Urban and rural planning virtual scene construction method and system based on game engine

Abstract

The application relates to the technical field of urban and rural planning, and discloses a method and system for constructing a virtual scene of urban and rural planning based on a game engine, which integrates geographical space data, planning design data, real scene perception data and human social data of a target urban and rural area, carries out format standardization, noise elimination and coordinate calibration on various types of data, and obtains a standardized fusion data set; wherein the geographical space data comprises a DEM (Digital Elevation Model), a tilt photography model and an orthographic image, the application designs a complete system architecture, covers full-process modules such as data collection and preprocessing, scene construction, interactive deduction, verification output and data storage, adopts a distributed storage architecture, supports efficient storage and parallel access of massive data, the system continuously runs for 24 hours without failure, the data access speed is less than or equal to 50 ms, and the stability, safety and availability of the system are guaranteed; compared with an existing system, the application is more perfect in function, more superior in performance, and is more suitable for large-scale popularization and application in the field of urban and rural planning.

Description

Technical Field

[0001] This invention relates to the field of urban and rural planning technology, and in particular to a method and system for constructing virtual scenes for urban and rural planning based on a game engine. Background Technology

[0002] Urban and rural planning is a core task for coordinating urban and rural spatial layout, optimizing resource allocation, and promoting integrated urban and rural development. Virtual scene construction, as an important technical means for urban and rural planning, can transform abstract planning schemes into intuitive three-dimensional visual scenes, providing strong support for planning demonstration, public participation, and decision-making. Currently, virtual scene construction technologies for urban and rural planning are mainly divided into two categories: one is the construction method based on traditional 3D modeling software (such as 3ds Max and ArcGIS), and the other is a simplified construction method based on conventional game engines.

[0003] However, existing technologies have many shortcomings and are unable to meet the high precision, high efficiency, and high interactivity requirements of modern urban and rural planning. Traditional construction methods based on 3ds Max + ArcGIS rely on manual modeling and data processing, which is not only labor-intensive and time-consuming, but also results in low scene detail reproduction, lack of dynamic interaction and planning simulation functions, and inability to intuitively present the implementation effects and future development trends of planning schemes. At the same time, various multi-source data (geospatial, planning and design, reality perception, etc.) have inconsistent formats and are difficult to integrate, resulting in low data processing efficiency and insufficient accuracy. This leads to poor geographical matching between virtual scenes and real areas, making it difficult to support accurate planning decisions.

[0004] Existing game engine-based construction methods mostly use conventional engines such as Unity, failing to fully utilize advanced rendering technologies (such as Nanite and Lumen), making it difficult to balance scene realism and rendering efficiency. Furthermore, they lack professional multi-source data fusion processing mechanisms and have not introduced procedural content generation (PCG) technology, so the construction of ecosystem scenes still relies on manual deployment, resulting in low efficiency. The planning and simulation functions are simple, only supporting basic parameter adjustments, and cannot achieve multi-scheme comparison, future scene prediction, or immersive interaction, making it difficult to meet the diverse needs of planning demonstration and public participation.

[0005] In addition, existing systems often suffer from imperfect architecture, poor compatibility with multiple terminals, and insecure data storage, which further limits their promotion and application in the field of urban and rural planning.

[0006] Therefore, developing a method and system for constructing virtual scenes for urban and rural planning based on a game engine, which can achieve efficient fusion of multi-source data, accurate scene construction, dynamic interactive simulation, and has a sound architecture and strong adaptability, has become an urgent technical problem to be solved. Summary of the Invention

[0007] In order to overcome the shortcomings of the existing technology, one of the objectives of this invention is to provide a method and system for constructing virtual scenes for urban and rural planning based on a game engine.

