a three-dimensional space scope parameter set; Semantic tree construction module: connected with the operation instruction generation module data, including: Topology relationship analysis unit: extracts the three-dimensional spatial topology relationship of the target object based on the half-edge data structure and generates parent node data; Scope calculation unit: calculates the weight distribution matrix of child nodes according to the spatial impact algorithm corresponding to the operation type; Conflict detection module: bidirectionally connected with local storage module and network communication module, including, Semantic tree comparer: performs spatial scope overlap detection between the local operational semantic tree and the received remote semantic tree; Priority arbiter: Built-in time and space dimension judgment logic circuit, triggering the arbitration process when overlap is detected; Operation merging execution module: connected to the output end of the conflict detection module, including: Rule base: stores a table of temporal and spatial priority coefficients, where structural modification P1, geometric deformation P2, and material editing P3 satisfy P1>P2>P3; Differential calculation unit: generates an operation increment package containing only the changes of vertices in the overlapping area; Log management module: communicates with all modules, including, Binary encoder: converts the operation record into a compressed log containing the hash value of the semantic tree; Local rollback controller: In response to a conflict signal, it performs coordinate rollback of the affected vertices based on the differential CRC checksum. Network communication module: adopts dual-channel design, including: Operation instruction transmission channel: encapsulates semantic tree data packets; Log synchronization channel: transmits compressed binary difference logs; The semantic tree construction module, the conflict detection module, and the operation merging execution module are interconnected at the hardware layer through a shared memory bus to form a real-time processing pipeline architecture.

[0021] In a third aspect, the present application provides an electronic device comprising a memory and a processor, wherein the memory is used to store computer program code, and the processor is used to execute the computer program code stored in the memory to implement the method in the above-mentioned first aspect and any one of the first aspect, or the above-mentioned second aspect and any possible implementation of the second aspect.

[0022] In a fourth aspect, the present application provides a computer-readable storage medium, which stores a computer program or instruction. When the computer program or instruction is executed, it implements the method in the above-mentioned first aspect and any one of the first aspect, or the above-mentioned second aspect and any possible implementation of the second aspect.

[0023] In summary, this application includes at least one of the following beneficial technical effects: 1. Improved conflict resolution accuracy. By operating on semantic trees to achieve spatial semantic parsing, the conflict misjudgment rate is effectively reduced compared to traditional methods. Simultaneously, the use of spatiotemporal priority rules greatly increases the probability of correct merging that complies with physical laws. 2. System performance optimization: The partial rollback mechanism reduces the amount of restored data to half that of a full rollback. The binary log compression rate is significantly improved compared to the JSON format, reducing storage overhead. 3. Real-time performance is guaranteed. The end-side processing delay is effectively controlled. Operational consistency can be maintained even in network fluctuation scenarios (RTT 100-500ms), and the user-perceived lag rate is significantly reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 A flowchart of a modeling method based on a dynamic 3D target provided in an embodiment of the present application.

[0025] Figure 2 A structural diagram of a dynamic 3D target-based modeling system provided in an embodiment of the present application.

[0026] Figure 3 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0027] The technical solutions of this application will be described below in conjunction with all the accompanying drawings. Obviously, the embodiments described are only part of the embodiments of this application, not all of them.

[0028] In the description of the embodiments of this application, unless otherwise specified, " / " means "or." For example, A / B can mean A or B. "And / or" in this article is merely a description of an association relationship between associated objects, indicating that three relationships can exist. For example, A and / or B can mean: A exists alone, A and B exist simultaneously, and B exists alone. In addition, in the description of the embodiments of this application, "plurality" or "a plurality" means two or more than two.

[0029] In the following, the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of the technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of this embodiment, unless otherwise specified, "plurality" means two or more.

[0030] The terms used in the following examples are only for the purpose of describing specific embodiments and are not intended to limit the present application. As used in the specification and appended claims of this application, the singular expressions "a," "an," "said," "above," "the," and "this" are intended to include expressions such as "one or more," unless the context clearly indicates otherwise. It should also be understood that in the following examples of this application, "at least one," "one or more" refer to one, two, or more than two.

[0031] References to "one embodiment" or "some embodiments" in this specification mean that a particular feature, structure, or characteristic described in conjunction with that embodiment is included in one or more embodiments of the present application. Thus, phrases such as "one embodiment," "some embodiments," "another embodiment," and "other embodiments" appearing in various places in this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically stated. The terms "including," "comprising," "having," and variations thereof mean "including but not limited to," unless otherwise specifically stated.

[0032] The embodiment of the present application provides a method for modeling a dynamic 3D object, which is performed by an electronic device, wherein the electronic device can be an independent physical electronic device, or an electronic device cluster or distributed system composed of multiple physical electronic devices, or a cloud electronic device that provides cloud computing services. The embodiment of the present application is not limited here, such as Figure 1 As shown, the method includes: S1. Generate 3D modeling operation instructions.

