Urban three-dimensional visualization interactive method and system based on virtual reality

Abstract

The present invention discloses a method and system for interactive three-dimensional visualization of a city based on virtual reality, and relates to the technical field of urban three-dimensional visualization data processing. The method comprises the following steps: dividing a three-dimensional city model into multiple sub-areas, pre-processing virtual reality three-dimensional hardware load-related data and virtual reality three-dimensional software load-related data, comprehensively evaluating the hardware load evaluation coefficient and the software load evaluation coefficient to obtain a rendering load evaluation coefficient, and executing a balanced rendering quality optimization scheme for the real-time rendering of the sub-areas based on the threshold comparison results of the rendering load evaluation coefficient. By executing a balanced rendering quality optimization scheme for the real-time rendering of the sub-areas based on the threshold comparison results of the rendering load evaluation coefficient, the present invention achieves the effect of accelerating real-time rendering speed and improving frame rate, and solves the problem of insufficient accuracy of virtual reality three-dimensional real-time rendering when complex virtual reality three-dimensional visualization interactions exist in the prior art.

Description

Technical Field

[0001] The present invention relates to the technical field of urban three-dimensional visualization data processing, and in particular to a method and system for urban three-dimensional visualization interaction based on virtual reality. Background Art

[0002] The complexity of modern urban planning and management is increasing. Traditional two-dimensional floor plans and static models can no longer meet people's needs for intuitive, dynamic, and multi-dimensional information display. With the continuous maturity of virtual reality technology, by building highly realistic three-dimensional virtual city environments, users can roam and explore freely in them and perform various interactive operations, such as modifying building designs and simulating traffic flows, thus providing a new and efficient tool and method for urban planning, design, management and other fields.

[0003] The existing visual interaction system is realized through three-dimensional modeling technology, real-time visualization technology and multi-source data acquisition and integration technology.

[0004] 3D modeling technology: Use laser scanning, photogrammetry and other technologies to obtain the 3D coordinates and texture information of objects in the real world and build realistic 3D models.

[0005] Real-time visualization technology: Integrate the established 3D model with traffic data, etc., and use graphics rendering engine and virtual reality technology to achieve realistic visualization effects.

[0006] Multi-source data collection and integration: Collect data from multiple data sources such as traffic sensors, cameras, GPS devices, mobile terminals, etc., and integrate and clean them to form a unified data format and standard.

[0007] For example, the invention patent with publication number CN113963113A discloses a three-dimensional visualization method for urban buildings, which includes: constructing a three-dimensional model of urban buildings; dividing the three-dimensional model into blocks, dividing the three-dimensional model of urban buildings into spatial models of different areas according to categories; performing overall visualization and regional visualization of the three-dimensional visualization of urban buildings, realizing data reading, and performing color rendering or texture mapping; integrating the spatial models after color rendering or texture mapping to form a complete three-dimensional visualization model of urban buildings.

[0008] For example, the invention patent with publication number CN118485784A discloses an efficient modeling and visualization method for urban-level scenes based on NeRF, including: training data partitioning, and also including the following steps: parallel training of sub-regions; generating a sparse octree and storing the weight data of each node; and performing interactive rendering.

[0009] However, in the process of implementing the technical solutions of the invention in the embodiments of the present application, the present application found that the above technology has at least the following technical problems:

[0010] In the existing technology, in the three-dimensional visualization interaction of urban scenes, when users interact with the city 3D model, for example, when switching between day and night, zooming in on details and weather changes, the buildings and sites in the city 3D model need to be presented in real time with finer details. This leads to the problem of insufficient accuracy of virtual reality 3D real-time rendering during complex interactions in virtual reality 3D visualization. Summary of the Invention

[0011] The embodiments of the present application provide a method and system for interactive three-dimensional city visualization based on virtual reality, thereby solving the problem of insufficient accuracy of three-dimensional real-time rendering in virtual reality during complex interactions in the prior art, and achieving the effect of accelerating real-time rendering speed and improving frame rate.

