Online multi-element cloud exhibition hall system based on 3D visualization

Abstract

The invention relates to an online multi-cloud exhibition hall system based on 3D visualization, and relates to the technical field of information processing technologies, full life cycle management of works and materials is realized through a workbench module, and a hotspot configuration module drives flexible definition of a multi-modal interaction component. The guide management module accurately controls scene switching and media playing rhythm based on a time axis global synchronization mechanism, the multi-theme module supports dynamic loading and mixed superposition of multiple sets of theme resources, and finally the rendering engine module integrates the instructions to realize high-precision 3D rendering and cross-terminal adaptation. Three-dimensional interactive display of texts / PDF / audios and videos is achieved through multi-modal information fusion, a dynamic Bessel algorithm is adopted to intelligently adjust a guide path and node duration so as to eliminate visual angle jump, layered rendering and a resource on-demand loading technology are combined, multi-theme seamless switching and same-screen comparison are supported, and the visual angle jump is eliminated. And the performance optimization is realized while the visual effect is ensured.

Description

Technical Field

[0001] The present application relates to the field of information processing technology, and in particular to a 3D visualization online multi-cloud exhibition hall system. Background Art

[0002] Currently, virtual tour guide technology is widely used in digital tourism, online education, virtual exhibition halls, and other fields. Its core value lies in providing users with an immersive information acquisition experience through three-dimensional scene roaming and multimedia interaction. Existing technologies generally use hotspot icons to mark key information points, integrate navigation nodes to achieve scene switching, and support basic theme configuration. However, as user demand for deeper interaction, more dense content, and more diverse scenes continues to increase, traditional solutions have gradually exposed systemic flaws in key technical aspects.

[0003] First, in terms of information presentation, existing hotspot card technology has long been stuck in the static display stage. While mainstream systems allow for the configuration of graphic and text content, they are unable to precisely bind visibility to the timeline. For example, when the guide node switches to the cultural relic display case, the associated restoration process video and background audio commentary must be manually triggered to play. The card display duration cannot automatically match the media duration, resulting in the card disappearing before the video is finished, or remaining on the screen after the audio ends. This "disconnection between display logic and content rhythm" makes the guide process lose narrative coherence.

[0004] Secondly, there is a significant "secondary processing" bottleneck in the tour guide production process. After the developer configures the camera path, the node time code needs to be exported to a third-party audio software (such as Audition) to match the commentary, and then re-imported into the system to calibrate the subtitle display. During this process, any fine-tuning of the path will trigger a chain reaction of realignment of sound and picture. What is more prominent is that the existing timeline management system can only linearly control a single element (such as lens movement or audio playback) and lacks a global synchronization mechanism across media types. When a user adds a 10-second close-up shot to an exhibit, the corresponding commentary segment must be manually extended and the timing of the subtitles must be adjusted frame by frame, and the complexity of the operation increases exponentially.

[0005] Furthermore, there is a serious imbalance between the need for coexistence of multiple themes and the resource management mechanism. Although the current solution supports theme switching, it uses a "full replacement" mode - when switching, all resources of the old theme need to be uninstalled and the new package needs to be loaded, resulting in a brief black screen and performance jitter in the scene. For example, when the museum tour switches from "day mode" to "night mode", the same building model is loaded repeatedly, resulting in memory waste. In addition, the theme style parameters (such as light intensity and material reflectivity) are adjusted in a global coverage manner, which makes it impossible to achieve comparative observation of the dual themes of "historical original appearance" and "modern restoration", restricting the application value of teaching and research scenarios. Summary of the Invention

[0006] The purpose of the present invention is to provide a 3D visualization online multi-cloud exhibition hall system, which solves the problems of fragmented multimedia information interaction, lag in theme switching and inefficient production process in the virtual tour system.

[0007] To achieve the above objectives, the present invention is implemented through the following technical solutions: A 3D visualization online multi-cloud exhibition hall system, including: The workbench module is used to manage user works and material resources, including work editing, material classification storage and recycling functions; the hotspot configuration module is connected to the workbench module and receives its material data, and is used to implement multimodal configuration of hotspot icons and card information and jump link management; A navigation management module is connected to the hotspot configuration module and receives its configuration rules, and is used to synchronously control the camera movement duration of the navigation node, the hotspot card display and hiding logic, and the media resource playback based on the timeline; A multi-theme module, connected to the workbench module and the navigation management module, is used to load multiple theme configuration packages and dynamically switch scene content according to user instructions; The rendering engine module is connected to the navigation management module and the multi-theme module, receives timeline instructions and theme rendering parameters, and performs 3D scene rendering and cross-platform interactive response.