[0033] Specifically, the operation instruction at least includes an operation type, a target object identifier, and a spatial scope parameter.

[0034] Furthermore, the operation type is defined as three types of basic operation codes preset by the system, for example: 0x01 = structure modification, such as cutting and drilling, corresponding to high-priority operations; 0x02 = geometric deformation, such as stretching and bending, corresponding to medium-priority operations; 0x03 = material editing, such as mapping and coloring, corresponding to low-priority operations.

[0035] The operation type is automatically mapped by tool type, for example, selecting "Cutting Tool" maps it to "Structure Modification".

[0036] Furthermore, the target object identifier uses a universally unique identifier (UUIDv4) to mark the model components, and a globally unique identification code is generated for each model component using a 36-bit mixed code (including numbers, letters and connectors).

[0037] The identification code includes the generation timestamp, device fingerprint and random entropy value to ensure global uniqueness.

[0038] Furthermore, the spatial scope parameters are collected by accurately locating the coordinates (x, y, z) of the operation center point through a three-dimensional picking algorithm, and automatically calculating the effective radius according to the tool properties, such as the brush radius and stretching range.

[0039] Among them, the effective radius of structural tools (such as drill bits) = physical diameter × 1.2 times the safety factor; the effective radius of surface treatment tools (such as spray guns) = preset parameters × operating pressure value.

[0040] In an embodiment of the present application, user operations are converted into standardized instructions that can be understood by the machine, and each operation is given a unique identity and spatial positioning information, just like creating a "digital ID card" for the modeling operation, thereby achieving a significant improvement in the efficiency of instruction generation. In a standard workstation environment, the time required to generate an instruction for a single operation does not exceed 3 milliseconds, and the time required to generate an instruction is significantly reduced compared to traditional solutions.

[0041] S2. Construct an operation semantic tree based on the operation instructions.

[0042] Specifically, the operation semantics tree node contains the following hierarchical structure: parent node, the three-dimensional spatial topological relationship of the target object; child node, the spatial scope weight value of the user operation.

[0043] Furthermore, the implementation steps of parent node construction (topological relationship) are: The 3D mesh topology data of the target component is extracted from the model database. The spatial connection relationship is recorded using a "vertex-half-edge-face" circular linked list structure. A local topology subgraph is constructed, and only elements within 1.5 times the operation influence radius are retained.

[0044] Furthermore, the implementation steps of child node calculation (scope weight) are: The influence of each vertex is calculated based on the distance attenuation principle. The closer the vertex is to the center of the operation, the higher the weight. The attenuation curve follows a Gaussian distribution, with weight values ranging from [0, 1].

[0045] Dynamically adjust the decay rate parameters to suit different operating characteristics. Structural modification uses fast decay (affecting a concentrated area), while material editing uses slow decay (affecting a diffuse area).

[0046] In an embodiment of the present application, the impact range and model structure of the operation are encoded into a "spatial relationship map" so that the computer can understand the spatial semantics of the operation.

[0047] S3. Detect the spatial scope overlap area by comparing the local operation semantic tree with the received remote operation semantic tree.

[0048] Specifically, the implementation steps include fast spatial index construction and overlap determination process.

[0049] Furthermore, fast spatial index construction involves creating a 3D axial bounding box (AABB) for each semantic tree, using an octree structure to accelerate spatial queries, and dividing the 3D space into a hierarchical grid.

[0050] Furthermore, the overlap determination process includes: Phase 1: Coarse detection - comparing bounding box intersections (time < 1ms); Phase 2: Precision detection - calculating the weighted heat map overlap rate (5-15ms); Judgment threshold setting: Conflict processing is triggered when the proportion of overlapping vertices exceeds 70%.

[0051] In an embodiment of the present application, the operation conflict area is identified by spatial relationship comparison, and a collision detection mechanism similar to radar scanning is implemented to realize a multi-level cache mechanism, so that the cache reuse rate of the most recent comparison result is improved and the repeated calculation overhead is greatly reduced.

[0052] S4. When overlapping areas are detected, the spatiotemporal dimension priority merging rule is applied.

[0053] Specifically, the priority merging rules for the time and space dimensions are as follows: The time dimension priority condition is that when the time difference between the operation initiation and the operation is less than the threshold Δt, the spatial dimension judgment is activated.

[0054] The spatial dimension priority condition assigns priority coefficients according to the operation type, where the structure modification operation priority P1>geometry deformation operation priority P2>material editing operation priority P3.

[0055] Furthermore, the time dimension determination step involves setting a dynamic time window threshold Δt, which is adaptively adjusted based on network latency. Specifically, Δt = 50 milliseconds in a 5G network environment and Δt = 100 milliseconds in a standard broadband environment. When the operation time difference is less than Δt, spatial priority arbitration is activated.