[0008] One of the objectives of this invention is achieved through the following technical solution: A method for constructing a virtual scene for urban and rural planning based on a game engine includes the following steps: S1. Multi-source data acquisition and preprocessing: Collect geospatial data, planning and design data, real-scene perception data, and human and social data of the target urban and rural areas. Standardize the format, remove noise, and calibrate the coordinates of various types of data to obtain a standardized fusion dataset. Among them, the geospatial data includes DEM digital elevation model, oblique photogrammetry model, and orthophoto map, which are acquired by UAV oblique photogrammetry technology combined with laser scanning technology. The real-scene perception data includes real-time traffic flow, environmental monitoring data, and population flow heat map data, which are collected in real time through Internet of Things sensing devices. S2. Game Engine Scene Initialization: Based on the game engine, a basic framework for a virtual scene is built, and scene rendering parameters are configured, including lighting parameters, texture parameters, and level of detail parameters. Using the game engine's world partitioning system, the target urban and rural areas are automatically divided into several grid partitions according to geographical coordinates, a unified coordinate system is established, and the automatic loading and unloading of scene grids is realized, solving the problem of lag when loading ultra-large scenes. S3. Terrain and Ecological Scene Construction: The preprocessed DEM digital elevation model is imported into the game engine, and terrain texture mapping is completed by combining orthophoto maps to generate a 3D terrain model that accurately matches the real geographic coordinates; based on procedural content generation (PCG) technology, the surface material is batch filled into the terrain model according to the land cover classification results, and ecological elements such as vegetation and water bodies are automatically placed to generate a virtual ecological environment scene; among them, the vegetation placement is combined with the vegetation type data of the target area to achieve accurate distribution and growth status simulation of different vegetation; S4. Precise Integration of Planning Elements: The pre-processed planning and design data is converted into a model format that the game engine can recognize. Lightweight processing and posture correction are performed on the planning element models such as buildings, roads, and pipelines to avoid scene lag caused by model redundancy. The processed planning element models are imported into the ecological environment virtual scene according to the real geographic coordinates to achieve seamless integration of planning elements with terrain and ecological scene, and generate the initial urban and rural planning virtual scene. S5. Dynamic Interaction and Deduction Optimization: Based on the real-time rendering capabilities of the game engine, it integrates a dynamic lighting system and a physical simulation system to realize the dynamic simulation of changes in light and shadow, weather, and vegetation growth in virtual scenes; it builds a planning and deduction module to support real-time adjustment of planning schemes, comparison of multiple schemes, and deduction of future scenes, while providing immersive interactive operations to support users to roam, measure, annotate, and perform collision detection in virtual scenes; S6. Scene Verification and Output: Establish a multi-dimensional verification index system to verify the geographic accuracy, planning rationality, and interactive smoothness of the virtual scene. If the verification fails, return to the corresponding step for adjustment. After the verification passes, output a standardized urban and rural planning virtual scene file, which supports multi-terminal display and data sharing.

[0009] Furthermore, the data preprocessing described in step S1 specifically includes: using AI-CG dual-modal data fusion recognition technology to intelligently decipher the planning and design drawings, accurately extracting internal building structure data such as walls, beams, columns, and pipelines; denoising and simplifying the oblique photography model, retaining key details while reducing model complexity; converting all data to the WGS84 coordinate system through a coordinate transformation algorithm to ensure data spatial consistency; and supplementing missing human and social data using interpolation algorithms to improve the completeness of the dataset.

[0010] Furthermore, the game engine mentioned in step S2 adopts Unreal Engine 5 (UE5), and the rendering parameter configuration specifically includes: using UE5's Nanite micropolygon geometry technology to achieve infinite detail rendering of planned element models without manually setting the Level of Detail (LOD), avoiding visual jumps when switching models; using the Lumen global illumination system to achieve real-time reflection of light and dynamic lighting changes in the virtual scene, simulating the light and shadow effects under different time periods and weather conditions, and improving the realism of the scene; configuring virtual texture technology to support efficient loading and streaming of massive high-resolution textures, achieving ultra-high-definition scene presentation under limited video memory.

[0011] Furthermore, the specific applications of the programmatic content generation (PCG) technology described in step S3 include: setting parameters such as vegetation type, density, and growth cycle based on the ecological environment data of the target area, and automatically generating vegetation communities that conform to the real environment in batches through the PCG toolset; automatically assigning corresponding surface materials according to terrain slope, altitude, and other information to achieve differentiated presentation of different terrains such as mountains, plains, and water bodies; and supporting users to customize PCG rules and flexibly adjust the presentation effect of ecological scenes.

[0012] Furthermore, the planning simulation module in step S5 specifically includes: supporting parameterized adjustment of planning schemes, allowing users to modify planning parameters such as building height, road width, and green space area in real time, with scene updates synchronized; providing a multi-scheme comparison function, which can load multiple planning schemes simultaneously and visually present the differences between different schemes; integrating future scene simulation algorithms, which simulate urban and rural development changes in the target area over the next 5-20 years based on historical data and development trends, providing data support for planning decisions; and supporting VR / AR device access for immersive scene roaming and planning operations.

[0013] Furthermore, the multi-dimensional verification index system mentioned in step S6 includes: geographic accuracy index, requiring the coordinate deviation between the virtual scene and the real area to be no more than 0.5 meters; planning rationality index, including the degree of conformity between parameters such as building density, plot ratio, and road red line and planning specifications; interaction smoothness index, requiring the scene roaming frame rate to be no less than 60fps and the model loading delay to be no more than 1 second; the verification process adopts a combination of automated verification and manual verification to improve verification efficiency and accuracy.

