Method for three-dimensional model particle splitting and dispersing and multi-mode feedback

Abstract

The invention relates to the technical field of three-dimensional model image data processing, and discloses a three-dimensional model particularization splitting deduction and multi-mode feedback method, which comprises the following steps: carrying out lightweight three-dimensional model reconstruction according to an equipment process structure diagram, and generating a glTF format data packet with a process information embedding characteristic; establishing a virtual three-dimensional scene through a three.js engine, and instantiating a three-dimensional model object in a glTF format; non-uniform granulation processing is carried out on the equipment model by adopting a splitting and deducting mechanism, discrete motion tracks of parts are solved through a response algorithm, and an equipment dynamic decomposition path display effect is constructed; a calibration sensing link framework is established, an interaction response mechanism based on user behaviors is constructed, and a multi-mode visual feedback system of a large screen is achieved; according to the method, the technical bottlenecks of slow model loading, high interaction delay and the like in traditional industrial large screen display are effectively solved, and the dynamic interaction display effect and user perception of an industrial equipment structure are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of three-dimensional model image data processing, and in particular to a method for particle splitting, propagation and multi-modal feedback of a three-dimensional model. Background Art

[0002] In the wave of digital technology innovation, the design of industrial large-screen displays is no longer limited to two-dimensional flat display technology, and has significant limitations in the presentation of data spatial relationships, dynamic process simulation, and user interaction dimensions. As the pace of industrial development accelerates, scenarios such as intelligent manufacturing and digital twins urgently need to use three-dimensional models to construct the spatial topological structure of equipment operating status, supplemented by particle motion effects to achieve real-time visualization of data flows, but the existing two-dimensional interface can hardly support such needs. Summary of the Invention

[0003] This invention proposes a method for three-dimensional model particle splitting and propagation and multimodal feedback, aiming to effectively solve the limitations of industrial large-screen displays through three-mode modeling, motion effect production and interactive technology, enhance the technological sense and visual impact of the large screen, and achieve a deeper immersive experience of human-computer interaction.

[0004] In order to solve the above technical problems, the technical solutions of the present invention are as follows:

[0005] A method for particle splitting and propagation of a three-dimensional model and multimodal feedback, specifically: Based on the equipment assembly process, lightweight 3D models of industrial equipment are reconstructed to generate glTF format data packets with process information embedded features. Build a virtual 3D scene using the three.js engine and instantiate 3D model objects in glTF format; A dynamic splitting mechanism is used to perform non-uniform particle processing on the equipment model. The discrete motion trajectory of the components is automatically solved through the response algorithm to construct a dynamic decomposition path display effect of the equipment. Build a calibration perception link framework, construct an interactive response mechanism based on user behavior, and realize a multimodal visual feedback system on a large screen.

[0006] The parametric modeling of the assembly process equipment is performed to reconstruct the lightweight three-dimensional model of the process equipment. The specific implementation steps are as follows: Using the physical device combined with the structural drawings as the input source, a new component layer is created through geometric extrusion, Boolean operations, scanning, and clone generator operations to achieve device white film modeling, create three-point lighting, import environment maps, and give the white film PBR material; Merge the created components into an assembly layer, standardize the naming of each layer, and calibrate the coordinate origin of each component to the geometric center to ensure the accuracy of motion path calculation during model deconstruction; Export 3D model data packages based on the glTF2.0 standard.

[0007] The virtual 3D scene is built based on the three.js engine, specifically: The supporting architecture is based on the three.js engine's graphics rendering framework and the front-end vue3 architecture system; Use the Scene container to create a basic container object for a virtual 3D scene to manage the hierarchical relationship of all rendered objects, control the global environment, and implement object organization and coordinate transformation; Use the WebGLRenderer renderer to achieve image conversion of 3D scenes, build an interactive virtual 3D environment infrastructure, and complete real-time rendering and visualization of 3D scenes; A PerspectiveCamera is used to simulate the natural visual perception of the human eye. The field of view is defined by parameters (such as fov field of view angle, aspect ratio, near clipping plane, far clipping plane), forming a perspective projection effect where objects appear larger near and smaller far away, thereby defining the 3D visualization perspective. Relying on the OrbitControls orbit controller, users can dynamically adjust the viewing angle through mouse or touch operations (such as rotation, zoom, and translation); Use Light ambient light and parallel light source to form a physical level lighting model to enhance the visual realism of the 3D scene; To ensure that the 3D scene can adapt to the adjustment of different device resolutions and window sizes, the browser resize event is dynamically monitored to capture the window size changes; Call the setPixelRatio method to detect the pixel density of the device screen and adjust the rendering resolution proportionally to avoid performance loss at high resolutions or blurring at low resolutions. Use setSize to synchronously update the renderer size and canvas properties to ensure that the rendering area accurately matches the browser window, ensuring the rendering efficiency and visual fidelity of the 3D scene on different terminals.