[0008] Preferably, the hotspot configuration module includes: The text configuration unit supports multi-level text structures of main titles, subtitles, and body text, and can be configured with word-by-word printing animation effects; the file plus text configuration unit supports embedding PDF, audio, and video files, and each file is associated with an independent text description panel; The jump link control unit supports four types of link configurations: scene jump, custom page, third-party link and platform content jump.

[0009] Preferably, the navigation management module includes: The navigation node control unit dynamically adjusts the node switching duration based on the time controller of the VR scene engine, and the lens motion trajectory is generated by the cubic Bezier curve interpolation algorithm; The timeline synchronization unit globally aligns the camera trajectory of the navigation node, the appearance and disappearance time of the hotspot card, and the playback progress of the media resource, with an alignment error of less than 0.1 second.

[0010] Preferably, the display duration T of the hotspot card in the time axis synchronization unit is card According to the associated audio duration T audio Dynamic calculation, the calculation rules are:

[0011] Preferably, the multi-theme module includes: Theme container unit, which stores independent theme configuration packages and dynamically loads resources through on-demand loading technology; The hybrid rendering unit assigns 3D models of different themes to independent rendering layers and realizes superimposed display through transparency blending; the resource release unit uses the LRU algorithm to release the memory resources of inactive themes according to the active status of the theme.

[0012] Preferably, in the hybrid rendering unit: The final pixel color of the transparency blend is calculated by weighting the alpha channels of the foreground color and the background color; Global illumination parameters are dynamically adjusted based on the theme, including high ambient light intensity in light mode and low ambient light intensity in dark mode.

[0013] Preferably, the weighted calculation formula of the Alpha channel of the foreground color and the background color is as follows: final =C foreground α+C background (1-α); Among them, C foreground Indicates the foreground color, C background Represents the background color, α is the transparency value of the foreground, α∈[0,1].

[0014] Preferably, the rendering engine module includes: The scene rendering unit optimizes the rendering accuracy of 3D models based on frustum clipping and LOD technology, where the frustum clipping priority is calculated based on the spatial distance between the model and the camera; The interactive response unit triggers the display of hotspot cards, jump logic, and media resource playback through the event monitoring mechanism; Cross-platform output unit adapts rendering results to web, mobile and desktop terminals.

[0015] Preferably, in the scene rendering unit: The threshold for switching the model's level of detail in LOD technology is adaptively adjusted based on the user's viewing speed. The priority of frustum clipping is calculated as follows: the closer the model is in space, the higher the priority.

[0016] In summary, this application includes at least one of the following beneficial technical effects: 1. This application uses a multi-level text structure and mixed file embedding technology to support the composite configuration of main and subtitles, PDF previews, audio and video resources, and text descriptions. Compared with the single text or image display format in traditional navigation systems, it breaks through the bottleneck of a single information carrier and limited display dimensions. Users can freely combine interactive elements such as printed animations and floating text descriptions to make the knowledge conveyed in the virtual scene more vivid and three-dimensional.

[0017] 2. This application is based on a dynamic Bezier path interpolation algorithm and a global timeline alignment mechanism. The invention can automatically adjust the lens movement speed and node dwell time according to the density of scene content. Existing technologies often suffer from insufficient presentation of key content or sluggish display of simple scenes due to a fixed camera movement rhythm. In this solution, the system analyzes the hotspot distribution density in real time, dynamically extending the navigation time of complex areas. At the same time, it smoothly connects perspective switching through curved paths, eliminating the sense of screen jumpiness and making the navigation process more relaxed.

[0018] 3. This application adopts layered rendering and on-demand loading technology. The present invention realizes the superimposed display and smooth switching of multiple theme resources in a unified scene. In traditional solutions, theme switching requires restarting the scene or clearing the cache, resulting in a fragmented experience. This technology allows "historical mode" and "science fiction style" theme elements to coexist through independent rendering layer allocation. Users can slide to adjust the transparency mixing ratio. When the system load increases, the high-precision models of inactive themes are automatically unloaded, and the low-polygon version of the core content is retained, taking into account both visual effects and operational stability. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 This is the framework diagram of this application system; Figure 2 This is the framework diagram of the hotspot configuration module of this application; Figure 3 This is the framework diagram of the navigation management module of this application; Figure 4 This is the framework diagram of the multi-theme module of this application; Figure 5 This is a framework diagram of the rendering engine module of this application. DETAILED DESCRIPTION

[0020] The following is combined with Figure 1 , further details of this application are given.