[0014] A virtual scene construction system for urban and rural planning based on a game engine, comprising: The data acquisition and preprocessing module is used to collect geospatial data, planning and design data, real-scene perception data, and human and social data of the target urban and rural areas, and to perform format standardization, noise removal, coordinate calibration, and missing value supplementation on various types of data, outputting a standardized fusion dataset; the data acquisition and preprocessing module includes a UAV oblique photography unit, a laser scanning unit, an IoT sensing unit, and a data processing unit, the data processing unit integrating an AI-CG dual-modal data fusion recognition module and a coordinate transformation module; The game engine scene initialization module is used to build a basic framework for virtual scenes based on the game engine, configure scene rendering parameters, divide the target area into grids using a world partitioning system, establish a unified coordinate system, and realize the automatic loading and unloading of scene grids; the game engine scene initialization module integrates a rendering parameter configuration unit and a grid partitioning management unit. The terrain and ecological scene construction module is used to import the preprocessed DEM digital elevation model and orthophoto map into the game engine, complete the terrain texture mapping, and generate a 3D terrain model; based on procedural content generation (PCG) technology, it completes the surface material filling and ecological element placement to generate a virtual ecological environment scene. The planning element fusion module is used to convert planning and design data into a model format that the game engine can recognize, perform lightweight processing and posture correction on the planning element model, import the planning element model into the ecological environment virtual scene according to the real geographic coordinates, achieve seamless fusion, and generate the initial urban and rural planning virtual scene. The dynamic interaction and simulation module integrates a dynamic lighting system and a physical simulation system to simulate the dynamic changes of virtual scenes; it also builds a planning and simulation module to support real-time adjustment of planning schemes, comparison of multiple schemes, and simulation of future scenes; and provides an immersive interactive operation interface to support users in roaming, measuring, labeling, and collision detection. The scene verification and output module is used to establish a multi-dimensional verification index system to verify the geographic accuracy, planning rationality, and interaction smoothness of virtual scenes. Scenes that fail the verification are returned to the corresponding module for adjustment. After the verification is passed, standardized virtual scene files are output, which support multi-terminal display and data sharing. The data storage module is used to store the collected raw data, preprocessed standardized fusion datasets, virtual scene model files, planning scheme data, and verification result data, and supports fast data query, update, and backup.

[0015] Furthermore, the game engine uses Unreal Engine 5 (UE5). The rendering parameter configuration unit in the game engine scene initialization module is specifically used to configure Nanite micropolygon geometry parameters, Lumen global illumination parameters, and virtual texture parameters to achieve ultra-high-definition and high-realism scene rendering. The mesh partition management unit adopts a dynamic mesh loading algorithm, which automatically loads mesh partitions within the current field of view and unloads mesh partitions outside the field of view based on the user's operation position, thereby reducing system resource consumption.

[0016] Furthermore, the dynamic interaction and simulation module also includes a data visualization unit, which is used to overlay and display planning simulation data and real-world perception data in the form of charts, heat maps, etc. in the virtual scene, intuitively presenting the implementation effect of the planning scheme and the regional development status; the interactive operation interface supports mouse, keyboard, VR / AR devices and touch devices to meet the operation needs of different users.

[0017] Furthermore, the system also includes a multi-terminal adaptation module, which converts the output virtual scene file into a format recognizable by different terminals, supporting display and interaction on PCs, mobile devices, VR devices, and large-screen terminals; the data storage module adopts a distributed storage architecture, supporting efficient storage and parallel access of massive amounts of data, ensuring data security and availability.

[0018] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. This invention employs AI-CG dual-modal data fusion and recognition technology to achieve intelligent deciphering of planning and design drawings and accurate extraction of building internal structure data, with an extraction accuracy rate of over 98%. Combined with efficient noise reduction and simplification, seven-parameter coordinate transformation, and interpolation supplementation algorithms, it achieves standardized fusion of multi-source data, improving data integrity to over 99% and controlling coordinate transformation error within 0.02m. Compared with traditional methods, data preprocessing time is reduced by more than 50%, effectively solving the problems of low data processing efficiency, poor accuracy, and insufficient integrity in existing technologies.

[0019] 2. This invention is based on Unreal Engine 5 (UE5), and uses Nanite micropolygon geometry technology and Lumen global illumination system to achieve ultra-high-definition and high-realism scene rendering without the need for manual LOD settings. It introduces PCG technology to achieve automatic allocation of surface materials, batch deployment of vegetation, and dynamic simulation of water bodies. The scene construction time is reduced by more than 75% compared with traditional methods, and the scene realism reaches more than 92%, solving the problems of low scene construction efficiency, poor realism, and inaccurate placement of ecological elements in existing technologies.