[0008] The visual three-dimensional model is specifically: Create a three.js-based GLTFLoader instance, call its asynchronous load method, dynamically obtain the complete glTF model data package including geometry, materials, animations, skeletons, etc. through the model URL resource path, and complete the parsing and processing of the 3D model to generate a 3D model object that can be rendered in the three.js scene; Encapsulate the loading logic of GLTFLoader based on Promise, monitor loading progress events in real time through the onProgress callback, and capture abnormal situations such as network request failures and data parsing errors through the onError callback, so as to achieve refined management and control of model loading progress, success and abnormal status; Use the add method of the Scene to mount the model entity to the scene to achieve visual rendering of the model in the virtual space.

[0009] The particle processing of the device model is specifically as follows: Filter model components and extract component geometry data, traverse the model's hierarchical structure and identify the required meshes to obtain vertex coordinate arrays; construct particle space mapping relationships based on component vertex attributes, and use BufferGeometry to efficiently convert the device model from solid meshes to discrete particles; Using a specific space segmentation algorithm, the particle cluster is cut and logically grouped to form several independent particle fragment units, and the particle model is mounted and rendered in the three-dimensional scene through the Scene add method.

[0010] The dynamic splitting mechanism is used to solve the discrete motion trajectory of the components and construct the dynamic decomposition path display effect of the equipment, specifically: Simulate the natural fragmentation trajectory of objects and realize the dynamic transformation effect of model shell to particle fragments; Use the physics engine gSap to define precise flight trajectories (initial position, target position, etc.) for each particle fragment group or model component; Set the motion parameters of each particle fragment and model component fragment object, configure nonlinear motion curves in combination with elastic functions, and simulate the effect of object shattering and scattering; The scene removes particle fragments outside the visual clipping plane, reducing rendering load and memory usage while ensuring visual effects.

[0011] The interactive response mechanism based on user behavior is constructed as follows: Extract API metadata based on the standardized interface description protocol and build a structured data model, including endpoint path, request method, parameter definition, and response structure, to form a mappable logical relationship network; Convert data into JSON format and perform data cleaning, filter invalid content, and repair abnormal situations.

[0012] The multi-modal visual feedback system for large screen is specifically: By registering DOM event listeners and linking the ray detection and anchoring functions of three.js, the model components selected by the user are determined in real time. For the selected target unit, its unique identifier and spatial metadata (position, rotation state, etc.) are extracted to form a structured calibration data package; When the system detects that the user operation matches the preset calibration completion rules, it automatically encapsulates the calibration data and constructs a standardized HTTP request, which is sent to the backend service through the browser's network interface; The back-end response processing related modules perform data verification and business logic processing, and generate response data containing the latest calibration status and actionable suggestions; after the front-end receives the response, it parses the results and updates the large-screen visualization content, intuitively feedbacking the calibration results through 3D scene adjustment, data panel refresh and dynamic visual effects.

[0013] The large screen displays the functions of self-recovery and continuous display when there is no interactive instruction, specifically: The system has a built-in intelligent no-operation detection mechanism. When the user does not generate any effective interaction for more than 120 seconds, the no-operation timeout logic is automatically triggered, the camera view is restored to the default position, the temporary annotation layer is hidden, the interactive prompt UI is closed, and the system switches to automatic demonstration mode. In automatic demonstration mode, the system plays key information and model effects of key equipment units in a loop according to the preset performance logic. The display content is precisely controlled through the timeline to ensure the smooth connection of each stage of the demonstration. At the same time, the display rhythm is dynamically adjusted according to the equipment parameters to meet the needs of different scenarios; the system continuously monitors user operations. Once new interactive behavior is detected, the current display is immediately terminated and restored to normal interactive mode to ensure the system's response speed and seamless switching of user experience.

[0014] The technical solution provided by the present invention has the following beneficial effects: The glTF data package generated by lightweight 3D reconstruction of the process structure diagram significantly improves GPU rendering efficiency, and the embedded process parameters can be directly parsed by the three.js engine, avoiding the resource loss of secondary data matching. The non-uniform particle splitting and spreading mechanism, combined with the response algorithm, enables real-time rendering of the device decomposition path, enhancing the technological sense and immersive experience of the large-scale visualization screen. The calibration perception link framework built can capture users' multimodal behaviors, trigger directional highlighting and motion trajectory feedback of equipment components, reduce the error rate of industrial large screens, and improve equipment structure cognition, providing key technical support for operation and maintenance training and fault diagnosis, and solving pain points such as low information density and single interaction. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Attachment Figure 1 This is a business flow chart of the present invention.