[0012] The embodiment of the present application provides a method for city three-dimensional visualization interaction based on virtual reality, comprising the following steps: dividing a city three-dimensional model into multiple sub-areas, a user selecting a sub-area, and performing visualization interaction on the model in the sub-area; during the user's visualization interaction, a virtual reality computer renders the sub-area in real time and collects virtual reality three-dimensional hardware load-related data and virtual reality three-dimensional software load-related data; preprocessing the virtual reality three-dimensional hardware load-related data and virtual reality three-dimensional software load-related data to obtain virtual reality three-dimensional hardware load preprocessing data and virtual reality three-dimensional software load preprocessing data; analyzing the virtual reality three-dimensional hardware load preprocessing data to obtain a hardware load evaluation coefficient, and analyzing the virtual reality three-dimensional software load preprocessing data to obtain a software load evaluation coefficient. An estimation coefficient is obtained by comprehensively evaluating the hardware load assessment coefficient and the software load assessment coefficient to obtain the rendering load assessment coefficient; the hardware load assessment coefficient, the software load assessment coefficient and the rendering load assessment coefficient are respectively compared with the hardware load assessment threshold, the software load assessment threshold and the rendering load assessment threshold; according to the threshold comparison result of the hardware load assessment coefficient, a hardware load optimization scheme is executed for the real-time rendering of the sub-area; if the hardware load assessment coefficient is still greater than or equal to the hardware load assessment threshold, a software load optimization scheme is executed for the real-time rendering of the sub-area; according to the threshold comparison result of the software load assessment coefficient, a rendering load optimization scheme is executed for the real-time rendering of the sub-area; according to the threshold comparison result of the rendering load assessment coefficient, a balanced rendering quality optimization scheme is executed for the real-time rendering of the sub-area.

[0013] Furthermore, the specific interactive process of visual interaction with the model in the sub-area is as follows: obtain the model in the sub-area through laser scanning, import the model in the sub-area into the VR development engine, add visual cues in the VR environment, and visually interact with the model in the sub-area through handle buttons, gestures and voice input. The visual interaction includes dragging and zooming, rotating to observe the internal structure, triggering animation demonstration, day and night switching and weather adjustment.

[0014] Furthermore, the specific collection process of virtual reality 3D hardware load related data and virtual reality 3D software load related data is: collecting virtual reality 3D hardware load related data and virtual reality 3D software load related data through the built-in tools of the operating system and using Unity's Profiler; virtual reality 3D hardware load related data includes the CPU core utilization of the VR computer, the GPU utilization of the VR computer, the percentage of video memory occupancy of the VR computer and the multi-threaded scheduling efficiency of the VR computer; virtual reality 3D software load related data includes the number of polygons in the city 3D during each frame rendering, the number of dynamic objects in the city 3D during each frame rendering, the interactive delay from input to screen update in each frame rendering and the data stream bandwidth in each frame rendering.

[0015] Furthermore, the specific process of preprocessing the virtual reality three-dimensional hardware load related data and the virtual reality three-dimensional software load related data is as follows: associating the virtual reality three-dimensional hardware load related data and the virtual reality three-dimensional software load related data through timestamps, using linear interpolation method to fill in short-term missing data in the virtual reality three-dimensional hardware load related data and the virtual reality three-dimensional software load related data, performing time interval detection on the virtual reality three-dimensional hardware load related data and the virtual reality three-dimensional software load related data, marking and deleting long-term missing data, using a low-pass filter to smooth the virtual reality three-dimensional hardware load related data and the virtual reality three-dimensional software load related data, and using Z-score normalization method to normalize the virtual reality three-dimensional hardware load related data and the virtual reality three-dimensional software load related data.

[0016] Furthermore, the specific analysis process of analyzing the virtual reality three-dimensional hardware load preprocessing data is as follows: obtaining the weight factor of the VR computer's CPU core utilization, the weight factor of the VR computer's GPU utilization, the weight factor of the VR computer's video memory occupancy percentage, and the weight factor of the VR computer's multi-threaded scheduling efficiency from the database, arranging the VR computer's CPU core utilization, the VR computer's GPU utilization, the VR computer's video memory occupancy percentage, and the VR computer's multi-threaded scheduling efficiency in a time series, assigning weights to the VR computer's CPU core utilization, the VR computer's GPU utilization, the VR computer's video memory occupancy percentage, and the VR computer's multi-threaded scheduling efficiency, comparing the standard values of the VR computer's CPU core utilization, the VR computer's GPU utilization, the VR computer's video memory occupancy percentage, and the VR computer's multi-threaded scheduling efficiency with the average values of the VR computer's CPU core utilization, the VR computer's GPU utilization, the VR computer's video memory occupancy percentage, and the VR computer's multi-threaded scheduling efficiency, and processing to obtain a hardware load evaluation coefficient.