[0056] Furthermore, the spatial priority arbitration step sets the basic priority coefficients: "Structural Modification" with a priority coefficient of 3 and a rigid constraint strength; "Geometry Deformation" with a priority coefficient of 2 and an elastic constraint strength; and "Material Editing" with a priority coefficient of 1 and an unconstrained constraint strength. Scene adaptability is enhanced: the structural modification coefficient in mechanical assembly scenarios is multiplied by 1.5, and the geometric deformation coefficient in character modeling scenarios is multiplied by 1.2.

[0057] In the embodiment of the present application, a multi-dimensional decision-making model is established to intelligently coordinate operation instructions with time and space conflicts, forming a two-level arbitration mechanism. First, operations with significant time differences are excluded, and then physical rationality assessments are performed on operations that are close in time and space to ensure the integrity of the model structure.

[0058] S5. Execute the merged operation instructions and generate a binary log containing the hash value of the operation semantic tree.

[0059] Specifically, the implementation steps for generating binary logs are to design log structures and storage optimization strategies.

[0060] Furthermore, log structure design includes, Header identifier: 2-byte magic number (0xAE01 identifies the operation log); Semantic tree fingerprint: 32-byte SHA-256 hash value; Operation parameter segment: variable-length bytes store coordinate changes; Check code: 4-byte CRC32 cyclic redundancy check.

[0061] Furthermore, the storage optimization strategy includes, Incremental record: only stores the difference value (Δ value) of vertex coordinates; Compression coding: using streaming compression algorithm, real-time compression rate ≥ 80%; Block storage: Every 1024 operation records are packaged into a data block.

[0062] In the embodiments of the present application, a log format with efficient access is designed to achieve full life cycle management of operation history, achieve a log write speed of 120,000 logs per second, and a stable read latency within 5 milliseconds.

[0063] S6. When an operation conflict is detected, perform a local rollback based on the differential CRC check value.

[0064] Specifically, the differential CRC value is calculated only for the overlapping area data blocks.

[0065] Specifically, the implementation steps of the local rollback mechanism include locating the affected area, calculating the difference value, and performing data recovery.

[0066] Furthermore, the impact area location step includes tracing back the impact path through the semantic tree to generate an affected vertex index list sorted by the spatial Morton code.

[0067] Furthermore, the difference value calculation step includes recording a vertex coordinate snapshot before the rollback, calculating a cyclic redundancy check code (CRC) of the coordinate value, and generating a difference check code.

[0068] Furthermore, the data recovery execution step includes locating the vertices to be recovered according to the index list, applying the ΔCRC difference to restore the original coordinate values, and locally redrawing the model to avoid global refresh.

[0069] In the embodiment of the present application, data recovery is achieved with pinnacle-level accuracy, the impact of rollback is minimized, and a double-check mode is adopted to ensure that the data recovery accuracy reaches 99.9999%.

[0070] In some embodiments, step S2 further includes: The three-dimensional spatial topological relationship of the parent node uses the half-edge data structure to record the connection relationship between vertex-edge-face.

[0071] Specifically, the half-edge data structure in computer graphics is used to record the topological connection relationship of the model mesh. Its core elements include: "vertices" store three-dimensional coordinates (x, y, z); "half-edges" record the connection relationship between adjacent facets, forming a bidirectional circular linked list; "faces" are closed polygons composed of half-edges connected at the head and tail.

[0072] Furthermore, the implementation steps for extracting the topological subgraph are to calculate the model bounding sphere with the operation center as the origin, extract the local topological structure with a radius covering 1.5 times the operation influence range, and only retain the vertex, edge, and face connection relationships within this range when constructing the parent node data.

[0073] It also includes that the spatial scope weight value of the child node is calculated through a Gaussian decay function, and the action radius σ is positively correlated with the preset influence range of the operation type.

[0074] Specifically, the weight calculation formula of the Gaussian attenuation function is:

[0075] Where d is the Euclidean distance from the vertex to the operation center, and the action radius σ is dynamically adjusted according to the operation type.

[0076] Specifically, the structural modification operation σ = 20% of the diagonal length of the model bounding box. Structural changes must cover the entire component. For example, when cutting an engine block, the cross-section must be complete.

[0077] The geometric deformation operation σ = 150% of the radius of the operating tool. Considering the physical diffusion effect of the tool force, for example, the impact range of a 10cm diameter carving knife extends to 15cm.

[0078] The material editing operation σ = 30% of the texture map UV expansion area. Material operations are expanded in UV space and need to match the texture distribution density.

[0079] In this application, the dual design of structured data encoding and spatial influence quantification transforms abstract modeling operations into a precisely computable spatial semantic relationship network, creating a "spatial DNA map" for the modeling process. This significantly reduces memory usage compared to traditional adjacency matrices and speeds up topology queries.