[0021] A 3D visualization online multi-cloud exhibition hall system, referring to Figure 1 ,include: The workbench module is used to manage user works and material resources, including work editing, material classification storage, and a recycling bin. As the core management unit of the 3D visualization online multi-cloud exhibition hall system, the workbench module is used to implement the full life cycle management of user works and material resource scheduling, specifically including work editing, material classification storage, and a recycling bin. This module interacts with the hotspot configuration module, the navigation management module, and the multi-theme module through a structured data interface. The work editing function is responsible for version control and collaborative editing of scene configuration files. The material classification storage function realizes the standardized processing and efficient retrieval of multimedia resources; The Recycle Bin function provides a two-stage data management mechanism of logical deletion and physical cleanup.

[0022] In some embodiments, the work editing function is built based on a distributed version control system, and its technical implementation is as follows: Scene configuration file format: It is defined in JSON structure and includes scene identifier (scene_id), camera parameters (coordinates, rotation angle), hotspot coordinate list and navigation node sequence; Version control mechanism: Git-LFS (Large File Storage) records incremental data for each edit operation, and the difference file hash value is stored in the patch file generated by gitdiff. For example, when a user modifies the camera angle parameters, only the changed camera_params field is stored. When rolling back a version, the patch is merged to restore the historical state.

[0023] In a possible implementation, the collaborative editing function resolves conflicts among multiple concurrent editors through an operation transformation algorithm (OT).

[0024] Specifically, when two users modify the hotspot coordinates of the same scene at the same time, the system automatically detects the conflicting areas and generates merge suggestions. The conflict resolution rules are defined as: The coordinates of user A and user B are both three-dimensional vectors (x, y, z).

[0025] Specifically, the material classification storage function supports standardized management of seven types of resource formats, including PDF documents, audio files (MP3 / WAV), video files (MP4), panoramic images, 3D models (glTF2.0), SVG vector icons, and Markdown text. Its technical details include: File preprocessing process: 3D model conversion: Convert FBX, OBJ and other formats to glTF2.0, and perform Draco compression through the gltf-pipeline tool. The compression rate calculation formula is: LOD generation: Creates a multi-level detail version of the 3D model, where the number of low-polygon faces is dynamically calculated based on the original face count, and the simplified scale factor β is defined as: Resource retrieval optimization: An inverted index is built based on Elasticsearch, supporting label-based queries (such as "material: metal" and "resolution: 4K"). Query response time is controlled within 50ms through sharding technology. The sharding rules are as follows: The recycle bin function adopts a two-stage mechanism of logical deletion and physical cleanup. The specific implementation methods include: Logical deletion phase: After the user performs a deletion operation, the file metadata is marked as is_deleted = 1, the physical storage path remains unchanged, and the retention period is fixed at 15 days (1296000 seconds); Physical cleanup phase: A scheduled background task scans the deleted_time field daily and irreversibly deletes expired files. The disk space release algorithm is as follows: In one possible implementation, the file recovery function ensures data consistency through a write-ahead log (WAL). For example, when recovering a deleted PDF file, the system performs the following atomic operations: Retrieve the original path, permissions, and version information of the file from the WAL log; Reset the is_deleted field to 0; Reindex the files into the Elasticsearch database.

[0026] The workbench module works in conjunction with other system modules in the following ways: Hotspot configuration module: calls the material library interface of the workbench module to obtain metadata (such as file size and duration) of PDF, audio, and video files, and downloads resources through the HTTP protocol; Navigation management module: Export the scene configuration file (JSON format) from the workbench module, parse the timeline_nodes field to generate the initial navigation timeline; Multi-theme module: Dynamically obtain theme resource packages from the workbench module through an on-demand loading interface (such as UnityAddressables), and the loading priority is dynamically adjusted according to user operations.

[0027] For example, when a user uploads an MP4 video file in the hotspot configuration module, the workbench module performs the following operations: Verify file format (H.264 encoding) and resolution (≥1080P); Generate video thumbnails (resolution 256×144) and store them in the library; Returns the CDN accelerated download link and metadata (duration, frame rate) of the video file.