[0017] Furthermore, the specific analysis process of analyzing the preprocessing data of the virtual reality three-dimensional software load is as follows: obtaining from the database the weight factor of the number of polygons in the three-dimensional city when rendering each frame, the weight factor of the number of dynamic objects in the three-dimensional city when rendering each frame, the weight factor of the interactive delay from input to screen update in each frame rendering, and the weight factor of the data flow bandwidth in each frame rendering, arranging the number of polygons in the three-dimensional city when rendering each frame, the number of dynamic objects in the three-dimensional city when rendering each frame, the interactive delay from input to screen update in each frame rendering, and the data flow bandwidth in each frame rendering in frame sequence, and performing the weight factor of the number of polygons in the three-dimensional city when rendering each frame. The weights of the number of polygons in the city 3D when rendering each frame, the number of dynamic three-dimensional objects in the city when rendering each frame, the interactive delay from input to screen update in the rendering each frame, and the data stream bandwidth in the rendering each frame are allocated. The standard values of the number of polygons in the city 3D when rendering each frame, the number of dynamic three-dimensional objects in the city when rendering each frame, the interactive delay from input to screen update in the rendering each frame, and the data stream bandwidth in the rendering each frame are compared with the average values of the number of polygons in the city 3D when rendering each frame, the number of dynamic three-dimensional objects in the city when rendering each frame, the interactive delay from input to screen update in the rendering each frame, and the data stream bandwidth in the rendering each frame, and the software load assessment coefficient is obtained through processing.

[0018] Furthermore, the specific analysis process for comprehensive evaluation of the hardware load assessment coefficient and the software load assessment coefficient is as follows: obtaining the weight factor of the hardware load assessment coefficient and the weight factor of the software load assessment coefficient from the database, assigning weights to the hardware load assessment coefficient and the software load assessment coefficient, combining the hardware load assessment coefficient and the software load assessment coefficient, and processing them to obtain the rendering load assessment coefficient.

[0019] Furthermore, the specific threshold comparison method for comparing the hardware load evaluation coefficient, the software load evaluation coefficient and the rendering load evaluation coefficient with the hardware load evaluation threshold, the software load evaluation threshold and the rendering load evaluation threshold is as follows: comparing the hardware load evaluation coefficient with the hardware load evaluation threshold; if the hardware load evaluation coefficient is less than the hardware load evaluation threshold, directly performing real-time rendering on the sub-area; if the hardware load evaluation coefficient is greater than or equal to the hardware load evaluation threshold, executing a hardware load optimization scheme for the real-time rendering of the sub-area; after executing the hardware load optimization scheme, continuing to compare the hardware load evaluation coefficient with the hardware load evaluation threshold; if the hardware load evaluation coefficient is still greater than or equal to the hardware load evaluation threshold, executing a software load optimization scheme for the real-time rendering of the sub-area; performing software load optimization scheme on the real-time rendering of the sub-area. Comparison of the load assessment coefficient and the software load assessment threshold. If the software load assessment coefficient is less than the software load assessment threshold, the sub-area is rendered in real time. If the software load assessment coefficient is greater than or equal to the software load assessment threshold, a rendering load optimization solution is implemented for the real-time rendering of the sub-area. Comparison of the rendering load assessment coefficient and the rendering load assessment threshold. If the rendering load assessment coefficient is less than the rendering load assessment threshold, the sub-area is rendered in real time. If the rendering load assessment coefficient is greater than or equal to the rendering load assessment threshold, a balanced rendering quality optimization solution is implemented for the real-time rendering of the sub-area. The hardware load optimization solution includes dynamic resolution adjustment. The software load optimization solution includes interactive event priority scheduling and multi-threaded parallel processing. The rendering load optimization solution includes cone-based GPU culling technology.

[0020] Furthermore, the specific process of the balanced rendering quality optimization plan is: if after executing the hardware load optimization plan, the software load optimization plan and the rendering load optimization plan, if the rendering load evaluation coefficient is still greater than or equal to the rendering load evaluation threshold, the balanced rendering quality optimization plan is executed for the real-time rendering of the sub-area, including performing polygon merging on the real-time rendering of the sub-area, reducing the number of polygons and not rendering the polygons facing away from the camera. If the rendering load evaluation coefficient is less than the rendering load evaluation threshold, the balanced rendering quality optimization plan will no longer be executed.

