Artificial intelligence-based three-dimensional model generation and printing service system
By using blockchain-based rights confirmation and multi-dimensional weighted algorithm scheduling, combined with an integrated cloud-based process, the problems of copyright protection and printing scheduling after 3D model generation are solved, achieving secure, efficient, and data-real integration of 3D model generation and printing services.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Filing Date
- 2026-05-08
- Publication Date
- 2026-07-14
AI Technical Summary
In existing technologies, 3D models lack copyright protection after generation, the ownership of secondary creations is unclear, the accuracy of distributed printing scheduling and matching is low, and the digital and physical manufacturing processes are separated, making it difficult to build an economic ecosystem for creators and resulting in low efficiency in physical manufacturing.
It employs a Web3 original and derivative copyright confirmation module, a distributed printing farm intelligent matching and scheduling module, a cloud-based 3D model generation and rendering module, a fault-tolerant and over-deduction billing module, an automatic slicing and abnormal circuit breaker module, and a cross-domain proxy rendering module to achieve blockchain-based confirmation of 3D model ownership, multi-dimensional weighted algorithm scheduling, and cloud-based integrated process management.
By leveraging blockchain for ownership verification and traceability, we can protect the copyright of original and derivative works, optimize the matching of printing nodes, build a secure and efficient 3D model generation and printing service system, and achieve seamless integration of the entire lifecycle of digital and physical models.
Smart Images
Figure CN122391497A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the fields of artificial intelligence-generated content, intelligent manufacturing, and digital copyright protection, and particularly to an artificial intelligence-based 3D model generation and printing service system. Background Technology
[0002] With the development of deep learning technology, the technology of automatically generating three-dimensional models using natural language or two-dimensional images (3DAIGC) has gradually matured.
[0003] However, existing technical solutions mainly focus on the "how to generate the model" stage, leaving significant technical gaps and pain points in the subsequent stages:
[0004] Lack of copyright confirmation and chaotic secondary creation: Existing AIGC platforms usually only provide model file downloads and lack copyright protection for the generated results; user-generated works are easily copied and misappropriated by others; in addition, "secondary creations" based on modifications to the original model are difficult to define the contribution ratio between original and secondary creations under the existing framework, resulting in unclear IP ownership and seriously hindering the construction of the creator economic ecosystem.
[0005] Inefficient physical manufacturing scheduling: Existing cloud-based 3D printing services mostly adopt a "centralized factory" model or a simple order dispatch logic; after a user submits a printing request, the system often fails to optimally match the specific requirements of the order (such as specific color, material, and precision) with the real-time capacity of distributed printing farms; for example, when a user needs to print a red and blue dual-color model, the existing system has difficulty automatically filtering out printers that are currently equipped with red and blue material trays and are available, resulting in users having to wait in long queues or communicate manually, leading to long manufacturing cycles and low resource utilization.
[0006] The separation between virtual and reality: Most current systems treat "digital model generation" and "physical printing" as two separate processes, lacking a unified system to manage the entire lifecycle of data from the birth (generation), ownership confirmation (on-chain) to physicalization (print scheduling).
[0007] Therefore, it is necessary to provide an AI-based 3D model generation and printing service system to solve the above-mentioned technical problems. Summary of the Invention
[0008] This invention provides an artificial intelligence-based 3D model generation and printing service system, which solves the problems of difficulty in confirming the copyright of existing 3D AIGC models, unclear ownership of secondary creations, low accuracy of distributed printing scheduling and matching, and the separation between digital models and physical manufacturing processes.
[0009] To address the aforementioned technical problems, the present invention provides an artificial intelligence-based 3D model generation and printing service system, comprising:
[0010] Web3 original and derivative copyright confirmation module, distributed printing farm intelligent matching and scheduling module, cloud 3D model generation and rendering module, fault-tolerant anti-over-deduction billing module, automatic slicing and abnormal circuit breaker module, cross-domain proxy rendering module;
[0011] The Web3 original and secondary creation copyright confirmation module is used to perform blockchain-based copyright confirmation of AI-generated 3D models and to trace the source of secondary creations and distribute rights and benefits.
[0012] The distributed printing farm intelligent matching and scheduling module is used to collect printing farm node data, parse order requirements, and complete intelligent allocation of printing nodes through a multi-dimensional weight algorithm.
[0013] The cloud-based 3D model generation and rendering module is used to control the cloud-based generation and front-end rendering of 3D models through multimodal commands.
[0014] The fault-tolerant and over-deductible billing module is used to enable zero-point departure for the 3D model generation task and formal deduction of fees after successful completion.
[0015] The automatic slicing and abnormal circuit breaking module is used to automatically slice the 3D model into G-code files and trigger circuit breaking and return when the model has unprintable defects.