[0080] In some embodiments, in step S4, dynamically adjusting the priority coefficient according to the model type includes: In mechanical assembly models, set the priority ratio P1:P2:P3 to 4:3:1. Structural modifications (such as tolerance adjustments) directly impact assembly functionality and therefore require the highest priority. For example, in transmission model testing, this effectively improves the accuracy of gear mesh conflict resolution.

[0081] In organic biological models, the P2 weight has been increased by 30%, for example, from an original P2 of 2 to an adjusted P2 of 2.6. Biomorphic modeling centers around surface deformation operations (such as muscle definition). For example, in human body modeling, this effectively improves the surface smoothness compliance rate.

[0082] In the architectural scene model, the lighting operation priority has been increased to P4 = 0.8 × P1. Lighting effects are strongly coupled with spatial structure (for example, window position affects daylighting). For example, if a user simultaneously modifies a wall structure (P1) and adjusts a spotlight angle (P4), the system prioritizes the wall modification while retaining 80% of the lighting operation weight.

[0083] Furthermore, the adjustment is based on the classification labels in the model file metadata.

[0084] Specifically, the implementation steps of model type identification are to parse the classification label from the model file metadata.

[0085] Among them, model types support three basic categories: mechanical assemblies, which are industrial components with precise structural matching relationships (such as engines and gearboxes); organic biological models, which are biological forms with complex surfaces (such as human organs and animal bodies); and architectural scenes, which are spatial structures that include lighting systems (such as interior design and urban planning).

[0086] In an embodiment of the present application, a scenario-adaptive priority model is established based on the physical characteristics and operational requirements of different modeling scenarios, so that the merging rules can intelligently adapt to the model type.

[0087] In some embodiments, in step S6, the local rollback includes: Locate the subtree node that needs to be rolled back based on the operational semantics tree.

[0088] Specifically, a conflict impact analysis is first performed, such as obtaining the spatial scope parameters of the overlapping area from the conflict detection engine, and matching the corresponding subtree nodes in the operational semantic tree, whose range covers 1.2 times the conflict area.

[0089] Furthermore, the subtree retrieval mechanism uses a depth-first search (DFS) to traverse the semantic tree nodes and locates the target subtree using a fast bounding box matching algorithm. The subtree bounding box center is ≤σ (operation impact radius) × 0.8 from the center of the conflicting region, keeping the retrieval time within 5ms.

[0090] The method also includes generating an index list V_list of the affected vertices, which is sorted by the spatial Morton code.

[0091] Specifically, the Morton code encoding principle is to convert the three-dimensional coordinates (x, y, z) into a linear one-dimensional code, where the encoding rule is to alternately take the binary bits of each coordinate axis to form a Z-shaped space-filling curve.

[0092] Specifically, the vertex list generation steps involve extracting the coordinates of all vertices under a subtree node, calculating the 64-bit Morton code for each vertex, sorting the vertices in ascending order by the Morton code, and generating the index list V_list. Vertex data is stored contiguously in memory based on spatial proximity, reducing the cache miss rate during coordinate rollback compared to traditional random access.

[0093] Also included is that coordinate fallback calculations are performed only for vertices in V_list.

[0094] Specifically, the fallback calculation process is to traverse each vertex index in V_list, read the original coordinates of the vertex from the historical snapshot, and apply a coordinate interpolation algorithm to smoothly transition to the historical state. The new coordinate = current coordinate × (1-α) + historical coordinate × α, where α gradually changes from 0 to 1. The transition animation takes 20ms to complete.

[0095] Furthermore, for historical version management, before each operation instruction is executed, a coordinate snapshot of the affected vertices is automatically saved. The snapshot data is stored differentially, and only the changes (Δx, Δy, Δz) are recorded.

[0096] Furthermore, for data consistency verification, after the rollback is completed, the CRC32 check value of the local area is calculated and compared with the benchmark check code stored on the server side, and the error rate is <0.001%.

[0097] In this application, a precise spatial backtracking algorithm is designed to address operational conflicts in 3D modeling, enabling "targeted repair" by restoring only the affected area and avoiding global data resets. This approach is similar to replacing faulty parts in complex machinery rather than disassembling the entire machine. Compared to a full tree traversal, this significantly increases positioning speed, optimizes memory access locality, and improves cache hit rates.

[0098] In some embodiments, before step S3 is executed, a preloading step is also included, specifically as follows: S301: predicting a model area that may be modified based on the type of tool currently operated by the user.

[0099] Specifically, a library of predefined tool types is provided. Each tool type is associated with a spatial influence mode. For example, a structural cutter predicts a cylindrical area with an influence radius coefficient of 2.0 × the tool length; a surface engraving brush predicts a spherical area with an influence radius coefficient of 1.5 × the brush radius; and a material spray gun predicts a conical diffusion area with an influence radius coefficient of 3.0 × the nozzle diameter.