[0021] The embodiment of the present application provides a virtual reality-based three-dimensional visualization interactive system for a city, which includes an interaction and data collection module, a data preprocessing module, a data analysis module, and an optimization module: the interaction and data collection module is used to divide a three-dimensional city model into multiple sub-areas, and the user selects a sub-area and performs visual interaction on the model in the sub-area. During the user's visual interaction, the virtual reality computer renders the sub-area in real time and collects virtual reality three-dimensional hardware load-related data and virtual reality three-dimensional software load-related data; the data preprocessing module is used to preprocess the virtual reality three-dimensional hardware load-related data and virtual reality three-dimensional software load-related data to obtain virtual reality three-dimensional hardware load preprocessing data and virtual reality three-dimensional software load preprocessing data; the data analysis module is used to analyze the virtual reality three-dimensional hardware load preprocessing data to obtain a hardware load evaluation coefficient. Analyze the virtual reality three-dimensional software load preprocessing data to obtain the software load evaluation coefficient, conduct a comprehensive evaluation of the hardware load evaluation coefficient and the software load evaluation coefficient to obtain the rendering load evaluation coefficient; optimization module: used to perform threshold comparison between the hardware load evaluation coefficient, the software load evaluation coefficient and the rendering load evaluation coefficient and the hardware load evaluation threshold, the software load evaluation threshold and the rendering load evaluation threshold respectively, and according to the threshold comparison result of the hardware load evaluation coefficient, execute the hardware load optimization plan for the real-time rendering of the sub-area; if the hardware load evaluation coefficient is still greater than or equal to the hardware load evaluation threshold, execute the software load optimization plan for the real-time rendering of the sub-area; according to the threshold comparison result of the software load evaluation coefficient, execute the rendering load optimization plan for the real-time rendering of the sub-area; according to the threshold comparison result of the rendering load evaluation coefficient, execute the balanced rendering quality optimization plan for the real-time rendering of the sub-area.

[0022] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages:

[0023] 1. By comparing the results of the threshold of the rendering load evaluation coefficient, a balanced rendering quality optimization scheme is implemented for the real-time rendering of sub-areas, thereby achieving the effect of accelerating real-time rendering speed and improving frame rate. This effectively solves the problem of insufficient accuracy of VR 3D real-time rendering during complex interactions in VR 3D visualization in existing technologies.

[0024] 2. By dividing the urban 3D model into multiple sub-areas, users can select a sub-area and perform visual interaction on the model in the sub-area, thereby improving the efficiency of visual interaction and effectively solving the problem of insufficient visual interaction efficiency in existing technologies.

[0025] 3. By preprocessing virtual reality 3D hardware load related data and virtual reality 3D software load related data, virtual reality 3D hardware load preprocessing data and virtual reality 3D software load preprocessing data are obtained, thereby achieving the effect of improving data quality and effectively solving the problem of insufficient data quality in the existing technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 A flowchart of a method for interactive three-dimensional visualization of a city based on virtual reality provided in an embodiment of the present application;

[0027] Figure 2 A schematic diagram of the hardware load evaluation coefficient for the virtual reality-based three-dimensional urban visualization interactive method provided in an embodiment of the present application;

[0028] Figure 3 This is a structural diagram of the virtual reality-based urban three-dimensional visualization interactive system provided in an embodiment of the present application. DETAILED DESCRIPTION

[0029] The embodiments of the present application provide a virtual reality-based three-dimensional urban visualization interaction method and system, which solves the problem of insufficient accuracy of virtual reality three-dimensional real-time rendering during complex virtual reality three-dimensional visualization interactions in the prior art. By comparing the threshold results of the rendering load evaluation coefficient, a balanced rendering quality optimization scheme is implemented for the real-time rendering of sub-areas, thereby achieving the effect of accelerating real-time rendering speed and improving frame rate.

[0030] The technical solution in the embodiments of the present application is to solve the above-mentioned problem of insufficient accuracy of virtual reality 3D real-time rendering during complex interactions of virtual reality 3D visualization. The overall idea is as follows:

[0031] By dividing the urban 3D model into multiple sub-areas, collecting virtual reality 3D hardware load-related data and virtual reality 3D software load-related data during the visualization interaction process, analyzing the virtual reality 3D hardware load preprocessing data to obtain the hardware load assessment coefficient, analyzing the virtual reality 3D software load preprocessing data to obtain the software load assessment coefficient, and comprehensively evaluating the hardware load assessment coefficient and the software load assessment coefficient to obtain the rendering load assessment coefficient, the effect of accelerating real-time rendering speed and improving frame rate is achieved.

[0032] In order to better understand the above technical solution, the above technical solution will be described in detail below with reference to the accompanying drawings and specific implementation methods.