[0016] The cross-domain proxy rendering module is used to bypass browser cross-domain restrictions and complete the secure rendering and loading of third-party cloud storage 3D assets.
[0017] Preferably, the Web3 original and derivative copyright confirmation module includes: an original copyright confirmation unit, used to extract the geometric feature hash of the AI-generated 3D model as a digital fingerprint, write the digital fingerprint, generation timestamp, and creator wallet address into the blockchain, and mint an NFT copyright certificate; a derivative copyright confirmation unit, used to identify the original on-chain ID of the derivative model, quantify the degree of model modification, and mint an NFT marked as a derivative work when the modification degree meets the standard, and associate it with the original contract address; and a profit-sharing unit, used to automatically distribute the revenue of derivative works between the original creator and the derivative creator according to a preset ratio through a smart contract.
[0018] Preferably, the distributed printing farm intelligent matching and scheduling module includes: a data acquisition unit, used to acquire in real time the geographical location, equipment model, material tray status matrix, and equipment operating status data of the printing farm nodes through an IoT interface; an order parsing unit, used to extract the target material, color requirements, model size, and expected delivery location parameters of the printing order; a multi-level filtering and scoring unit, used to first filter by hard indicators of material and size, then calculate the color matching degree for filtering, and finally calculate and sort the node scores using a comprehensive weight formula; and a dynamic allocation unit, used to allocate orders to the highest-scoring node, synchronously distribute slice data, and estimate delivery time.
[0019] Preferably, the comprehensive score formula of the multi-level screening and scoring unit is: S=W1×Distance+W2×ColorMatch+W3×IdleRate+W4×QueueLength; where Distance is the geographical distance between the user and the farm, ColorMatch is the color matching coefficient, IdleRate is the device idle rate, QueueLength is the number of queued orders, and W1 to W4 are configurable weight coefficients.
[0020] Preferably, the cloud-based 3D model generation and rendering module includes: a task dispatch unit, used to verify user points, store data in the local database, and generate task tracking credentials; a cloud-based proxy rendering unit, used to establish a server-side relay channel and send multi-file dependent 3D resources to the front end; a billing circuit breaker unit, used to issue a formal deduction instruction only after the 3D model is successfully parsed and verified; and a physical deletion unit, used to physically delete invalid assets from cloud storage space using UUID.
[0021] Preferably, the fault-tolerant anti-over-deduction billing module execution process is as follows: pre-inspection and access, after verifying that the user has sufficient points, execute zero-point departure; credential acquisition, encapsulate the user instruction and send it to the large model server and obtain the remote task ID; image persistence, generate a record in the local task table and return the ID to the client; status recovery, automatically resume the task polling progress after the user disconnects and reconnects; circuit breaker determination, activate the deduction after monitoring the task success and obtaining the geometric file direct link; formal settlement, execute point deduction, record the billing timestamp and store the model in the user asset database.
[0022] Preferably, the automatic slicing and abnormal circuit breaking module is used to slice the 3D model into G-code files according to the device parameters before printing. When a non-manifold geometric defect is detected in the model, the abnormal circuit breaking is automatically triggered, the model repair module is returned, and the user is notified.
[0023] Preferably, the execution flow of the cross-domain proxy rendering module is as follows: Session authorization: The client requests a loading instruction from the system, and the system generates a URL with a security signature and a session authorization token; Path mapping: The token is associated with the cloud resource path and time-limited management is set; Dynamic proxy request: The front end sends a request to the proxy gateway with the token and file name; Security resolution: The proxy gateway verifies the token and merges it to generate the real resource address; Streaming pass-through: The 3D data is passed through to the front end through a high-bandwidth node with micro-slice buffer; Sandbox reconstruction: The front end completes the loading and rendering display of 3D assets within the browser sandbox.
[0024] Preferably, the distributed printing farm intelligent matching and scheduling module requires the use of a user terminal, which includes a tablet computer body. The outer side of the tablet computer body is provided with a protective component, the back of the tablet computer body is provided with a connecting frame, the back of the connecting frame is provided with a limit component, and the back of the limit component is provided with a handle.
[0025] Preferably, the protective component includes four protective corners and four elastic bands. The four elastic bands are fixedly installed between the four protective corners. The four protective corners are respectively located at the four corners of the tablet computer body. The limiting component includes a limiting frame, which is fixedly installed on the back of the connecting frame. A rotating column is connected inside the back of the limiting frame. A movable plate is fixedly installed at one end of the rotating column. Multiple positioning balls are fixedly installed on one side of the movable plate. Multiple positioning grooves are opened on one side of the inner side of the fixed frame. The multiple positioning balls are engaged with the inner side of the multiple positioning grooves. A spring is sleeved on the outer side of the rotating column. The other end of the rotating column is fixedly installed with the handle.