[0020] 3. This invention integrates a dynamic lighting and physical simulation system to achieve real-time simulation of dynamic effects such as light and shadow, weather, and vegetation growth, with a dynamic delay of ≤0.1 seconds; it also builds a fully functional planning and simulation module that supports real-time adjustment of planning parameters, comparison of multiple schemes, and scenario simulation for the next 5-20 years (simulation accuracy ≥95%), and supports VR immersive interaction, meeting the diversified needs of planning demonstration, public participation, and decision support, and making up for the shortcomings of existing technologies in terms of insufficient dynamic interaction and simple simulation functions.

[0021] 4. This invention establishes a multi-dimensional verification index system and adopts a combination of automated and manual verification, achieving a verification pass rate of over 98.5% and an average coordinate deviation of ≤0.32m, ensuring the geographical accuracy and planning rationality of the virtual scene; it also sets up a multi-terminal adaptation module, which can convert virtual scene files into formats recognizable by PCs, mobile devices, VR devices, and large-screen terminals, with an adaptation pass rate of 100%, solving the problems of low verification efficiency and poor multi-terminal compatibility in existing technologies.

[0022] 5. This invention designs a complete system architecture, covering the entire process modules such as data acquisition and preprocessing, scenario construction, interactive simulation, verification output, and data storage. It adopts a distributed storage architecture, supports efficient storage and parallel access of massive amounts of data, and the system can run continuously for 24 hours without failure. The data access speed is ≤50ms, ensuring the stability, security, and availability of the system. Compared with existing systems, it has more complete functions, better performance, and is more suitable for large-scale application in the field of urban and rural planning.

[0023] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described in detail below with reference to the accompanying drawings. Attached Figure Description

[0024] Figure 1 This is a flowchart of this embodiment. Detailed Implementation

[0025] The present invention will now be further described in conjunction with the accompanying drawings and specific embodiments. It should be noted that, without conflict, the various embodiments or technical features described below can be arbitrarily combined to form new embodiments.

[0026] It should be noted that when a component is described as "fixed to" another component, it can be directly on the other component or may have a component in between. When a component is considered "connected to" another component, it can be directly connected to the other component or may have a component in between. When a component is considered "set on" another component, it can be directly set on the other component or may have a component in between. The terms "vertical," "horizontal," "left," "right," and similar expressions used in this document are for illustrative purposes only.

[0027] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0028] 1. Overview of Implementation Examples This embodiment uses a county-level urban-rural fringe area (118°20′-118°35′E, 32°10′-32°25′N) as the target area, with a total area of ​​approximately 80 km². 2 The system encompasses three terrain types: mountains, plains, and water areas. It includes four functional zones: urban built-up areas, rural settlements, industrial parks, and ecological protection zones. The planning goal is to construct a virtual urban and rural planning scenario that is "ecologically livable and integrates industry and city," for use in planning scheme demonstration, public participation, and decision-making support. This embodiment uses Unreal Engine 5 (UE5) as the core game engine to build an urban and rural planning virtual scenario construction system based on this invention. The system fully executes the construction method described in claim 1, verifying the feasibility, advancement, and practicality of this invention.

[0029] 2. Specific Implementation Steps 2.1 Multi-source data acquisition and preprocessing Four types of core data were collected from the target area, as detailed below: (1) Geospatial data: The DJI Phantom 4RTK drone was used for oblique photography, and the data was acquired by the RIEGLVZ-400 laser scanner to generate an orthophoto map with a resolution of 0.1m, an oblique photography model with an accuracy of 0.05m, and a DEM digital elevation model with a resolution of 10m. (2) Planning and design data: Collect the urban and rural master plan drawings (CAD format), building construction drawings, road and pipeline design drawings, ecological protection red lines and basic farmland protection zone delineation documents for the area from 2025 to 2035; (3) Real-world perception data: 30 IoT sensor nodes are deployed in the target area, including traffic flow detectors (12), environmental monitors (10, monitoring PM2.5, temperature and humidity), and pedestrian flow thermal collectors (8), to collect data in real time at a sampling frequency of 1 time / 5 minutes; (4) Humanities and social data: Collect population statistics (2020-2024), GDP data, industrial distribution data and resident travel survey data for the region.