[0033] like Figure 13D city visualization interaction method based on virtual reality provided by an embodiment of the present application, the method is applied to a city 3D visualization interaction system based on virtual reality, the method comprising the following steps: dividing a city 3D model into a plurality of sub-areas, a user selecting a sub-area, performing visualization interaction on the model in the sub-area, during which the user performs visualization interaction, a virtual reality computer performs real-time rendering on the sub-area and collects virtual reality 3D hardware load related data and virtual reality 3D software load related data; pre-processing the virtual reality 3D hardware load related data and virtual reality 3D software load related data to obtain virtual reality 3D hardware load pre-processed data and virtual reality 3D software load pre-processed data; analyzing the virtual reality 3D hardware load pre-processed data to obtain a hardware load evaluation coefficient, and analyzing the virtual reality 3D hardware load pre-processed data. The software load preprocessing data is measured to obtain the software load assessment coefficient, and the hardware load assessment coefficient and the software load assessment coefficient are comprehensively evaluated to obtain the rendering load assessment coefficient; the hardware load assessment coefficient, the software load assessment coefficient and the rendering load assessment coefficient are threshold-compared with the hardware load assessment threshold, the software load assessment threshold and the rendering load assessment threshold respectively; according to the threshold comparison result of the hardware load assessment coefficient, the hardware load optimization scheme is executed for the real-time rendering of the sub-area; if the hardware load assessment coefficient is still greater than or equal to the hardware load assessment threshold, the software load optimization scheme is executed for the real-time rendering of the sub-area; according to the threshold comparison result of the software load assessment coefficient, the rendering load optimization scheme is executed for the real-time rendering of the sub-area; according to the threshold comparison result of the rendering load assessment coefficient, the balanced rendering quality optimization scheme is executed for the real-time rendering of the sub-area.

[0034] Furthermore, the specific interactive process of visual interaction with the model in the sub-area is as follows: obtain the model in the sub-area through laser scanning, import the model in the sub-area into the VR development engine, add visual cues in the VR environment, and visually interact with the model in the sub-area through handle buttons, gestures and voice input. The visual interaction includes dragging and zooming, rotating to observe the internal structure, triggering animation demonstration, day and night switching and weather adjustment.

[0035] In this embodiment, laser scanning technology is used to scan an existing urban park to obtain precise three-dimensional data of a specific building in the park, such as a pavilion. The scanned three-dimensional model of the building is imported into a VR development engine, such as Unity or Unreal Engine. In the VR environment, visual cues are added to the building, such as highlighting areas to be renovated or marking different functional areas with different colors. Users can drag and scale the building model using buttons on the VR controller to observe the model from different angles and distances. Users can use gesture control, such as making a fist and opening the palm, to rotate the building model and observe its internal structure. When the user wants to view the effect of the renovation, they can trigger animations using specific controller buttons, such as simulating the opening of the pavilion top or the dynamic effect of a sculpture. Users can use voice input commands, such as "switch to night", to observe the visual effects of the building under different lighting conditions. Similarly, users can use voice commands to adjust the weather in the VR environment, such as simulating rain or sunny days, to evaluate the impact of weather on the park user experience.

[0036] Furthermore, the specific collection process of virtual reality 3D hardware load related data and virtual reality 3D software load related data is: collecting virtual reality 3D hardware load related data and virtual reality 3D software load related data through the built-in tools of the operating system and using Unity's Profiler; virtual reality 3D hardware load related data includes the CPU core utilization of the VR computer, the GPU utilization of the VR computer, the percentage of video memory occupancy of the VR computer and the multi-threaded scheduling efficiency of the VR computer; virtual reality 3D software load related data includes the number of polygons in the city 3D during each frame rendering, the number of dynamic objects in the city 3D during each frame rendering, the interactive delay from input to screen update in each frame rendering and the data stream bandwidth in each frame rendering.

[0037] In this embodiment, the VR computer's CPU core utilization, GPU utilization, video memory usage percentage, multi-thread scheduling efficiency, and the number of polygons in the three-dimensional city during each frame rendering, the number of dynamic objects, the interactive delay from input to screen update, and the data stream bandwidth can be obtained through real-time monitoring and performance analysis tools. These tools include the operating system's built-in task manager and performance monitor, as well as the performance analyzer provided by professional VR development engines (such as Unity or Unreal Engine).

[0038] A VR computer's CPU core utilization refers to the degree to which each core of the computer's central processing unit (CPU) is used while a VR application is running, typically expressed as a percentage. This can be obtained through the operating system's built-in task manager (such as Windows' Task Manager or macOS's Activity Monitor) or third-party performance monitoring tools (such as CPU-Z and HWMonitor).

[0039] A VR computer's GPU utilization refers to how busy the graphics processing unit (GPU) is when performing graphics rendering and other computing tasks, expressed as a percentage. You can monitor GPU utilization using tools provided by your GPU manufacturer, such as NVIDIA's GPU Control Panel and Performance Monitor in GeForce Experience, or AMD's Performance Monitor in Radeon Settings.

[0040] The percentage of video memory usage on a VR computer refers to the extent to which the GPU's video memory is being used by currently running applications, also expressed as a percentage. You can check video memory usage using tools provided by the GPU manufacturer or third-party monitoring software.

[0041] The multi-threaded scheduling efficiency of a VR computer refers to the operating system's ability to effectively allocate and schedule multiple threads to run on multiple CPU cores, and usually requires evaluation using professional performance analysis tools (such as Intel VTune Amp Level 1 and AMD uProf).