[0026] Compared with related technologies, the artificial intelligence-based 3D model generation and printing service system provided by this invention has the following advantages:
[0027] This invention provides an AI-based 3D model generation and printing service system. Through the coordinated efforts of modules including Web3 original and derivative copyright confirmation, distributed printing farm intelligent matching and scheduling, cloud-based 3D model generation and rendering, fault-tolerant and anti-over-deduction billing, automatic slicing and anomaly circuit breaking, and cross-domain proxy rendering, this system enables copyright confirmation, traceability, and automatic profit sharing for original and derivative models via blockchain. It achieves efficient scheduling by intelligently matching optimal printing nodes through a multi-dimensional weighting algorithm and connects the entire lifecycle of digital models and physical printing through a cloud-integrated process. This solves the problems of difficult copyright confirmation for AI 3D models, unclear ownership of derivative works, low accuracy of distributed printing scheduling and matching, and the disconnect between digital and physical processes, thus constructing a secure, efficient, and data-integrated 3D model generation and printing service system. Attached Figure Description
[0028] Figure 1 A flowchart illustrating the first embodiment of the AI-based 3D model generation and printing service system provided by the present invention;
[0029] Figure 2 A schematic diagram of the user terminal structure used in the artificial intelligence-based 3D model generation and printing service system provided by the present invention;
[0030] Figure 3 for Figure 2 The diagram shows another perspective of the structure.
[0031] Figure 4 for Figure 3 The diagram shows a cross-sectional view of the limiting component.
[0032] The following are the labels in the diagram: 1. Tablet PC body, 2. Protective components, 21. Protective feet, 22. Elastic band, 3. Connecting frame, 4. Limiting components, 41. Limiting frame, 42. Rotating column, 43. Moving plate, 44. Spring, 45. Positioning ball, 46. Positioning groove, 5. Handle. Detailed Implementation
[0033] The present invention will be further described below with reference to the accompanying drawings and embodiments.
[0034] First Embodiment
[0035] Please refer to the following: Figure 1 ,in, Figure 1 This is a flowchart of the first embodiment of the artificial intelligence-based 3D model generation and printing service system provided by the present invention.
[0036] The AI-based 3D model generation and printing service system includes: a Web3 original and derivative copyright confirmation module, a distributed printing farm intelligent matching and scheduling module, a cloud-based 3D model generation and rendering module, a fault-tolerant and over-deduction-prevention billing module, an automatic slicing and abnormal circuit breaker module, and a cross-domain proxy rendering module.
[0037] The Web3 original and secondary creation copyright confirmation module is used to perform blockchain-based copyright confirmation of AI-generated 3D models and to trace the source of secondary creations and distribute rights and benefits.
[0038] The distributed printing farm intelligent matching and scheduling module is used to collect printing farm node data, parse order requirements, and complete intelligent allocation of printing nodes through a multi-dimensional weight algorithm.
[0039] The cloud-based 3D model generation and rendering module is used to control the cloud-based generation and front-end rendering of 3D models through multimodal commands.
[0040] The fault-tolerant and over-deductible billing module is used to enable zero-point departure for the 3D model generation task and formal deduction of fees after successful completion.
[0041] The automatic slicing and abnormal circuit breaking module is used to automatically slice the 3D model into G-code files and trigger circuit breaking and return when the model has unprintable defects.
[0042] The cross-domain proxy rendering module is used to bypass browser cross-domain restrictions and complete the secure rendering and loading of third-party cloud storage 3D assets.
[0043] The Web3 original and derivative copyright confirmation module includes: an original copyright confirmation unit, used to extract the geometric feature hash of the AI-generated 3D model as a digital fingerprint, write the digital fingerprint, generation timestamp, and creator wallet address into the blockchain, and mint an NFT copyright certificate; a derivative copyright confirmation unit, used to identify the original on-chain ID of the derivative model, quantify the degree of model modification, and mint an NFT marked as a derivative work when the modification degree meets the standard, and associate it with the original contract address; and a profit-sharing unit, used to automatically distribute the revenue of derivative works between the original creator and the derivative creator according to a preset ratio through a smart contract.
[0044] Originality Confirmation Unit: Model Generation and Fingerprint Extraction: After a user generates a 3D model using AI, the system automatically extracts the model's geometric feature hash as a "digital fingerprint"; On-chain Minting: The system calls a smart contract interface to write the digital fingerprint, generation timestamp, creator's wallet address, and other information into the blockchain (such as Ethereum Layer 2 or a consortium blockchain); Copyright Certificate Generation: After successful minting, a unique NFT asset certificate is generated, establishing the user's original author identity and copyright ownership; Cloud Proxy Rendering Module: Utilizing a relay channel established on the server side, the browser's Cross-Origin Resource Restriction (CORS) is bypassed, and 3D resources with multiple file dependencies are delivered to the front end.