[0030] The data preprocessing process was strictly performed according to claim 2: AI-CG dual-modal data fusion and recognition technology was used to intelligently decipher CAD planning drawings, accurately extracting internal building structure data such as walls, beams, columns, and pipelines, with an extraction accuracy of 98.2%; Gaussian filtering algorithm was used to denoise the oblique photogrammetry model, combined with QuadricEdgeCollapseDecimation algorithm for simplification, reducing the number of triangular faces in the model from 120 million to 28 million while retaining more than 95% of the key details, and compressing the model size from 8.6GB to 1.9GB; a seven-parameter coordinate transformation algorithm was used to uniformly convert all data to the WGS84 coordinate system, with the coordinate transformation error controlled within 0.02m; for the missing rural population data in 2023, a linear interpolation algorithm was used to supplement it, improving the data integrity to 99.5%, and finally a standardized fusion dataset with a total size of 12.3GB was obtained.

[0031] 2.2 Game Engine Scene Initialization The virtual scene framework is built based on Unreal Engine 5 (UE5), and the rendering parameters are configured according to claim 3: enable Nanite micropolygon geometry technology, disable manual LOD settings, and realize infinite detail rendering of planned element models; enable Lumen global illumination system, set the number of light bounces to 4, and simulate the light and shadow effects of four time periods: sunrise, noon, sunset, and night, as well as three weather conditions: sunny, cloudy, and rainy; configure virtual texture technology, set the texture resolution to 4K, enable texture streaming function, and control the video memory usage to within 8GB.

[0032] By utilizing UE5's world partitioning system, the target area is automatically divided into 80 grid partitions with a size of 1km×1km. A unified WGS84 coordinate system is established, and a dynamic grid loading algorithm is adopted to realize the automatic loading and unloading of scene grids. When the user's roaming speed does not exceed 10m / s, the grid loading delay is controlled within 0.3 seconds, thus solving the problem of lag when loading ultra-large scenes.

[0033] 2.3 Construction of Topography and Ecological Scenes The preprocessed DEM digital elevation model is imported into UE5, and terrain texture mapping is completed by combining it with orthophoto maps to generate a 3D terrain model that accurately matches the real geographic coordinates, with the terrain elevation error controlled within 0.1m. Based on the PCG toolset of UE5, ecological scene construction is carried out according to the PCG technology application method of claim 4: (1) Vegetation layout: Based on the vegetation type data of the target area (trees: poplar, willow; shrubs: Amorpha fruticosa, holly; herbs: Bermuda grass, Kentucky bluegrass), different vegetation densities were set (20-30 trees / 100m²). 2 50-80 shrubs per 100m 2 With full coverage of herbaceous plants and a growth cycle of 5-30 years for trees and 3-10 years for shrubs, vegetation communities can be automatically generated in batches using the PCG toolset, increasing the generation efficiency by more than 85% compared to manual layout. (2) Surface material allocation: Based on the terrain slope (slope <5° is plain, 5°≤slope <25° is hill, slope ≥25° is mountain) and altitude (altitude <50m is water area, 50m≤altitude <200m is plain / hill, altitude ≥200m is mountain), the corresponding surface material is automatically allocated (plain: farmland, construction land; hill: forest, grassland; mountain: forest, rock; water area: wetland, tidal flat), to achieve differentiated presentation of different terrains; (3) Water body layout: Import the vector data of rivers and reservoirs in the target area, automatically generate water body models through PCG tools, configure water body physical parameters (flow velocity, transparency), simulate the flow effect of natural water bodies, and finally generate a virtual scene of ecological environment with a scene realism of over 92%.

[0034] 2.4 Precise integration of planning elements The preprocessed planning and design data (CAD format) was converted into FBX format recognizable by UE5 using 3ds Max software. Lightweight processing was carried out on the planning element models such as buildings, roads, and pipelines: the number of triangular faces in the building model was reduced by an average of 60% using a model simplification algorithm, the road model was simplified by 55%, and the pipeline model was simplified by 70%; the rotation angle and position coordinates of the model were calibrated using an attitude correction algorithm, and the attitude correction error was controlled within 0.03m.

[0035] The processed planning element model is imported into the ecological environment virtual scene according to the real geographic coordinates. The building model fits the terrain with a degree of 99.8%, the slope adaptation error of the road model is ≤0.5°, and the connection error between the pipeline network model and the buildings and roads is ≤0.1m. This achieves seamless integration of planning elements with terrain and ecological scene, generating the initial urban and rural planning virtual scene.