[0016] Attachment Figure 2 This is a flow chart of the method of Example 1 of the present invention.

[0017] Attachment Figure 3 This is a flow chart of the method of embodiment 2 of the present invention.

[0018] Attachment Figure 4 This is a system structure block diagram of Example 3 of the present invention. DETAILED DESCRIPTION

[0019] The following is a clear and complete description of the technical solutions used in the implementation of the present invention. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of them. All other embodiments derived by persons of ordinary skill in the art based on the embodiments of the present invention without creative effort are within the scope of protection of this application.

[0020] Example 1: As shown in the attached Figure 2 As shown, an embodiment of the present invention discloses a method for particle splitting, spreading and automatic demonstration of a three-dimensional model, including: S101 builds a virtual 3D model scene based on three.js, obtains the model data package through the load method of GLTFLoader, generates a 3D model object that can be rendered in the three.js scene, and mounts the model entity to the scene; S102 obtains a vertex coordinate array based on the model entity, constructs a particle space mapping relationship based on the vertex attributes of the components, realizes the transformation of the model from a solid mesh to discrete particles, and cuts and logically groups the particle cluster to form several independent particle fragment units; S103 sets the initial position, target position, and motion parameters of each particle fragment and model component fragment object, and then configures a nonlinear motion curve in combination with an elastic function to simulate the effect of object fragmentation and scattering; S104 plays key information and model effects of key equipment units in a loop according to the preset performance logic. The display content is precisely controlled through the timeline to ensure the smooth connection of each stage of the demonstration. At the same time, the display rhythm is dynamically adjusted according to the equipment parameters to meet the needs of different scenarios.

[0021] Example 2: As shown in the attached Figure 3 As shown, an embodiment of the present invention discloses a method for multimodal feedback of a three-dimensional model, comprising: S201 registers a DOM event listener to capture user behavior. By associating with the ray detection and anchoring functions of three.js, it determines the model parts selected by the user in real time, extracts their unique identifiers and spatial metadata, and forms a structured calibration data package. S202 determines whether it is necessary to send a related request based on the set parameters. If it is necessary to send a related request, then execute S204 first and then execute S203. Otherwise, execute S203. S203 analyzes the calibration results, intuitively feeds back the calibration results through 3D scene visual window adjustment, data panel refresh, and dynamic visual effects, and updates the large-screen visualization content; S204 maps the logical relationship network based on the user calibration request, extracts and cleans the calibration response data, converts the data into JSON format and returns it to the large screen; S205 has a built-in intelligent detection no-operation timeout mechanism, which monitors user behavior in real time, counts the length of time the large screen is in an inactive state, and determines whether the user has no operation behavior within a preset time period. If so, execute S206, otherwise repeat the step of S201 and reset the statistical time; S206 When the system triggers the no-operation mechanism, the camera angle is restored to the default position, the temporary annotation layer is hidden, and the interactive prompt UI is closed; S207 The system switches to automatic demonstration mode and executes S104.

[0022] Example 3: As shown in the attached Figure 4 As shown, an embodiment of the present invention discloses a method for particle splitting and propagation of a three-dimensional model and multimodal feedback, including: Model building and export: Perform parametric modeling based on assembly process equipment, reconstruct lightweight 3D models of process equipment, and generate 3D model data packages; Virtual scene construction: Using three.js and vue3 as the supporting framework, we build a scene container and use the renderer as the core to achieve 3D scene image conversion. We then use the camera and track controller to adjust the 3D perspective, and combine lighting to enhance the visual realism of the 3D scene, thus achieving a large-screen virtual 3D effect. Model mounting and motion effect implementation: Load the model data package, parse it, and mount it to the scene. Build a particle space mapping relationship based on the vertex attributes of the model entity components, implement discrete particleization and grouping, set the motion parameters of the fragment objects, configure nonlinear motion curves, and simulate the effect of objects breaking and scattering. User behavior parsing mechanism: monitors and captures user behavior, and through the association with three.js's ray detection and anchoring functions, determines the selected model components in real time, extracts their unique identifiers and spatial metadata, and forms a structured calibration data package; Data calibration and cleaning: Based on user calibration requests, map the logical relationship network, extract and clean the calibration response data, analyze the calibration results on the large screen and update the visualization content; Monitoring and demonstration: Monitor screen changes and user behavior, dynamically adjust the display effect of the large-screen visualization window, and loop the key information and model effects of key equipment units according to the preset performance logic in demonstration mode to achieve dynamic interactive display and cognitive assistance of the internal structure of industrial equipment.