[0045] Secondary Creation Rights Confirmation Unit: Lineage Relationship Identification: When User B modifies User A's model (secondary creation), the system reads the original model's on-chain ID when loading the model; Contribution Calculation: The system compares the model's geometric data before and after modification, calculating parameters such as volume change and texture change rate to quantify the extent of modification. Derivative Product Minting: If the modification exceeds a preset threshold (e.g., 30%), the system allows the secondary creator to mint new NFTs, but these NFTs are marked as "derivative works" on-chain and forcibly associated with the original work's contract address; Profit Sharing: If a secondary creation is traded or printed for a fee, the smart contract will automatically distribute royalties according to a preset ratio (e.g., 30% for the original author, 70% for the secondary creator), protecting the rights of the original IP.
[0046] The distributed printing farm intelligent matching and scheduling module includes: a data acquisition unit, used to acquire in real time the geographical location, equipment model, material tray status matrix, and equipment operating status data of the printing farm nodes through an IoT interface; an order parsing unit, used to extract the target material, color requirements, model size, and expected delivery location parameters of the printing order; a multi-level filtering and scoring unit, used to first filter by hard indicators of material and size, then calculate the color matching degree for filtering, and finally calculate and sort the node scores using a comprehensive weight formula; and a dynamic allocation unit, used to allocate orders to the highest-scoring node, synchronously distribute slice data, and estimate delivery time.
[0047] The comprehensive score formula for the multi-level screening and scoring unit is: S=W1×Distance+W2×ColorMatch+W3×IdleRate+W4×QueueLength; where Distance is the geographical distance between the user and the farm, ColorMatch is the color matching coefficient, IdleRate is the device idle rate, QueueLength is the number of queued orders, and W1 to W4 are configurable weight coefficients.
[0048] Data Acquisition Unit: The system accesses data from each printing farm node in real time via an IoT interface; data dimensions include: Geographic location: latitude and longitude coordinates of the farm; Equipment model: Supports equipment models for different processes such as FDM, SLA, and SLM; Pad status matrix: Records the current pad color (e.g., red-blue-white combination), remaining weight, and material type (PLA / ABS / TPU) for each printer in real time; Operating status: Idle, Printing (remaining time), Under maintenance.
[0049] Order parsing unit: The system parses the print orders submitted by users and extracts key constraint parameters: target material, color requirements (single color or multi-color), model size, and expected delivery location.
[0050] Multi-level screening and scoring unit: Level 1 screening (hard indicator filtering): Excludes farm nodes whose equipment models do not support the material or whose print size is smaller than the model size; Level 2 screening (color matching calculation): If the order requires a multi-color model, the system calculates the matching degree between the "pawl color combination" of each farm node and the order requirement; for example: if the order requires red and blue, farm A equipment is currently filled with red + blue (matching degree 100%), farm B equipment is filled with red + white (matching degree 50%, material needs to be changed); prioritize nodes with high matching degree to reduce material change and cleaning time; Level 3 scoring (comprehensive weight ranking): calculate the comprehensive score S for the screened nodes: S=W1×Distance1+W2×ColorMatch+W3×IdleRate+W4×QueueLength where Distance is the geographical distance between the user address and the farm (the closer the distance, the higher the score), ColorMatch is the color matching coefficient, IdleRate is the equipment idle rate, QueueLength is the current queued order quantity (the less queued, the higher the score); W1 to W4 are configurable weight coefficients.
[0051] Dynamic allocation unit: The system automatically assigns orders to the farm node with the highest score and sends model slice data simultaneously; at the same time, the system estimates logistics time based on geographical location and provides feedback to the user on the estimated delivery time.
[0052] The cloud-based 3D model generation and rendering module includes: a task dispatch unit, used to verify user points, store data in the local database, and generate task tracking credentials; a cloud-based proxy rendering unit, used to establish a server-side relay channel to send multi-file dependent 3D resources to the front end; a billing circuit breaker unit, used to issue a formal deduction instruction only after the 3D model is successfully parsed and verified; and a physical deletion unit, used to physically delete invalid assets from cloud storage space using UUID.