[0028] Please refer to Figure 2 , a hotspot configuration module, connected to the workbench module and receiving its material data, for realizing multimodal configuration of hotspot icons and card information and jump link management; The hotspot configuration module connects to the workbench module via a data interface, receiving its categorized and stored resource assets (including PDF documents, audio and video files, and 3D models). It also implements multimodal configuration of hotspot icons and card information, as well as jump link management. This module generates interactive content based on hierarchical configuration rules and interacts with the navigation management module and rendering engine module via standardized protocols, ensuring dynamic adaptation of information display and interactive logic within the user's virtual scene.

[0029] In some embodiments, the text configuration unit supports a multi-level text structure of main title, subtitle, and body text, and its technical implementation includes the following core logic: Hierarchical text rendering: Use Markdown syntax to define the text structure, and generate a DOM tree after parsing, where the main title node corresponds to HTML <h1> Label, subtitle corresponding< / h1> <h2>Tags, corresponding to the text content Label.

[0030] Word-by-word printing animation control: total animation duration T animation The number of characters n and the word interval t char Determine, the calculation formula is: T animation =n·t char (t char ∈[0.1,1.0] seconds).

[0031] Specifically, when the user sets t char = 0.2 seconds and the number of text characters n = 50, the total animation duration is 10 seconds.

[0032] Transparency gradual effect: the transparency value α is linearly interpolated from the initial value 0 to 1, and the interpolation step Δα is based on the display time T fade calculate: For example, if T fade =2 seconds and the frame rate is 60FPS, then the transparency increment per frame is 1 / (2×60)≈0.0083.

[0033] In a possible implementation, the file plus text configuration unit realizes the integrated display of multimedia resources and text descriptions in the following manner: PDF file embedding and rendering: Render the PDF page to the Canvas canvas, and scale the page to the canvas width W. canvas The original width of the PDF is W pdf Dynamic calculation: For example, when the canvas width is 800px and the original PDF width is 1600px, the scaling factor is 0.5, ensuring that the content is fully displayed.

[0034] Streaming playback control: Audio / video files use the HLS protocol to achieve adaptive bit rate switching, and the resolution selection is based on the network bandwidth BB and threshold B threshold : If the real-time bandwidth is 1500Kbps, select standard definition resolution to ensure smooth playback.

[0035] Text description panel positioning: The screen coordinates (x, y) of the text panel are based on the hotspot icon position (x hotspot ,y hotspot ) and offset Δx, Δy calculation: x = x hotspot +Δx,y=y hotspot +Δy; Among them, x hotspot 、y hotspot The coordinates of the hotspot icon in the screen coordinate system are in pixels (px); Δx and Δy are the horizontal / vertical offsets of the text panel relative to the hotspot icon. The default value is 10 pixels and is adaptively adjusted according to the screen resolution.

[0036] The offset can be adaptively adjusted according to the screen resolution. For example, in portrait mode on mobile devices, Δy is increased to 20px to avoid occlusion.

[0037] Specifically, the jump link control unit supports the configuration and execution of four types of jump logic: Scene jump: Call the Unity engine's SceneManager.LoadSceneAsync() method to asynchronously load the target scene. The loading progress is synchronized to the rendering engine module through the event callback function.

[0038] Custom page jump: The HTML page size is scaled according to the device screen ratio. The scaling factor k is calculated as follows: Among them, W device 、H device The width and height of the user's device screen are in pixels (px); W page 、H page is the original design size of the custom HTML page, in pixels (px); k is the page scaling factor, which ensures that the content fits the device screen.

[0039] For example, if the device screen is 1080×1920px and the page design size is 1200×1600px, the zoom factor is 0.9.

[0040] Third-party link jump: Obtain user authorization token through OAuth2.0 protocol, the token is valid for T token Fixed at 3600 seconds, and automatically refreshed after timeout.

[0041] Platform content jump: Connected with the DICTSHOP platform, product data is transmitted through Protobuf protocol encoding, and rendered to the 3D interface after decoding. The product model scaling ratio is uniformly 0.8 times.

[0042] The hotspot configuration module interacts with other system modules in the following ways: Get materials from the workbench module: Pull metadata (such as file size and duration) of PDF, audio, and video files through HTTP GET requests, and download resources through CDN accelerated links.

[0043] Pass parameters to the navigation management module: set the display time of the hotspot card to T card The animation parameters are encapsulated as JSON instructions and aligned to the global timeline via the timeline synchronization unit.

[0044] Linked with the rendering engine module: The jump link configuration instruction is passed to the interactive response unit through the event bus, triggering scene switching or page loading operations.