[0100] Specifically, the dynamic prediction model monitors the tool's motion trajectory and operating pressure in real time. Using a Kalman filter, the model predicts the likely operating area for the next five frames (approximately 80ms). Inputs include the tool's velocity vector and the force intensity value, and outputs a three-dimensional bounding box (AABB) for the predicted area.

[0101] S302: Pull the spatial topology data of the area from the server in advance.

[0102] Specifically, cache data is organized in a quadtree structure, and the cache hit rate reaches over 85%.

[0103] Furthermore, the spatial partitioning rule of the quadtree structure is to project the model onto a two-dimensional plane (XZ axis) at the top level, dividing it into four quadrants. Each quadrant is recursively subdivided until the number of vertices in a node is ≤ 500. Its caching strategy is to retain frequently accessed nodes in memory and compress less frequently accessed nodes to the SSD cache. A Morton code index is also established to achieve O(logn) query complexity.

[0104] Specifically, the data request only requests a subset of topological data within the prediction area, including vertex coordinates and connection relationships, material UV mapping information, and physical property parameters (such as stiffness and density).

[0105] Specifically, the data prefetching process is: When the user selects the engraving pen, the future impact spherical area (radius R = 15cm) is predicted, and the quadtree node data of this area (node levels L3-L5) is requested from the server. The cached data packet size is approximately 2.8MB.

[0106] When the user switches to the structural cutter, the low-priority cache (such as material data) is dynamically released and the new predicted columnar area data (priority weight + 30%) is loaded.

[0107] In an embodiment of the present application, a forward-looking data pre-fetching system is established to intelligently predict user operation intentions and load necessary data for potential modification areas in advance, just like equipping the modeling process with a "spatial navigation radar", significantly reducing the impact of network latency on real-time collaboration.

[0108] In some embodiments, step S4 further includes: S401. Collect historical operation conflict data to train an LSTM prediction model.

[0109] Specifically, the collection and preprocessing of historical operation conflict data involves collecting six months of historical operation logs. Key fields include operation type, spatiotemporal coordinates, tool parameters, conflict occurrence markers, resolution time, and impact range. A three-dimensional spatiotemporal feature vector is then constructed. The temporal dimension includes the sequence of operation intervals (the time difference between the last ten operations); the spatial dimension includes a curve showing the overlap rate of the operation areas; and the semantic dimension includes a correlation score for the combination of operation types.

[0110] Specifically, the LSTM prediction model architecture utilizes a two-layer LSTM network (Long Short-Term Memory). The input layer consists of a 128-dimensional feature vector (including spatiotemporal semantic features); the hidden layer consists of 256 LSTM units with a dropout rate of 0.3; and the output layer uses a sigmoid activation function to output conflict probabilities (0–1). The training strategy employed is a weighted cross-entropy loss function (with a 3x weighting for high-conflict samples); and the optimizer is AdamW (with a learning rate of 1e-4 and a weight decay of 1e-5).

[0111] Specifically, the model's online learning mechanism is to deploy edge computing nodes to collect new conflict cases in real time, perform incremental training every 24 hours, and maintain the model's prediction accuracy >92%.

[0112] S402: Actively restrict non-critical operations when a high conflict probability (>70%) is detected.

[0113] Specifically, the operation classification strategy is divided into critical operations, non-critical operations, and normal operations. The network bandwidth is allocated in a ratio of 7:2:1 (critical: normal: non-critical).

[0114] Among them, key operations include structural modification and physical parameter adjustment (priority P1); non-key operations include view rotation, annotation addition, and material preview (priority P3); and common operations include geometric deformation and lighting adjustment (priority P2).

[0115] Specifically, when the LSTM predicts a conflict probability greater than 70%, non-critical operations are delayed (queue waiting), the non-critical operation channel is slowed down to 10Mbps, and the interface displays a "High Conflict Risk - Some Operations Restricted" warning icon.

[0116] S403 , dynamically adjust the Δt threshold: Δt=base value 50ms×(1-collision probability).

[0117] Specifically, the basic value is initially set to Δt=50ms (corresponding to the median RTT of the 5G network), which is dynamically calibrated according to the network type: in the 5G network environment, the basic Δt=50ms, and the adjustment coefficient is 1.0; in the Wi-Fi6 network environment, the basic Δt=65ms, and the adjustment coefficient is 0.8; in the 4G network environment, the basic Δt=100ms, and the adjustment coefficient is 0.5.

[0118] Specifically, the dynamic calculation formula for conflict probability responsive adjustment is Δt_dynamic = Δt_base × (1 - conflict probability). For example, when the conflict probability is 20%, Δt = 50 × 0.8 = 40ms (accelerated arbitration); when the conflict probability is 80%, Δt = 50 × 0.2 = 10ms (strict arbitration).

[0119] Furthermore, a boundary protection mechanism is established, setting the minimum Δt value to 10ms (to avoid oversensitivity). When Δt is less than 20ms, network quality detection is started. If the packet loss rate is greater than 3%, the Δt adjustment is frozen and an alarm is triggered.