[0053] Task dispatch unit: The entry point for user multimodal commands, responsible for verifying points and storing them in the local database, and generating a unique task tracking credential;
[0054] Cloud-based proxy rendering unit: Utilizes a relay channel established on the server side to bypass the browser's cross-domain resource restrictions (CORS) and deliver 3D resources with multiple file dependencies to the front end;
[0055] Billing and circuit breaker module unit: Based on asynchronous polling of task table records, it only issues a formal deduction instruction to the core account system after the three-dimensional entity text has been successfully parsed and verified;
[0056] Physical stripping unit: For assets marked as deleted or invalid, perform physical-level data removal from cloud storage space via UUID addressing.
[0057] The fault-tolerant and over-deduction-prevention billing module execution process is as follows: pre-inspection and access: after verifying that the user has sufficient points, the vehicle is launched with zero points; credential acquisition: the user's instruction is encapsulated and sent to the large model server and the remote task ID is obtained; image persistence: a record is generated in the local task table and the ID is returned to the client; status recovery: the task polling progress is automatically restored after the user reconnects from the network; circuit breaker determination: after monitoring the success of the task and obtaining the geometric file direct link, the deduction is activated; formal settlement: the points are deducted, the billing timestamp is recorded, and the model is stored in the user asset library.
[0058] Suppose a 3D model generation platform P has a user account U1 with a total account points of S1. User U1 now initiates a request command C to "generate a complex 3D model from text," which is expected to consume S2 points. Due to the long time required for remote model generation and the risk of network fluctuations, robust billing can be implemented in the following way:
[0059] Pre-screening and access: User U1 sends instruction C to platform P; platform P intercepts the request and verifies that S1≥S2. If the condition is met, the request is approved and the system executes "zero-point departure", that is, no points are deducted temporarily.
[0060] Credential Acquisition: Platform P encapsulates instruction C and sends it to a third-party large model server to obtain the remote task's unique ID1;
[0061] Image persistence: Platform P generates a record (containing U1 identifier, ID1 and estimated score S2) in the local database task table (studio_jobs), and returns ID1 to the client. The client then disconnects the long connection and switches to lightweight polling.
[0062] Status recovery: If user U1 closes the webpage or loses internet connection during the generation process, the front end will automatically restore the polling progress of ID1 by querying the studio_jobs table after logging back in, ensuring the continuity of task status;
[0063] Circuit break determination: Platform P continuously monitors the ID1 status. When the remote end returns SUCCEED and successfully retrieves the actual geometry file URL, the system activates the circuit breaker callback function.
[0064] Formal settlement: The circuit breaker function executes the actual deduction logic (deducting S2 from S1) and prints an immutable billing timestamp (charged_at), and then formally writes the results into U1's permanent asset database.
[0065] The automatic slicing and abnormal circuit breaking module is used to slice the 3D model into G-code files according to the equipment parameters before printing. When a non-manifold geometric defect is detected in the model, the abnormal circuit breaking is automatically triggered, the model repair module is returned, and the user is notified.
[0066] Before the print command is issued, the system's cloud-based slicing engine will automatically slice the 3D model into G-code files based on the target device's model parameters. If non-printable defects such as non-manifold geometry (broken surfaces) are found in the model during the slicing process, the system will automatically trigger an "abnormal circuit breaker," return the model to the model repair module, and notify the user for confirmation to avoid wasting printing materials.
[0067] The execution flow of the cross-domain proxy rendering module is as follows: Session authorization: The client requests a loading command from the system, and the system generates a URL with a security signature and a session authorization token; Path mapping: The token is associated with the cloud resource path and time-limited management is set; Dynamic proxy request: The front end sends a request to the proxy gateway with the token and file name; Security resolution: The proxy gateway verifies the token and merges it to generate the real resource address; Streaming pass-through: The 3D data is passed through to the front end through a high-bandwidth node with micro-slice buffer; Sandbox reconstruction: The front end completes the loading and rendering display of 3D assets within the browser sandbox.
[0068] When the generated system 3D file URL is stored in a third-party public cloud (such as OSS / COS), the following pipeline procedure is executed to resolve cross-domain restrictions and texture loss issues caused by browser sandboxing:
[0069] Session authorization: When a client needs to render 3D assets, it does not directly access the original URL, but instead requests a loading instruction from platform P; platform P extracts the full URL with a secure signature and establishes a session-level authorization token in memory;
[0070] Path mapping: The system associates this token with the corresponding cloud source root directory path and assigns a timestamp management mechanism to ensure timeliness;
[0071] Dynamic proxy request: The client's WebGL rendering engine sends an HTTP request to a specific reverse proxy gateway of platform P, carrying a token and the target file name (such as scene.obj or texture.png).
[0072] Security Analysis: After receiving the request, the proxy gateway verifies the token's validity and the anti-hotlinking mechanism, and then merges them to generate the underlying real pull address;
[0073] Streaming pass-through: Platform P uses aiohttp to establish concurrent streams on high-bandwidth nodes in the intranet, and wraps remote binary data into the system response body with micro-slice buffers (such as 64KB Chunks), and directly passes it back to the front end.