Claims

1. A method for particle splitting and multimodal feedback of a three-dimensional model, characterized in that: include: Based on the assembly process, lightweight 3D models of industrial equipment are reconstructed to generate glTF format data packets with process information embedded features. A virtual 3D scene is built using the three.js engine, and 3D model objects in glTF format are instantiated. A dynamic splitting mechanism is used to perform non-uniform particle processing on the device model. The discrete motion trajectories of components are automatically calculated using a response algorithm to create a dynamic decomposition path display effect for the device. Build a calibration perception link framework, construct an interactive response mechanism based on user behavior, and realize a multimodal visual feedback system on a large screen.

2. The method for granularization, splitting, propagation and multimodal feedback of a three-dimensional model according to claim 1, characterized in that: Based on parametric modeling of assembly process equipment, lightweight 3D model reconstruction of process equipment is carried out, including: Using the physical equipment structure drawing as the input source, the model is lightweight reconstructed through geometric extrusion, Boolean operations, and clone generator operations; Merge component hierarchies, standardize layer naming, and reset component coordinates to zero to ensure the accuracy of motion path calculation during model deconstruction; Export 3D model data packages based on the glTF2.0 standard.

3. The method for granularization, splitting, propagation and multimodal feedback of a three-dimensional model according to claim 1, characterized in that: Build a virtual 3D scene based on the three.js engine, including: A graphics rendering framework based on the three.js engine, which establishes a 3D space coordinate system through the Scene container and uses the WebGLRenderer renderer to achieve initialization and real-time rendering of 3D scenes, building an interactive virtual 3D environment infrastructure. Use the PerspectiveCamera to define the 3D visualization perspective, combine it with the OrbitControls controller to implement user interactive perspective adjustment, and configure the Light ambient light and parallel light source to form a physical level lighting model to enhance the visual realism of the 3D scene; By monitoring the browser resize event, the window size changes are dynamically acquired, the setPixelRatio method is called to adjust the rendering resolution according to the device screen characteristics, and the setSize method is used to synchronously update the renderer size and canvas adaptation to ensure the rendering efficiency and visual fidelity of the 3D scene on different terminals.

4. A method for granularization, splitting, propagation and multimodal feedback of a three-dimensional model according to claim 2 or 3, characterized in that: Instantiate a 3D model object in glTF format, including: Create a GLTFLoader instance based on three.js, call its asynchronous load method, dynamically obtain glTF data through the model URL resource path, and complete the parsing and processing of the 3D model; Build callback functions based on Promise to achieve refined control of loading progress, success and abnormal status; Use the add method of the Scene to mount the model entity to the scene to achieve visual rendering of the model in the virtual space.

5. The method for granularization, splitting, propagation and multimodal feedback of a three-dimensional model according to claim 4, characterized in that: Particle processing of equipment models, including: Filter model components and extract component geometry data, use BufferGeometry to generate a particle system, build a particle space mapping relationship based on component vertex attributes, and implement particle processing of the device model; A specific algorithm is used to cut and group the ionized model to form ionized fragments, and the particle model is mounted and rendered in the three-dimensional scene through the Scene add method.

6. The method for granularization, splitting, propagation and multimodal feedback of a three-dimensional model according to claim 5, characterized in that: The dynamic splitting mechanism is used to solve the discrete motion trajectory of components and construct a dynamic decomposition path display effect of the equipment, including: Use the physics engine gSap to set flight paths for particle model fragments and model components; Set the flight time and flight effects of each particle fragment and model component to simulate the effect of objects breaking and scattering.

7. The method for granularization, splitting, propagation and multimodal feedback of a three-dimensional model according to claim 1, characterized in that: Build an interactive response mechanism based on user behavior, including: Extract API metadata based on the standardized interface description protocol and build a structured data model, including endpoint path, request method, parameter definition and response structure, to form a mappable logical relationship network; Convert data into JSON format and perform data cleaning.

8. The method for granularization, splitting, propagation and multimodal feedback of a three-dimensional model according to claim 7, characterized in that: Build a calibration perception link framework to implement a large-screen multimodal visualization feedback system, including: By capturing user behavior, rapid calibration of equipment units is achieved; Based on the command mapping engine, the calibration data is converted into 3D scene control parameters and rendered in real time to the display terminal; When the user operation meets the preset behavior rules, the relevant request is sent through the browser HTTP protocol; The backend responds to the calibration data and completes the visual feedback on the large screen.

9. The method for granularization, splitting, propagation and multimodal feedback of a three-dimensional model according to claim 1, characterized in that: The large screen displays the functions of self-recovery and continuous display when there is no interactive instruction, including: Based on the preset display plan, if the user does not perform any operation for more than 120 seconds, the display status reset operation is triggered; Through established logical rules, the 3D model components can be continuously displayed.

Images