[0042] The number of polygons in the three-dimensional city during each frame rendering refers to the total number of polygons (usually triangles) contained in the three-dimensional city model during each frame rendering process, which can be obtained through the statistical function provided by the three-dimensional engine or rendering API (such as OpenGL, DirectX).

[0043] The number of urban 3D dynamic objects in each frame rendering refers to the number of dynamically changing objects in the urban 3D scene in each frame, such as moving vehicles, pedestrians or other animation elements. Dynamic objects can be counted in the rendering loop by programming.

[0044] The interaction delay from input to screen update in each frame rendering refers to the time delay from user input (such as mouse clicks, keyboard keys or VR controller actions) to the corresponding screen update being displayed on the screen. The time difference between input and rendering updates can be measured by high-performance timing tools (such as QueryPerformanceCounter in Windows).

[0045] The data flow bandwidth in each frame rendering refers to the data transferred from the main memory to the GPU memory during each frame rendering process. The GPU performance analysis tool can be used to monitor the bandwidth of data transmission.

[0046] Furthermore, the specific process of preprocessing the virtual reality three-dimensional hardware load related data and the virtual reality three-dimensional software load related data is as follows: associating the virtual reality three-dimensional hardware load related data and the virtual reality three-dimensional software load related data through timestamps, using linear interpolation method to fill in short-term missing data in the virtual reality three-dimensional hardware load related data and the virtual reality three-dimensional software load related data, performing time interval detection on the virtual reality three-dimensional hardware load related data and the virtual reality three-dimensional software load related data, marking and deleting long-term missing data, using a low-pass filter to smooth the virtual reality three-dimensional hardware load related data and the virtual reality three-dimensional software load related data, and using Z-score normalization method to normalize the virtual reality three-dimensional hardware load related data and the virtual reality three-dimensional software load related data.

[0047] In this embodiment, for example, at a specific time point t1, the CPU core utilization is 60%, and the number of polygons in the 3D city rendered per frame is 1000. If the CPU core utilization data at time point t2 is missing, linear interpolation is used to estimate the CPU core utilization at time point t2 based on the CPU core utilization data at time points t1 and t3 (assuming they are 60% and 70%, respectively). If the GPU utilization data from time points t4 to t6 is completely missing, this missing period is determined to be a long-term absence. This period is marked and excluded from the analysis to avoid interference with the overall performance evaluation. To reduce noise in the data, a low-pass filter is used to smooth the CPU and GPU utilization data to obtain a smoother data curve for easier analysis. For example, if the average video memory utilization is 50% and the standard deviation is 10%, then the data point with a video memory utilization of 60% at time point t7 will be converted to a Z-score of 1, indicating that it is 1 standard deviation above the average.

[0048] Furthermore, the specific analysis process of analyzing the virtual reality three-dimensional hardware load preprocessing data is as follows: obtaining the weight factor of the VR computer's CPU core utilization, the weight factor of the VR computer's GPU utilization, the weight factor of the VR computer's video memory occupancy percentage, and the weight factor of the VR computer's multi-threaded scheduling efficiency from the database, arranging the VR computer's CPU core utilization, the VR computer's GPU utilization, the VR computer's video memory occupancy percentage, and the VR computer's multi-threaded scheduling efficiency in a time series, assigning weights to the VR computer's CPU core utilization, the VR computer's GPU utilization, the VR computer's video memory occupancy percentage, and the VR computer's multi-threaded scheduling efficiency, comparing the standard values of the VR computer's CPU core utilization, the VR computer's GPU utilization, the VR computer's video memory occupancy percentage, and the VR computer's multi-threaded scheduling efficiency with the average values of the VR computer's CPU core utilization, the VR computer's GPU utilization, the VR computer's video memory occupancy percentage, and the VR computer's multi-threaded scheduling efficiency, and processing to obtain a hardware load evaluation coefficient.

[0049] In this embodiment, the specific method for obtaining the hardware load evaluation coefficient is:

[0050]

[0051] Where HLA represents the hardware load assessment coefficient, which is used to evaluate the hardware load during rendering. Several time monitoring points are set, S0 = 1, 2, 3, ..., a, a represents the total number of time monitoring points, represents the CPU core utilization of the VR computer at the S0th time monitoring point, CPU_CU0 represents the CPU core utilization of the standard VR computer, and α represents the weight factor of the CPU core utilization of the VR computer. represents the GPU utilization of the VR computer at the S0th time monitoring point, β represents the weight factor of the GPU utilization of the VR computer, GPU_U0 represents the GPU utilization of the standard VR computer, It represents the percentage of video memory usage of the VR computer at the S0th time monitoring point, γ represents the weight factor of the percentage of video memory usage of the VR computer, GPU_MUP0 represents the percentage of video memory usage of the standard VR computer, represents the multi-thread scheduling efficiency of the VR computer at the S0th time monitoring point, δ represents the weight factor of the multi-thread scheduling efficiency of the VR computer, and MSE0 represents the multi-thread scheduling efficiency of the standard VR computer.