[0074] Sandbox Reconstruction: The front-end rendering engine receives a response stream with a standard same-origin header, identifies it as a safe source, and successfully loads complex spatial coordinates and texture meshes in the sandbox, completing the local reconstruction and display of the 3D asset.
[0075] A Web3-based mechanism for confirming the ownership of original and derivative works: This system introduces blockchain technology to generate a unique "digital fingerprint" for each generated 3D model. For original works, ownership is directly confirmed on the blockchain; for derivative works, the system automatically identifies their association with the original model through smart contracts and records the contributions of the derivative creators, constructing a clear IP traceability chain, effectively protecting the rights and interests of designers, and promoting the prosperity of the creative ecosystem.
[0076] Intelligent matching of print farms based on multi-dimensional parameters: This invention abandons the traditional static allocation mode and proposes a dynamic scheduling algorithm. The system can capture the geographical location of each print farm node in real time (optimizing logistics timeliness), the number and model of printers (matching process requirements), the color combination of the material trays (matching appearance requirements), and the equipment operating status (matching production capacity), and calculate the optimal allocation scheme. This greatly reduces the user's waiting time and improves the equipment utilization rate of distributed print farms.
[0077] Robust asynchronous scheduling and a "zero-cancellation" billing mechanism: Addressing the pain points of long processing times and frequent timeouts in generating 3D assets from large models, the system constructs a dual-link polling system based on the studio_jobs persistent database. The front-end uses a "real-state driven + smooth tweening" technique to visualize progress; the underlying billing executes strict circuit breaker logic: requests are dispatched immediately without penalty, and a billing timestamp (charged_at) is only recorded when the model task is successfully confirmed to be stored in the database. This mechanism fundamentally eliminates the problem of duplicate penalty points and cancelled orders caused by network fluctuations or interface disconnections.
[0078] An end-to-end automated production chain integrating data and reality: The system breaks down the software barriers between traditional 3D design and manufacturing, achieving a seamless connection from "AI simulation-panoramic preview" to "physical manufacturing." Users only need to input language or image commands, and the system can automatically connect model inference calls, WebGL real-time rendering, and cloud-based manufacturing preprocessing. Through standardized interfaces, the system can automatically verify the model's topology and calculate its volume, sending the data directly to a remote printer cluster. Users can complete the entire process from creative conception to physical delivery without switching slicing software or manually converting formats.
[0079] The community circulation and secondary creation mechanism based on "genetic tracing" differs from traditional platforms that only distribute "dead meshes (OBJ)". This system endows 3D assets with a traceability map. When an asset enters the community circulation, it not only contains geometric data but also carries "original genes" such as material trees, generation parameters, and prompts. Secondary creators can use the interface to instantly recall the original generation environment for iterative fine-tuning, realizing an ecological evolution from "one-time generation and download" to "unlimited controllable relay creation", greatly enhancing the reuse value of AIGC assets.
[0080] A closed-loop data system that integrates virtual and physical elements: The system links virtual digital assets (NFTs) with physical print orders. Users not only own the digital model, but their physical print records are also traceable, achieving transparent management from virtual creation to physical delivery.
[0081] Compared with related technologies, the artificial intelligence-based 3D model generation and printing service system provided by this invention has the following advantages:
[0082] By employing a collaborative structure that integrates modules for Web3 original and derivative works copyright confirmation, distributed printing farm intelligent matching and scheduling, cloud-based 3D model generation and rendering, fault-tolerant and over-deduction-prevention billing, automatic slicing and anomaly circuit breaking, and cross-domain proxy rendering, this system enables AI 3D model generation and printing services. It utilizes blockchain technology for copyright confirmation, traceability, and automatic profit sharing of original and derivative models, employs multi-dimensional weighting algorithms for intelligent matching of optimal printing nodes for efficient scheduling, and achieves seamless integration of the entire lifecycle of digital models and physical printing through a cloud-integrated process. This addresses the challenges of difficult copyright confirmation for AI 3D models, unclear ownership of derivative works, low accuracy in distributed printing scheduling and matching, and the disconnect between digital and physical processes, thus constructing a secure, efficient, and data-integrated 3D model generation and printing service system.
[0083] Second Embodiment
[0084] Please refer to the following: Figure 2 , Figure 3 and Figure 4 Based on the AI-based 3D model generation and printing service system provided in the first embodiment of this application, the second embodiment of this application proposes another AI-based 3D model generation and printing service system. The second embodiment is merely a preferred embodiment of the first embodiment, and its implementation will not affect the separate implementation of the first embodiment.