[0045] For example, when a user configures a hotspot containing a PDF file and a word-by-word printing animation, the module executes the following process: Download PDF files from the workbench module and parse the pages; According to the number of characters n = 100 and the word interval t char = 0.3 seconds, calculate the total animation duration T animation =30 seconds; The PDF scaling factor s=0.6 and the animation parameters are sent to the rendering engine module.

[0046] Please refer to Figure 3 , a navigation management module, connected to the hotspot configuration module and receiving its configuration rules, for synchronously controlling the camera movement duration of the navigation node, the hotspot card display and hiding logic, and the media resource playback based on the timeline; The navigation management module connects to the hotspot configuration module via a data bus, receiving its configuration rules (including hotspot card display parameters, jump link logic, and media resource metadata). Based on a timeline synchronization mechanism, it implements camera trajectory control for navigation nodes, hotspot card display and hiding logic, and media playback progress alignment. This module uses a dynamic interpolation algorithm and error compensation mechanism to ensure the timing consistency of all elements in the virtual tour process, with a global error strictly limited to within 0.1 seconds.

[0047] In a possible implementation, the navigation node control unit generates a camera motion trajectory through the following steps: Control point calculation: According to the navigation node sequence set by the user (starting point P start , end point P end ), two intermediate control points P1 and P2 are automatically generated, and their coordinate calculation formula is: in, is the unit direction vector from the starting point to the end point; d offset The control point offset, the default value is the path length D path 30% (ie 0.3D path ).

[0048] Path length calculation: cubic Bezier curve path length D path Approximate calculation by numerical integration with a step size of Δt = 0.01: Among them, t i is the discrete sampling point parameter of numerical integration, with a total of 100 equally spaced points covering the interval [0,0.99]; ||B(t i +Δt)-B(t i )|| is the adjacent sampling point t i and t i +Δt, where: Among them, (x i ,y i ,z i ) is B(t i )’s three-dimensional coordinate values; B(t) is the three-dimensional coordinate value of the cubic Bezier curve at parameter t, which is generated by interpolation of control points P0, P1, P2, and P3. The calculation formula is: B(t) = (1-t) 3 P0+3t(1-t) 2 P1+3t 2 (1-t)P2+t 3 P3(t∈[0,1]).

[0049] Specifically, the node switching time T switch According to the scene complexity coefficient k complexity Dynamic adjustment, the calculation rules are as follows: Among them, k complexity k is the scene complexity coefficient, ranging from 1.0 to 3.0, and is calculated by weighting the following factors: complexity =1.0+0.5·N hotspot +0.3·L audio ; N hotspot is the number of hotspots activated in the current scene; L audio is the total duration of the associated audio file (in seconds), normalized to the interval [0,1]; v base The base speed of the camera is fixed at 2 m / s.

[0050] Example: If there are 5 hotspots in the scene and the audio duration is 10 seconds, then: k complexity =1.0+0.5·5+0.3·(10 / 60)≈3.55 (because it exceeds the upper limit, take 3.0); In the timeline synchronization unit, the hotspot card displays the duration T card Associated with audio duration T audio The matching rules are: Audio duration parsing: extract duration from audio file header information (MP3 files are parsed through ID3 tags, and WAV files are read through RIFF chunks); Forced display of lower limit: When T audio When the display time is ≤5 seconds, the card display duration is fixed at 5 seconds to avoid incomplete information display.

[0051] Specifically, timeline synchronization errors are eliminated by: Linear interpolation compensation: When a time deviation Δt is detected between the appearance and disappearance of a hotspot card, its transparency interpolation weight w is adjusted: If Δt = 0.05 seconds and T card = 8 seconds, then w = 0.00625, and the transparency of the next frame is adjusted to: α next =α current ±w·(α end -α start ); Among them, α current is the current transparency value; α end is the target transparency value; α start is the initial value of transparency; w is the interpolation weight (dimensionless), which is used to adjust the rate of change of transparency; Dynamic adjustment of audio playback rate: When the audio playback progress deviation Δt audio When the time is greater than 0.1 seconds, adjust the playback rate r: Where Δt audio The audio playback progress deviation, in seconds; sign(Δt audio ) is the deviation direction (+1 means leading, -1 means lagging); the rate adjustment range is limited to [0.9, 1.1] to avoid pitch distortion.