[0036] 2.5 Dynamic Interaction and Inference Optimization Based on UE5's real-time rendering capabilities, it integrates a dynamic lighting system and a physics simulation system: the dynamic lighting system can switch the lighting and shadow effects of different time periods and weather conditions in real time, with a lighting and shadow change delay of ≤0.1 seconds; the physics simulation system can simulate dynamic effects such as vegetation swaying in the wind, raindrops falling, and vehicles moving, with a physics simulation frame rate of ≥60fps.

[0037] Build a planning and simulation module, configured according to the functions in claim 5: (1) Parametric adjustment: Users can modify planning parameters such as building height (adjustment range 3-30m), road width (adjustment range 4-60m), and green area (adjustment range 5%-40%) in real time. The scene is updated synchronously with an update delay of ≤0.2 seconds. (2) Multi-scheme comparison: Supports loading three planning schemes at the same time (scheme 1: focusing on industrial development, scheme 2: focusing on ecological protection, scheme 3: integration of industry and city), and intuitively presents the differences in building density, green space ratio, traffic accessibility and other aspects of different schemes through split-screen display, color marking and other methods; (3) Future scenario projection: Integrating the ARIMA time series forecasting algorithm, based on historical data such as population, GDP, and industrial distribution of the target area from 2020 to 2024, the simulation projected urban and rural development changes in 2030 and 2035, with projection error ≤5%; (4) Immersive interaction: Connect to MetaQuest3VR device to support users to immerse themselves in virtual scenes (roaming speed adjustable from 0.5 to 10 m / s), and also support measurement (measurement accuracy 0.01 m), annotation (text and graphic annotations can be added) and collision detection (collision detection response time ≤ 0.05 seconds), to meet the needs of planning demonstration and public participation.

[0038] 2.6 Scene Validation and Output The virtual scene is verified according to the multi-dimensional verification index system of claim 6: (1) Geographic accuracy verification: Select 20 feature points (including bridges, landmark buildings, and road intersections) and compare the coordinates of the virtual scene with those of the real area. The average deviation is 0.32m, which is less than the requirement of 0.5m. (2) Planning rationality verification: The combination of automated and manual verification was used to verify the conformity of parameters such as building density, plot ratio, and road red line with local urban and rural planning standards. The verification pass rate reached 98.5%. (3) Interaction smoothness verification: On a PC with a normal configuration (CPU: Intel i7-13700K, GPU: NVIDIA RTX4070, memory: 32GB), the average frame rate of scene roaming was 72fps and the average model loading delay was 0.45 seconds, both of which met the requirements.

[0039] After successful verification, standardized virtual scene files (formats: .uasset, .umap) are output. Simultaneously, through a multi-terminal adaptation module, these files are converted into files recognizable by PCs (EXE format), mobile devices (APK format), VR devices (.apk format), and large-screen terminals (MP4 format), supporting multi-terminal display and data sharing.

[0040] 2.7 System Deployment The specific configuration of each module in the system described in claim 7 is as follows: (1) Data acquisition and preprocessing module: Deploys drone oblique photography unit (DJI Phantom 4RTK), laser scanning unit (RIEGLVZ-400), IoT sensing unit (30 sensing nodes), and data processing unit adopts Intel Xeon E5-2690 processor, integrating AI-CG dual-modal data fusion recognition module and seven-parameter coordinate transformation module; (2) Game engine scene initialization module: Deployed on GPU server (NVIDIA RTX A6000, 48GB video memory), integrating rendering parameter configuration unit (configuring Nanite, Lumen, virtual texture parameters) and mesh partition management unit (using dynamic mesh loading algorithm); (3) Terrain and ecological scene construction module and planning element integration module: deployed on the same GPU server as the game engine scene initialization module, sharing hardware resources; (4) Dynamic interaction and deduction module: integrates data visualization unit, supports the overlay display of charts and heat maps, and the interactive operation interface supports mouse, keyboard, MetaQuest3VR device and touch device access; (5) Scene verification and output module: deployed on a CPU server (Intel Xeon E5-2690, 64GB memory) to achieve collaborative work between automated verification and manual verification; (6) Data storage module: It adopts a distributed storage architecture (3 storage servers, each with a storage capacity of 10TB), supports efficient storage and parallel access of massive data, and the data backup frequency is once a day; (7) Multi-terminal adaptation module: Deployed on the application server, it supports converting virtual scene files into formats that can be recognized by different terminals, and adapts to PC, mobile, VR devices and large screen terminals.

[0041] The above embodiments are merely preferred embodiments of the present invention and should not be construed as limiting the scope of protection of the present invention. Any non-substantial changes and substitutions made by those skilled in the art based on the present invention shall fall within the scope of protection claimed by the present invention.