[0085] Specifically, the difference between the AI-based 3D model generation and printing service system provided in the second embodiment of this application is that the distributed printing farm intelligent matching and scheduling module of the AI-based 3D model generation and printing service system requires the use of a user terminal. The user terminal includes a tablet computer body 1, a protective component 2 is provided on the outer side of the tablet computer body 1, a connecting frame 3 is provided on the back of the tablet computer body 2, a limiting component 4 is provided on the back of the connecting frame 3, and a handle 5 is provided on the back of the limiting component 4.
[0086] The protective component 2 includes four protective corners 21 and four elastic bands 22. The four elastic bands 22 are fixedly installed between the four protective corners 21. The four protective corners 21 are respectively located at the four corners of the tablet computer body 1. The limiting component 4 includes a limiting frame 41. The limiting frame 41 is fixedly installed on the back of the connecting frame 3. A rotating column 42 is connected inside the back of the limiting frame 41. A movable plate 43 is fixedly installed at one end of the rotating column 42. A plurality of positioning balls 45 are fixedly installed on one side of the movable plate 43. A plurality of positioning grooves 46 are opened on one side of the inner side of the fixed frame 41. The plurality of positioning balls 45 are engaged with the inner side of the plurality of positioning grooves 46. A spring 44 is sleeved on the outer side of the rotating column 42. The other end of the rotating column 42 is fixedly installed with the handle 5.
[0087] The working principle of the AI-based 3D model generation and printing service system provided by this invention is as follows:
[0088] Users can input commands, view models, and place print orders via the tablet computer 1; the protective corners 21 and elastic bands 22 provide anti-collision and anti-drop protection for the four corners of the equipment; pulling the handle 5 can drive the rotating column 42 and the moving plate 43, causing the positioning ball 45 to disengage from the positioning groove 46, rotate to a suitable angle, and then be released. The spring 44 pushes the positioning ball 45 into the positioning groove 46 to lock it, realizing multi-angle handheld support, which is convenient for on-site mobile operation and viewing.
[0089] Compared with related technologies, the artificial intelligence-based 3D model generation and printing service system provided by this invention has the following advantages:
[0090] Through the coordinated structure of protective corner 21, elastic band 22, connecting frame 3, limiting frame 41, rotating column 42, moving plate 43, spring 44, positioning ball 45 and positioning groove 46, when the tablet computer body 1 is moved on-site, the protective corner 21 and elastic band 22 can protect the four corners of the equipment from collision and drop, thereby allowing the rotating column 42 to drive the positioning ball 45 to lock into the positioning groove 46 for angle locking, realizing multi-angle handheld support, which is convenient for users to operate and view the model on-site, thereby increasing the convenience and protection of the user terminal.
[0091] The above description is merely an embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.
Claims
1. A 3D model generation and printing service system based on artificial intelligence, characterized in that, include: Web3 original and derivative copyright confirmation module, distributed printing farm intelligent matching and scheduling module, cloud 3D model generation and rendering module, fault-tolerant anti-over-deduction billing module, automatic slicing and abnormal circuit breaker module, cross-domain proxy rendering module; The Web3 original and secondary creation copyright confirmation module is used to perform blockchain-based copyright confirmation of AI-generated 3D models and to trace the source of secondary creations and distribute rights and benefits. The distributed printing farm intelligent matching and scheduling module is used to collect printing farm node data, parse order requirements, and complete intelligent allocation of printing nodes through a multi-dimensional weight algorithm. The cloud-based 3D model generation and rendering module is used to control the cloud-based generation and front-end rendering of 3D models through multimodal commands. The fault-tolerant and over-deductible billing module is used to enable zero-point departure for the 3D model generation task and formal deduction of fees after successful completion. The automatic slicing and abnormal circuit breaking module is used to automatically slice the 3D model into G-code files and trigger circuit breaking and return when the model has unprintable defects. The cross-domain proxy rendering module is used to bypass browser cross-domain restrictions and complete the secure rendering and loading of third-party cloud storage 3D assets.
2. The artificial intelligence-based 3D model generation and printing service system according to claim 1, characterized in that, The Web3 original and derivative copyright confirmation module includes: an original copyright confirmation unit, used to extract the geometric feature hash of the AI-generated 3D model as a digital fingerprint, write the digital fingerprint, generation timestamp, and creator wallet address into the blockchain, and mint an NFT copyright certificate; a derivative copyright confirmation unit, used to identify the original on-chain ID of the derivative model, quantify the degree of model modification, and mint an NFT marked as a derivative work when the modification degree meets the standard, and associate it with the original contract address; and a profit-sharing unit, used to automatically distribute the revenue of derivative works between the original creator and the derivative creator according to a preset ratio through a smart contract.