[0120] In the embodiment of the present application, a conflict management framework with "prophetic capabilities" is constructed through a machine learning-driven conflict prediction system and adaptive time window adjustment technology, transforming traditional passive response into active defense, just like installing "smart traffic lights" for the collaborative modeling system.

[0121] In some embodiments, in step S5, the binary log is compressed using differential encoding, specifically including: Create a dictionary encoding for the operation type field, with 1 byte replacing the original 4-byte enumeration value.

[0122] Specifically, the dictionary library is constructed by predefining 128 common operation types (such as cutting, stretching, mapping, etc.), each of which is assigned a unique 1-byte code.

[0123] Furthermore, under the structure cutting operation type, the code is 0x01, and an example scenario is an engine cylinder hole opening; under the surface stretching operation type, the code is 0x02, and an example scenario is a car fender shaping; under the material replacement operation type, the code is 0x03, and an example scenario is a building exterior wall paint replacement.

[0124] At the same time, 32 custom coding bits (0x80-0x9F) are reserved. When a new operation type is detected, an extended code is automatically assigned and synchronized to all clients.

[0125] It also includes that the spatial coordinates are stored using a delta value relative to the previous operation.

[0126] Specifically, the relative coordinate calculation step is to record the difference (Δx, Δy, Δz) between the current operation coordinates and the previous operation, and the absolute coordinates of the first operation are stored as the reference value.

[0127] For example, if the previous coordinates are (1.2, 3.4, -0.5) and the current coordinates are (1.3, 3.35, -0.48), the Δ value stored is (+0.1, -0.05, +0.02).

[0128] Furthermore, when the Δ value exceeds the type range, an absolute coordinate reference point is inserted and a precision mode upgrade (such as switching from int16 to float32) is automatically triggered.

[0129] Also included is streaming compression using the DEFLATE algorithm.

[0130] Specifically, the compressed log format is: |header identifier (0xAE) | compression flag | dictionary encoding | Δ coordinate value | CRC check |.

[0131] The header identifier (0xAE) is a magic number that identifies the log type and assists in quick retrieval. It also reserves extended identifier bits (version number, encryption flag, etc.).

[0132] Among them, the bits in the compression flag are defined as: [7]: compression (0-native / 1-compressed); [6:4]: compression algorithm (000-DEFLATE); [3:0]: reserved bits.

[0133] Among them, CRC check uses CRC32 algorithm to calculate data integrity, and verifies the range through dictionary coding + Δ coordinate value field. At the same time, the error detection rate is 100% (single-bit error).

[0134] Specifically, the compression algorithm chosen is the DEFLATE algorithm (LZ77+Huffman coding), the data stream is processed in blocks, a 32KB sliding window is set, and repeated patterns within a length of 256 bytes are matched.

[0135] Specifically, the real-time compression process steps are as follows: input: original byte stream of dictionary encoding + Δ coordinates; block: every 1024 operation records are a compression unit; output: compressed data stream (compression ratio typical value 5:1).

[0136] In the embodiment of the present application, efficient storage and transmission of three-dimensional operation logs are achieved through multi-layer compression strategies and incremental storage technology, which is like equipping massive modeling data with "smart compression clothing", greatly reducing resource consumption while ensuring data integrity.

[0137] The following introduces a modeling system based on a dynamic 3D target provided in an embodiment of the present application. The modeling system based on a dynamic 3D target described below and the modeling method based on a dynamic 3D target described above can refer to each other.

[0138] refer to Figure 2 , a dynamic 3D target-based modeling system includes: Operation instruction generation module 1: configured to receive user input and generate three-dimensional modeling operation instructions, the instructions at least containing an operation type field, a target object unique identifier and a three-dimensional space scope parameter set.

[0139] Semantic tree construction module 2: data connection with operation instruction generation module 1, including: Topology relationship analysis unit: extracts the three-dimensional spatial topology relationship of the target object based on the half-edge data structure and generates parent node data; Scope calculation unit: Calculates the weight distribution matrix of child nodes based on the spatial impact algorithm corresponding to the operation type.

[0140] Conflict detection module 3: bidirectionally connected with local storage module and network communication module 6, including: Semantic tree comparer: performs spatial scope overlap detection between the local operational semantic tree and the received remote semantic tree; Priority arbiter: Built-in time and space dimension judgment logic circuit, triggering the arbitration process when overlap is detected.

[0141] Operation merging execution module 4: connected to the output end of conflict detection module 3, including: Rule base: stores a table of temporal and spatial priority coefficients, where structural modification P1, geometric deformation P2, and material editing P3 satisfy P1>P2>P3; Differential calculation unit: Generates an operation increment package containing only the changes in vertices in the overlapping area.