[0052] Example: If the audio is ahead by 0.2 seconds and the total duration is T audio = 10 seconds, then: r=1-min(0.1,0.2 / 10)=1-0.02=0.98; Data interaction with the hotspot configuration module: Receive configuration parameters: Parse the data packet sent by the hotspot configuration module through the Protobuf protocol. The fields include: hotspot_id: hotspot unique identifier (UUID format); display_rules: Display rules (including T card , associated audio path); jump_link: jump link configuration (scene ID / URL).

[0053] Data verification: Verify the existence of the audio file and user permissions (read permission verification is performed through SHA-256 signature).

[0054] Passing instructions to the rendering engine module: Bezier curve parameter transmission: control point coordinates P0~P3 and node switching time T switch Encapsulated as a binary instruction stream, the structure is as follows: [Header: 4 bytes] [P0.x: 4 bytes] [P0.y: 4 bytes] [P0.z: 4 bytes]... [T_switch: 4 bytes] Real-time synchronization signal: Send a timeline synchronization signal (including global timestamp T) every 0.1 seconds via the WebSocket protocol global ).

[0055] Exception handling mechanism: Path generation failed: If the Bezier curve control point calculation is abnormal (such as path length D path =0), then switch to linear interpolation algorithm: P(t)=P start +t·(P end -P start )(t∈[0,1]); Audio parsing failed: If the audio file header information is damaged, re-download the file from the workbench module and try again (up to 3 times).

[0056] Please refer to Figure 4 , a multi-theme module, connected to the workbench module and the navigation management module, for loading multiple theme configuration packages and dynamically switching scene content according to user instructions; The multi-theme module is data-connected to the workbench module and the navigation management module. It loads multiple theme configuration packages and dynamically switches scene content based on user input, enabling the overlay and switching of diverse display themes within the same virtual scene. This module utilizes layered rendering technology and a dynamic resource scheduling mechanism to ensure smooth and seamless theme switching, seamlessly integrating with the navigation management module's camera tracking and hotspot card display and disappearance logic.

[0057] In some embodiments, the theme container unit uses on-demand loading technology to manage independent theme configuration packages. Its technical implementation includes the following core logic: Theme package structure: Each theme package is a ZIP compressed file, which contains a scene configuration file (JSON format), a 3D model (glTF2.0), a material atlas, and metadata (such as theme ID and version number).

[0058] Dynamic loading rules: When the user triggers a theme switch, the target theme package is downloaded from the Workbench module's asset library, decompressed, and only the assets in the visible area are loaded. For example, if the user switches to the "Dark Theme," low-polygon (LOD) models and compressed textures are loaded first.

[0059] Memory pre-allocation: based on the theme package size S theme (Unit: MB) Pre-allocated memory space, calculated as: (Unit: MB) Pre-allocated memory space, calculated as: Among them, 1.2 is the redundancy coefficient, which is used to avoid frequent memory expansion; S theme The total size of the theme configuration package in megabytes (MB); prealloc The size of the pre-allocated memory space in megabytes (MB).

[0060] Specifically, the hybrid rendering unit realizes multi-theme overlay display through layered rendering and transparency mixing. Its technical details include: Independent render layer assignment: The 3D model of each theme is assigned to an independent rendering layer (RenderLayer), and the layer ID is bound to the theme ID.

[0061] The transparency of the rendering layer α is dynamically adjusted by user configuration or scene lighting conditions. For example, in bright mode, the transparency of the foreground layer α = 0.8 and the background layer α = 0.3.

[0062] Pixel color mixing calculation: Final pixel color C final By foreground color C foreground With background color C background Alpha channel weighted calculation: C final =C foreground α+C background (1-α)(α∈[0,1]); Among them, C foreground is the foreground color (RGBA format, including transparency channel); C background is the background color (RGBA format); α is the foreground transparency, 0 is completely transparent, and 1 is completely opaque.

[0063] Global illumination dynamic adjustment: Ambient light intensity I ambient Configure according to the theme style, for example: Lighting parameters are passed to the GPU rendering pipeline in real time through ShaderUniform variables.

[0064] In one possible implementation, the resource release unit uses an LRU (Least Recently Used) algorithm to manage inactive topic resources. The technical details include: Active status mark: Each theme package records the last access timestamp T last , the timestamp is updated when the user operates or the navigation node switches.

[0065] Memory release rule: When the system memory usage exceeds the threshold M threshold =80%, the longest unused theme package resources are released. The release priority calculation formula is: Among them, T current is the current system time, priority P release The higher the value, the higher the release priority.