3. The artificial intelligence-based 3D model generation and printing service system according to claim 1, characterized in that, The distributed printing farm intelligent matching and scheduling module includes: a data acquisition unit, used to acquire in real time the geographical location, equipment model, material tray status matrix, and equipment operating status data of the printing farm nodes through an IoT interface; an order parsing unit, used to extract the target material, color requirements, model size, and expected delivery location parameters of the printing order; a multi-level filtering and scoring unit, used to first filter by hard indicators of material and size, then calculate the color matching degree for filtering, and finally calculate and sort the node scores using a comprehensive weight formula; and a dynamic allocation unit, used to allocate orders to the highest-scoring node, synchronously distribute slice data, and estimate delivery time.
4. The artificial intelligence-based 3D model generation and printing service system according to claim 3, characterized in that, The comprehensive score formula for the multi-level screening and scoring unit is: S=W1×Distance+W2×ColorMatch+W3×IdleRate+W4×QueueLength; where Distance is the geographical distance between the user and the farm, ColorMatch is the color matching coefficient, IdleRate is the device idle rate, QueueLength is the number of queued orders, and W1 to W4 are configurable weight coefficients.
5. The artificial intelligence-based 3D model generation and printing service system according to claim 1, characterized in that, The cloud-based 3D model generation and rendering module includes: a task dispatch unit, used to verify user points, store data in the local database, and generate task tracking credentials; a cloud-based proxy rendering unit, used to establish a server-side relay channel to send multi-file dependent 3D resources to the front end; a billing circuit breaker unit, used to issue a formal deduction instruction only after the 3D model is successfully parsed and verified; and a physical deletion unit, used to physically delete invalid assets from cloud storage space using UUID.
6. The artificial intelligence-based 3D model generation and printing service system according to claim 1, characterized in that, The fault-tolerant and over-deduction billing module execution process is as follows: pre-inspection and access, after verifying that the user has sufficient points, execute zero-point departure; credential acquisition, encapsulate the user instruction and send it to the large model server and obtain the remote task ID; image persistence, generate a record in the local task table and return the ID to the client. Status recovery: The task polling progress will automatically resume after the user reconnects to the network. Circuit breaker determination: The fee will be activated after the monitoring task is successfully completed and the geometric file direct link is obtained. Formal settlement involves deducting points, recording the billing timestamp, and storing the model in the user asset database.
7. The artificial intelligence-based 3D model generation and printing service system according to claim 1, characterized in that, The automatic slicing and abnormal circuit breaking module is used to slice the 3D model into G-code files according to the equipment parameters before printing. When a non-manifold geometric defect is detected in the model, the abnormal circuit breaking is automatically triggered, the model repair module is returned, and the user is notified.
8. The artificial intelligence-based 3D model generation and printing service system according to claim 1, characterized in that, The execution flow of the cross-domain proxy rendering module is as follows: Session authorization: The client requests a loading instruction from the system, and the system generates a URL with a security signature and a session authorization token; Path mapping: The token is associated with the cloud resource path and time-limited management is set; Dynamic proxy request: The front end sends a request to the proxy gateway with the token and file name; Security analysis: The proxy gateway verifies the token and merges them to generate the real resource address; streaming pass-through: 3D data is passed through to the front end via micro-slice buffering through high-bandwidth nodes; sandbox reconstruction: The front end completes the loading and rendering display of 3D assets within the browser sandbox.
9. The artificial intelligence-based 3D model generation and printing service system according to claim 1, characterized in that, The distributed printing farm intelligent matching and scheduling module requires the use of a user terminal, which includes a tablet computer body. The outer side of the tablet computer body is provided with a protective component, and the back of the tablet computer body is provided with a connecting frame. The back of the connecting frame is provided with a limit component, and the back of the limit component is provided with a handle.
10. The artificial intelligence-based 3D model generation and printing service system according to claim 9, characterized in that, The protective assembly includes four protective corners and four elastic bands. The four elastic bands are fixedly installed between the four protective corners. The four protective corners are respectively located at the four corners of the tablet computer body. The limiting assembly includes a limiting frame, which is fixedly installed on the back of the connecting frame. A rotating column is connected inside the back of the limiting frame. A movable plate is fixedly installed at one end of the rotating column. Multiple positioning balls are fixedly installed on one side of the movable plate. Multiple positioning grooves are opened on one side of the inner side of the fixed frame. The multiple positioning balls are engaged with the inner side of the multiple positioning grooves. A spring is sleeved on the outer side of the rotating column. The other end of the rotating column is fixedly installed with the handle.