[0142] Log management module 5: communicates with all modules, including, Binary encoder: converts the operation record into a compressed log containing the hash value of the semantic tree; Local rollback controller: In response to a conflict signal, it performs coordinate rollback of the affected vertices based on the differential CRC checksum.

[0143] Network communication module 6: adopts dual-channel design, including: Operation instruction transmission channel: encapsulates semantic tree data packets; Log synchronization channel: transmits compressed binary difference logs.

[0144] Local storage module: integrated NVMe SSD and persistent memory, including, Operation cache: LRU algorithm manages the most recent 1000 operation instructions; Semantic tree snapshot library: versioned storage of topology data compressed in CSR format; Log buffer: A circular queue that stores the history of encryption operations.

[0145] Among them, the semantic tree construction module 2, the conflict detection module 3, and the operation merging execution module 4 are interconnected through a shared memory bus at the hardware layer to form a real-time processing pipeline architecture.

[0146] In this embodiment, the modeling system utilizes a modular, heterogeneous architecture, breaking down the core logic of 3D modeling into six functionally decoupled modules. By integrating hardware-level acceleration with intelligent algorithms, it enables real-time collaborative modeling of dynamic 3D objects. Each module is interconnected via a high-speed data bus, forming a closed-loop "perception-decision-execution" workflow.

[0147] The present application embodiment provides an electronic device, such as Figure 3 As shown, Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. Figure 3 The electronic device 300 shown includes a processor 301 and a memory 303. The processor 301 and the memory 303 are connected, for example, via a bus 302. Optionally, the electronic device 300 may further include a transceiver 304. It should be noted that in actual applications, the number of transceivers 304 is not limited to one, and the structure of the electronic device 300 does not constitute a limitation on the embodiments of the present application.

[0148] Processor 301 may be a CPU (Central Processing Unit), a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It may implement or execute the various exemplary logic blocks, modules, and circuits described in conjunction with the embodiments disclosed herein. Processor 301 may also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, and the like.

[0149] The bus 302 may include a path for transmitting information between the above components. The bus 302 may be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus. The bus 302 may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 3 Only one thick line is used in the diagram, but this does not mean that there is only one bus or one type of bus.

[0150] The memory 303 may be a ROM (Read Only Memory) or other types of static storage devices that can store static information and instructions, a RAM (Random Access Memory) or other types of dynamic storage devices that can store information and instructions, or an EEPROM (Electrically Erasable Programmable Read Only Memory), a CD-ROM (Compact Disc Read Only Memory) or other optical disk storage, optical disk storage (including compact discs, laser discs, optical discs, digital versatile discs, Blu-ray discs, etc.), a magnetic disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto.

[0151] The memory 303 is used to store application code for executing the solution of the embodiment of the present application, and the execution is controlled by the processor 301. The processor 301 is used to execute the application code stored in the memory 303 to implement the content shown in the above method embodiment.

[0152] Among them, electronic devices include but are not limited to: mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 3 The electronic device shown is merely an example and should not limit the functions and scope of use of the embodiments of the present application.

[0153] An embodiment of the present application provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the above-mentioned modeling method based on dynamic 3D targets are implemented.

[0154] Since the embodiments of the computer-readable storage medium part and the embodiments of the method part correspond to each other, the embodiments of the computer-readable storage medium part refer to the description of the embodiments of the method part.

[0155] It should be understood that although the steps in the flowcharts of the accompanying drawings are shown in sequence as indicated by the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some of the steps in the flowcharts of the accompanying drawings may include multiple sub-steps or multiple stages, and these sub-steps or stages are not necessarily executed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be executed in turn or alternately with other steps or at least a portion of the sub-steps or stages of other steps.

[0156] The above are only some of the implementation methods of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A modeling method based on dynamic 3D objects, characterized in that: include: Generate a three-dimensional modeling operation instruction, wherein the operation instruction at least includes an operation type, a target object identifier, and a spatial scope parameter; An operation semantics tree is constructed based on the operation instruction, and the operation semantics tree node includes the following hierarchical structure: The three-dimensional spatial topological relationship between the parent node and the target object, Child node, the spatial scope weight value of the user operation; By comparing the local operation semantic tree with the received remote operation semantic tree, the spatial scope overlapping area is detected; When overlapping areas are detected, the spatiotemporal dimension priority merging rules are applied as follows: Time dimension priority condition: when the time difference between operations initiated is less than the threshold Δt, the spatial dimension judgment is activated; Spatial dimension priority condition, assigning priority coefficients according to the operation type, where the structure modification operation priority P1> geometric deformation operation priority P2> material editing operation priority P3; Execute the merged operation instructions and generate a binary log containing the hash value of the operation semantic tree; When an operation conflict is detected, a local rollback is performed based on a differential CRC check value calculated only for the overlapping area data blocks.