[0052] If the CPU core utilization of a VR computer is too high, the GPU of the VR computer will have to wait for data, resulting in a decrease in utilization. Efficient scheduling can evenly distribute tasks to the CPU cores (for example, separating UI threads, physics threads, and rendering instruction submission threads), avoiding single-core overload, improving CPU utilization, and reducing GPU waiting time. A high percentage of video memory usage (for example, a complex city model exceeds the video memory capacity) will trigger memory-to-video memory data swapping, significantly increasing latency and causing the GPU to have a decrease in utilization due to waiting for data.

[0053] When the system is running, it retrieves a mapping table of weight factors from the database and quickly extracts corresponding weight factors based on the current VR computer's CPU core utilization, GPU utilization, video memory usage percentage, and multi-threaded scheduling efficiency. These factors include the weight factor for the VR computer's CPU core utilization, the weight factor for the VR computer's GPU utilization, the weight factor for the VR computer's video memory usage percentage, and the weight factor for the VR computer's multi-threaded scheduling efficiency. This mapping table defines a clear set of association rules that converts the specific values of the VR computer's CPU core utilization, GPU utilization, video memory usage percentage, and multi-threaded scheduling efficiency into their corresponding weight factors. This mechanism effectively enables dynamic acquisition of weight factors, whether achieving a one-to-one exact match or a many-to-one relationship where multiple parameters are aggregated into a single weight.

[0054] In a specific embodiment, examples of hardware load evaluation coefficient data are shown in the following table.

[0055] Table 1 Hardware load evaluation coefficient data example table

[0056]

[0057] When the weight factors of the VR computer's CPU core utilization, the VR computer's GPU utilization, the VR computer's video memory occupancy percentage, and the VR computer's multi-threaded scheduling efficiency are 0.2, 0.3, 0.25, and 0.25, respectively, the standard values of the VR computer's CPU core utilization, the VR computer's GPU utilization, the VR computer's video memory occupancy percentage, and the VR computer's multi-threaded scheduling efficiency are 55%, 50%, 65%, and 80%, respectively. Figure 2 From the data in Table 1, it can be seen that when the CPU core utilization, GPU utilization, and video memory occupancy percentage of the VR computer remain unchanged, the greater the multi-thread scheduling efficiency of the VR computer, the smaller the hardware load evaluation coefficient.

[0058] Furthermore, the specific analysis process of analyzing the preprocessing data of the virtual reality three-dimensional software load is as follows: obtaining from the database the weight factor of the number of polygons in the three-dimensional city when rendering each frame, the weight factor of the number of dynamic objects in the three-dimensional city when rendering each frame, the weight factor of the interactive delay from input to screen update in each frame rendering, and the weight factor of the data flow bandwidth in each frame rendering, arranging the number of polygons in the three-dimensional city when rendering each frame, the number of dynamic objects in the three-dimensional city when rendering each frame, the interactive delay from input to screen update in each frame rendering, and the data flow bandwidth in each frame rendering in frame sequence, and performing the weight factor of the number of polygons in the three-dimensional city when rendering each frame. The weights of the number of polygons in the city 3D when rendering each frame, the number of dynamic three-dimensional objects in the city when rendering each frame, the interactive delay from input to screen update in the rendering each frame, and the data stream bandwidth in the rendering each frame are allocated. The standard values of the number of polygons in the city 3D when rendering each frame, the number of dynamic three-dimensional objects in the city when rendering each frame, the interactive delay from input to screen update in the rendering each frame, and the data stream bandwidth in the rendering each frame are compared with the average values of the number of polygons in the city 3D when rendering each frame, the number of dynamic three-dimensional objects in the city when rendering each frame, the interactive delay from input to screen update in the rendering each frame, and the data stream bandwidth in the rendering each frame, and the software load assessment coefficient is obtained through processing.