[0066] Resource uninstallation process: Destroy the GPU resources (such as textures and vertex buffers) corresponding to the rendering layer; Remove the decompressed theme package data from memory; Preserve metadata and thumbnails for fast reloading.

[0067] The multi-theme module interacts with other modules of the system in the following ways: Get resources from the workbench module: Download the theme package via HTTPS, verify the digital signature (SHA-256), and then decompress it to a temporary directory; Collaborate with the navigation management module: receive navigation node switching events and dynamically adjust the transparency α of the theme rendering layer to match the camera movement speed; Send instructions to the rendering engine module: encapsulate the layered rendering parameters (such as layer ID, transparency) into a binary instruction stream and submit it through the GPU command queue.

[0068] Example: When the user switches from "Default Theme" to "Holiday Theme": The theme container unit downloads the holiday theme package (size 150MB) from the workbench module and pre-allocates 180MB of memory; The hybrid rendering unit creates a new rendering layer (layer ID = 2) and sets the initial transparency α = 0.5; The navigation management module synchronously updates the camera trajectory to ensure that the camera movement speed matches the animation rhythm of the festival theme.

[0069] Please refer to Figure 5 The rendering engine module is connected to the navigation management module and the multi-theme module, receives timeline instructions and theme rendering parameters, and executes 3D scene rendering and cross-platform interactive response.

[0070] The rendering engine module is data-connected to the navigation management module and the multi-theme module, receiving timeline synchronization instructions and theme rendering parameters to perform efficient 3D scene rendering and cross-platform interactive response. This module utilizes dynamic optimization algorithms and a layered rendering mechanism to ensure visual fidelity and real-time performance during virtual tours, while also supporting multi-device adaptation. The scene rendering unit, interactive response unit, and cross-platform output unit work in tandem, seamlessly integrating with the navigation node control and multi-theme switching logic of the aforementioned modules.

[0071] In some embodiments, the scene rendering unit optimizes rendering efficiency based on frustum culling and level of detail (LOD) technology, and its technical details include: LOD adaptive switching: model detail level switching threshold d threshold According to the user's perspective movement speed v user Dynamic adjustment, the calculation formula is: Among them, d threshold is the basic distance threshold, the default value is 10 meters; v user is the user's viewing speed (unit: m / s), calculated by camera displacement difference; v ref is the reference speed, fixed at 2 m / s.

[0072] Example: When v user =3 m / s, d threshold =10×(1+3 / 2)=25 meters, the distant model switches to the low-precision LOD level.

[0073] View frustum clipping priority calculation: model rendering priority P render Based on its spatial distance d from the camera model calculate: (∈=0.01 meter, prevention is zero); Models with higher priorities (closer distances) are rendered first to prevent distant models from occupying too many GPU resources.

[0074] Specifically, the interactive response unit implements the binding of user operations with system logic through the event bus mechanism, and its technical implementation includes: Hotspot card visibility control: Receive the display time parameter T sent by the navigation management module card , the card is controlled to gradually appear / fade through the transparency interpolation algorithm. The interpolation formula is: Where α(t) is the transparency value at time t.

[0075] Jump link trigger: When the user clicks a hotspot card, the corresponding interface is called according to the jump type (scene jump, third-party link, etc.): Scene jump: call the SceneManager.LoadSceneAsync() method of the Unity engine; Third-party links: Open the URL via the system browser and append the OAuth2.0 token verification.

[0076] In a possible implementation, the cross-platform output unit adapts to different terminals in the following manner: Web rendering: Using WebGL2.0 standard, 3D scenes are rendered to HTML5Canvas elements. The resolution is dynamically scaled according to the browser viewport. The scaling ratio is k web Calculated as: Among them, W native ×H native is the original design resolution (such as 1920×1080), W viewport ×H viewport The current viewport size.

[0077] Mobile Optimization: Touch operations are mapped to virtual joystick events, and the joystick sensitivity is adaptively adjusted according to the screen DPI: (DPI is device pixel density); The rendering resolution is downgraded to 75% of the native resolution (e.g., a 1080P device renders at 810P) to reduce GPU load.

[0078] The rendering engine module interacts with other system modules in the following ways: Receive instructions from the navigation management module: parse the timeline synchronization signal (including the global timestamp T global ), drives the hotspot card display and hiding animation and media playback progress; Get parameters from the multi-theme module: receive layered rendering instructions (such as rendering layer ID, transparency α) and synchronize them to the GPU rendering pipeline; Feedback status to the workbench module: reporting rendering error logs (such as Shader compilation failure) and performance indicators (frame rate, GPU occupancy).