2. The method according to claim 1, characterized in that The step of constructing an operational semantic tree based on the operation instruction includes: The three-dimensional spatial topology of the parent node uses the half-edge data structure to record the connection relationship between vertices, edges and faces; The spatial scope weight value of the child node is calculated by the Gaussian decay function, and the action radius σ is positively correlated with the preset influence range of the operation type, where Structural modification operation σ = 20% of the diagonal length of the model bounding box; Geometric deformation operation σ = 150% of the radius of the operating tool; Material editing operation σ = 30% of the texture map UV unfolding area.

3. The method according to claim 1, characterized in that When an overlapping area is detected, in the step of applying the spatiotemporal dimension priority merging rule, the priority coefficient is dynamically adjusted according to the model type, including: In the mechanical assembly model, P1:P2:P3=4:3:1; The P2 weight in the organic biological model increased by 30%; Added lighting operation priority P4=0.8×P1 in the building scene model; The adjustment is based on the classification label in the model file metadata.

4. The method according to claim 1, wherein When an operation conflict is detected, performing a local rollback based on the differential CRC check value includes: Locate the subtree nodes that need to be rolled back based on the operational semantics tree; Generate an index list V_list of the affected vertices, sorted by the spatial Morton code; Perform coordinate fallback calculations only for vertices in V_list.

5. The method according to claim 1, wherein Before the step of detecting the spatial scope overlap area by comparing the local operational semantics tree with the received remote operational semantics tree is performed, a preloading step is also included: Predict the model area that may be modified based on the type of tool currently operated by the user; Pull the spatial topology data of the area from the server in advance; The cache data is organized in a quadtree structure, and the cache hit rate is over 85%.

6. The method according to claim 1, characterized in that The step of applying the spatiotemporal dimension priority merging rule when an overlapping area is detected further includes: Collect historical operation conflict data to train the LSTM prediction model; Proactively restrict non-critical operations when a high probability of conflict (>70%) is detected; Dynamically adjust the Δt threshold: Δt = base value 50ms × (1-collision probability).

7. The method according to claim 1, characterized in that In the step of executing the merged operation instructions and generating a binary log containing a hash value of the operation semantics tree, the binary log is compressed using differential encoding, including: Create a dictionary encoding for the operation type field, with 1 byte replacing the original 4-byte enumeration value; The spatial coordinates are stored using the Δ value relative to the previous operation; Use DEFLATE algorithm for streaming compression; The compressed log format is: |Header identifier (0xAE)|Compression flag|Dictionary encoding|Δ coordinate value|CRC check|.

8. A modeling system based on dynamic 3D objects, characterized in that: include: An operation instruction generation module (1) is configured to receive user input and generate a three-dimensional modeling operation instruction, wherein the instruction at least includes an operation type field, a target object unique identifier, and a three-dimensional space scope parameter set; Semantic tree construction module (2): connected to the operation instruction generation module (1), including: Topology relationship analysis unit: extracts the three-dimensional spatial topology relationship of the target object based on the half-edge data structure and generates parent node data; Scope calculation unit: calculates the weight distribution matrix of child nodes according to the spatial impact algorithm corresponding to the operation type; Conflict detection module (3): bidirectionally connected with the local storage module and the network communication module, including: Semantic tree comparer: performs spatial scope overlap detection between the local operational semantic tree and the received remote semantic tree; Priority arbiter: Built-in time and space dimension judgment logic circuit, triggering the arbitration process when overlap is detected; An operation merging execution module (4) is connected to the output end of the conflict detection module (3), and includes: Rule base: stores a table of temporal and spatial priority coefficients, where structural modification P1, geometric deformation P2, and material editing P3 satisfy P1>P2>P3; Differential calculation unit: generates an operation increment package containing only the changes of vertices in the overlapping area; Log management module (5): communicates with all modules, including, Binary encoder: converts the operation record into a compressed log containing the hash value of the semantic tree; Local rollback controller: In response to a conflict signal, it performs coordinate rollback of the affected vertices based on the differential CRC checksum. Network communication module (6): adopts dual-channel design, including: Operation instruction transmission channel: encapsulates semantic tree data packets; Log synchronization channel: transmits compressed binary difference logs; Local storage module (7): integrated NVMe SSD and persistent memory, including, Operation cache: LRU algorithm manages the most recent 1000 operation instructions; Semantic tree snapshot library: versioned storage of topology data compressed in CSR format; Log buffer: a circular queue that stores the history of encryption operations; The semantic tree construction module (2), the conflict detection module (3), and the operation merging execution module (4) are interconnected via a shared memory bus at the hardware layer to form a real-time processing pipeline architecture.

9. An electronic device, characterized in that: include: one or more processors; one or more memories; and one or more computer programs, wherein the one or more computer programs are stored in the one or more memories, and the one or more computer programs include instructions that, when executed by the one or more processors, cause the electronic device to perform the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that The storage medium stores a program or instruction, and when the program or instruction is executed, the method according to any one of claims 1 to 7 is implemented.

Images