Claims

1. A method for constructing a virtual scene for urban and rural planning based on a game engine, characterized in that, Includes the following steps: S1. Multi-source data acquisition and preprocessing: Collect geospatial data, planning and design data, real-scene perception data, and human and social data of the target urban and rural areas. Standardize the format, remove noise, and calibrate the coordinates of various types of data to obtain a standardized fusion dataset. Among them, the geospatial data includes DEM digital elevation model, oblique photogrammetry model, and orthophoto map, which are acquired by UAV oblique photogrammetry technology combined with laser scanning technology. The real-scene perception data includes real-time traffic flow, environmental monitoring data, and population flow heat map data, which are collected in real time through Internet of Things sensing devices. S2. Game Engine Scene Initialization: Based on the game engine, a basic framework for a virtual scene is built, and scene rendering parameters are configured, including lighting parameters, texture parameters, and level of detail parameters. Using the game engine's world partitioning system, the target urban and rural areas are automatically divided into several grid partitions according to geographical coordinates, a unified coordinate system is established, and the automatic loading and unloading of scene grids is realized, solving the problem of lag when loading ultra-large scenes. S3. Terrain and Ecological Scene Construction: The preprocessed DEM digital elevation model is imported into the game engine, and terrain texture mapping is completed by combining orthophoto maps to generate a 3D terrain model that accurately matches the real geographic coordinates; based on procedural content generation (PCG) technology, the surface material is batch filled into the terrain model according to the land cover classification results, and ecological elements such as vegetation and water bodies are automatically placed to generate a virtual ecological environment scene; among them, the vegetation placement is combined with the vegetation type data of the target area to achieve accurate distribution and growth status simulation of different vegetation; S4. Precise Integration of Planning Elements: The pre-processed planning and design data is converted into a model format that the game engine can recognize. Lightweight processing and posture correction are performed on the planning element models such as buildings, roads, and pipelines to avoid scene lag caused by model redundancy. The processed planning element models are imported into the ecological environment virtual scene according to the real geographic coordinates to achieve seamless integration of planning elements with terrain and ecological scene, and generate the initial urban and rural planning virtual scene. S5. Dynamic Interaction and Deduction Optimization: Based on the real-time rendering capabilities of the game engine, it integrates a dynamic lighting system and a physical simulation system to realize the dynamic simulation of changes in light and shadow, weather, and vegetation growth in virtual scenes; it builds a planning and deduction module to support real-time adjustment of planning schemes, comparison of multiple schemes, and deduction of future scenes, while providing immersive interactive operations to support users to roam, measure, annotate, and perform collision detection in virtual scenes; S6. Scene Verification and Output: Establish a multi-dimensional verification index system to verify the geographic accuracy, planning rationality, and interactive smoothness of the virtual scene. If the verification fails, return to the corresponding step for adjustment. After the verification passes, output a standardized urban and rural planning virtual scene file, which supports multi-terminal display and data sharing.

2. The method according to claim 1, characterized in that, The data preprocessing described in step S1 specifically includes: using AI-CG dual-modal data fusion and recognition technology to intelligently decipher planning and design drawings, accurately extracting internal building structure data such as walls, beams, columns, and pipelines; denoising and simplifying the oblique photography model, retaining key details while reducing model complexity; converting all data to the WGS84 coordinate system through coordinate transformation algorithms to ensure data spatial consistency; and supplementing missing human and social data using interpolation algorithms to improve the completeness of the dataset.

3. The method according to claim 1, characterized in that, The game engine used in step S2 is Unreal Engine 5 (UE5). The specific rendering parameter configuration includes: using UE5's Nanite micropolygon geometry technology to achieve infinite detail rendering of planned element models without manually setting Level of Detail (LOD), avoiding visual jumps when switching models; using the Lumen global illumination system to achieve real-time light bounce and dynamic lighting changes in the virtual scene, simulating the light and shadow effects under different time periods and weather conditions, and improving scene realism; configuring virtual texture technology to support efficient loading and streaming of massive high-resolution textures, achieving ultra-high-definition scene presentation with limited video memory.

4. The method according to claim 1, characterized in that, The specific applications of the programmatic content generation (PCG) technology described in step S3 include: setting parameters such as vegetation type, density, and growth cycle based on the ecological environment data of the target area, and automatically generating vegetation communities that conform to the real environment in batches through the PCG toolset; automatically assigning corresponding surface materials according to information such as terrain slope and altitude to achieve differentiated presentation of different terrains such as mountains, plains, and water bodies; and supporting users to customize PCG rules and flexibly adjust the presentation effect of ecological scenes.