[0079] Example: When the navigation management module triggers node switching: The scene rendering unit updates the frustum clipping range according to the camera position of the new node; The interactive response unit resets the transparency interpolation timer of the hotspot card; The cross-platform output unit dynamically adjusts the resolution to ensure smooth viewing on mobile devices.

[0080] The examples of this specific embodiment are all preferred embodiments of this application and are not intended to limit the scope of protection of this application. Identical components are represented by the same reference numerals. Therefore, any equivalent changes made based on the structure, shape, and principle of this application should be included in the scope of protection of this application. < / h2>

Claims

1. A 3D visualization online multi-cloud exhibition hall system, characterized by: include: The workbench module is used to manage user works and material resources, including work editing, material classification storage and recycle bin functions; A hotspot configuration module, connected to the workbench module and receiving its material data, is used to implement multimodal configuration of hotspot icons and card information and jump link management; A navigation management module is connected to the hotspot configuration module and receives its configuration rules, and uses timeline-based synchronization to control the camera movement duration of navigation nodes, hotspot card display and hiding logic, and media resource playback; A multi-theme module, connected to the workbench module and the navigation management module, is used to load multiple theme configuration packages and dynamically switch scene content according to user instructions; The rendering engine module is connected to the navigation management module and the multi-theme module, receives timeline instructions and theme rendering parameters, and performs 3D scene rendering and cross-platform interactive response.

2. The 3D visualization online multi-cloud exhibition hall system according to claim 1 is characterized in that: The hotspot configuration module includes: The text configuration unit supports a multi-level text structure of main title, subtitle and body text, and can be configured with word-by-word printing animation effects; File plus text configuration unit, supports embedding PDF, audio, and video files, and each file is associated with an independent text description panel; The jump link control unit supports four types of link configurations: scene jump, custom page, third-party link and platform content jump.

3. The 3D visualization online multi-cloud exhibition hall system according to claim 1 is characterized in that: The navigation management module includes: The navigation node control unit dynamically adjusts the node switching duration based on the time controller of the VR scene engine, and the lens motion trajectory is generated by the cubic Bezier curve interpolation algorithm; The timeline synchronization unit globally aligns the camera trajectory of the navigation node, the appearance and disappearance time of the hotspot card, and the playback progress of the media resource, with an alignment error of less than 0.1 second.

4. The 3D visualization online multi-cloud exhibition hall system according to claim 3 is characterized in that: The display duration T of the hotspot card in the timeline synchronization unit card According to the associated audio duration T audio Dynamic calculation, the calculation rules are:

5. The 3D visualization online multi-cloud exhibition hall system according to claim 1 is characterized in that: The multi-topic modules include: Theme container unit, which stores independent theme configuration packages and dynamically loads resources through on-demand loading technology; The hybrid rendering unit assigns 3D models of different themes to independent rendering layers and realizes superimposed display through transparency blending; the resource release unit uses the LRU algorithm to release the memory resources of inactive themes according to the active status of the theme.

6. The 3D visualization online multi-cloud exhibition hall system according to claim 5 is characterized in that: In the hybrid rendering unit: The final pixel color of the transparency blend is calculated by weighting the alpha channels of the foreground color and the background color; Global illumination parameters are dynamically adjusted based on the theme, including high ambient light intensity in light mode and low ambient light intensity in dark mode.

7. The 3D visualization online multi-cloud exhibition hall system according to claim 6 is characterized in that: The weighted calculation formula for the alpha channel of the foreground color and the background color is as follows: C final =C foreground ·a+C background ·(1-a); Among them, C foreground Indicates the foreground color, C background Represents the background color, and α is the transparency value of the foreground.

8. The 3D visualization online multi-cloud exhibition hall system according to claim 1 is characterized in that: The rendering engine module includes: The scene rendering unit optimizes the rendering accuracy of 3D models based on frustum clipping and LOD technology, where the frustum clipping priority is calculated based on the spatial distance between the model and the camera; The interactive response unit triggers the display of hotspot cards, jump logic, and media resource playback through the event monitoring mechanism; Cross-platform output unit adapts rendering results to web, mobile and desktop terminals.

9. The 3D visualization online multi-cloud exhibition hall system according to claim 8, characterized in that: In the scene rendering unit: The threshold for switching the model's level of detail in LOD technology is adaptively adjusted based on the user's viewing speed. The priority of frustum clipping is calculated as follows: the closer the model is in space, the higher the priority.

Images