5. The method according to claim 1, characterized in that, The planning simulation module in step S5 specifically includes: supporting parametric adjustments to planning schemes, allowing users to modify planning parameters such as building height, road width, and green space area in real time, with scene updates synchronized; providing a multi-scheme comparison function, enabling the simultaneous loading of multiple planning schemes and visually presenting the differences between different schemes; integrating future scene simulation algorithms, based on historical data and development trends, simulating urban and rural development changes in the target area over the next 5-20 years, providing data support for planning decisions; and supporting VR / AR device access for immersive scene roaming and planning operations.

6. The method according to claim 1, characterized in that, The multi-dimensional verification index system mentioned in step S6 includes: geographic accuracy index, requiring the coordinate deviation between the virtual scene and the real area to be no more than 0.5 meters; planning rationality index, including the degree of conformity between parameters such as building density, plot ratio, and road red line and planning specifications; interaction smoothness index, requiring the scene roaming frame rate to be no less than 60fps and the model loading delay to be no more than 1 second; the verification process adopts a combination of automated verification and manual verification to improve verification efficiency and accuracy.

7. A virtual scene construction system for urban and rural planning based on a game engine, characterized in that, include: The data acquisition and preprocessing module is used to collect geospatial data, planning and design data, real-scene perception data, and human and social data of the target urban and rural areas, and to perform format standardization, noise removal, coordinate calibration, and missing value supplementation on various types of data, outputting a standardized fusion dataset; the data acquisition and preprocessing module includes a UAV oblique photography unit, a laser scanning unit, an IoT sensing unit, and a data processing unit, the data processing unit integrating an AI-CG dual-modal data fusion recognition module and a coordinate transformation module; The game engine scene initialization module is used to build a basic framework for virtual scenes based on the game engine, configure scene rendering parameters, divide the target area into grids using the world partitioning system, establish a unified coordinate system, and realize the automatic loading and unloading of scene grids. The game engine scene initialization module integrates a rendering parameter configuration unit and a grid partition management unit; The terrain and ecological scene construction module is used to import the preprocessed DEM digital elevation model and orthophoto map into the game engine, complete the terrain texture mapping, and generate a 3D terrain model; based on procedural content generation (PCG) technology, it completes the surface material filling and ecological element placement to generate a virtual ecological environment scene. The planning element fusion module is used to convert planning and design data into a model format that the game engine can recognize, perform lightweight processing and posture correction on the planning element model, import the planning element model into the ecological environment virtual scene according to the real geographic coordinates, achieve seamless fusion, and generate the initial urban and rural planning virtual scene. The dynamic interaction and simulation module integrates a dynamic lighting system and a physical simulation system to simulate the dynamic changes of virtual scenes; it also builds a planning and simulation module to support real-time adjustment of planning schemes, comparison of multiple schemes, and simulation of future scenes; and provides an immersive interactive operation interface to support users in roaming, measuring, labeling, and collision detection. The scene verification and output module is used to establish a multi-dimensional verification index system to verify the geographic accuracy, planning rationality, and interaction smoothness of virtual scenes. Scenes that fail the verification are returned to the corresponding module for adjustment. After the verification is passed, standardized virtual scene files are output, which support multi-terminal display and data sharing. The data storage module is used to store the collected raw data, preprocessed standardized fusion datasets, virtual scene model files, planning scheme data, and verification result data, and supports fast data query, update, and backup.

8. The system according to claim 7, characterized in that, The game engine uses Unreal Engine 5 (UE5). The rendering parameter configuration unit in the game engine's scene initialization module is specifically used to configure Nanite micro-polygon geometry parameters, Lumen global illumination parameters, and virtual texture parameters to achieve ultra-high-definition, high-realism scene rendering. The mesh partition management unit uses a dynamic mesh loading algorithm to automatically load mesh partitions within the current field of view and unload mesh partitions outside the field of view based on the user's operation position, thereby reducing system resource consumption.

9. The system according to claim 7, characterized in that, The dynamic interaction and simulation module also includes a data visualization unit, which overlays and displays planning simulation data and real-world perception data in the form of charts, heat maps, etc., in the virtual scene to intuitively present the implementation effect of the planning scheme and the regional development status; the interactive operation interface supports mouse, keyboard, VR / AR devices and touch devices to meet the operation needs of different users.

10. The system according to claim 7, characterized in that, The system also includes a multi-terminal adaptation module, which converts the output virtual scene files into a format recognizable by different terminals, supporting display and interaction on PCs, mobile devices, VR devices, and large-screen terminals; the data storage module adopts a distributed storage architecture, supporting efficient storage and parallel access of massive amounts of data, ensuring data security and availability.

Images