[0059] In this embodiment, the specific method for obtaining the software load evaluation coefficient is:

[0060]

[0061] Where SLA represents the software load assessment coefficient, which is used to evaluate the load of the software during rendering. Several frame monitoring points are set, R0 = 1, 2, 3, ..., b, b represents the total number of frame monitoring points, It represents the number of polygons in the three-dimensional city when rendering each frame under the R0 frame monitoring point, 3D_NOP0 represents the number of polygons in the three-dimensional city when rendering each frame, and θ represents the weight factor of the number of polygons in the three-dimensional city when rendering each frame. Indicates the number of urban 3D dynamic objects per frame rendering under the R0th frame monitoring point, μ indicates the weight factor of the number of urban 3D dynamic objects per frame rendering, 3D_NOD0 indicates the number of urban 3D dynamic objects per frame rendering, represents the interactive delay from input to screen update in each frame rendering under the R0 frame monitoring point, ρ represents the weight factor of the interactive delay from input to screen update in each frame rendering, 3D_FUL0 represents the interactive delay from input to screen update in standard each frame rendering, represents the data stream bandwidth in each frame rendering under the R0-th frame monitoring point, σ represents the weight factor of the data stream bandwidth in each frame rendering, and DSB0 represents the data stream bandwidth in standard each frame rendering.

[0062] The number of polygons directly affects the rendering burden of the GPU. The more polygons that need to be processed per frame, the greater the pressure on the GPU's vertex shading, rasterization, and pixel filling, which may cause the frame rate to drop. High-precision models (high polygon counts) require larger storage and transmission bandwidth. If the model data needs to be dynamically loaded or streamed from the server (such as cloud rendering), high-polygon models will take up more data stream bandwidth, and dynamic objects (such as vehicles and pedestrians) need to update their status (position, animation, physical simulation) per frame, which will increase the logical calculation burden of the CPU. If the CPU processing time is too long, it may delay the start time of GPU rendering, resulting in increased interactive delays.

[0063] When the system is running, it retrieves a mapping table of weight factors from a database and quickly extracts corresponding weight factors based on the current number of polygons in the 3D city per frame, the number of dynamic objects in the 3D city per frame, the interaction delay from input to screen update per frame, and the data flow bandwidth per frame. For example, these factors are the weight factor for the number of polygons in the 3D city per frame, the weight factor for the number of dynamic objects in the 3D city per frame, the weight factor for the interaction delay from input to screen update per frame, and the weight factor for the data flow bandwidth per frame. This mapping table defines a clear set of association rules that converts the specific values of the number of polygons in the 3D city per frame, the number of dynamic objects in the 3D city per frame, the interaction delay from input to screen update per frame, and the data flow bandwidth per frame into their corresponding weight factors. This mechanism effectively enables dynamic acquisition of weight factors, whether achieving a one-to-one exact match or a many-to-one relationship where multiple parameters are aggregated into a single weight.

[0064] Furthermore, the specific analysis process for comprehensive evaluation of the hardware load assessment coefficient and the software load assessment coefficient is as follows: obtaining the weight factor of the hardware load assessment coefficient and the weight factor of the software load assessment coefficient from the database, assigning weights to the hardware load assessment coefficient and the software load assessment coefficient, combining the hardware load assessment coefficient and the software load assessment coefficient, and processing them to obtain the rendering load assessment coefficient.

[0065] In this embodiment, the specific method for obtaining the rendering load evaluation coefficient is:

[0066]

[0067] Where RLEF is the rendering load evaluation coefficient, which is used to evaluate the impact of hardware load and software load on rendering load. HLA is the hardware load evaluation coefficient. It is expressed as the weight factor of the hardware load assessment coefficient, SLA is expressed as the software load assessment coefficient, and ω is expressed as the weight factor of the software load assessment coefficient.

[0068] When the system is running, it retrieves a mapping table of weight factors from the database and quickly extracts the corresponding weight factors based on the current hardware and software load assessment coefficients, such as the weight factor for the hardware and software load assessment coefficients. This mapping table defines a clear set of association rules that convert the hardware and software load assessment coefficients into their corresponding weight factors. This mechanism effectively enables dynamic acquisition of weight factors, whether achieving a one-to-one exact match or a many-to-one relationship where multiple parameters are aggregated into a single weight.

[0069] Furthermore, the specific threshold comparison method for comparing the hardware load evaluation coefficient, the software load evaluation coefficient and the rendering load evaluation coefficient with the hardware load evaluation threshold, the software load evaluation threshold and the rendering load evaluation threshold is as follows: comparing the hardware load evaluation coefficient with the hardware load evaluation threshold; if the hardware load evaluation coefficient is less than the hardware load evaluation threshold, directly performing real-time rendering on the sub-area; if the hardware load evaluation coefficient is greater than or equal to the hardware load evaluation threshold, executing a hardware load optimization scheme for the real-time rendering of the sub-area; after executing the hardware load optimization scheme, continuing to compare the hardware load evaluation coefficient with the hardware load evaluation threshold; if the hardware load evaluation coefficient is still greater than or equal to the hardware load evaluation threshold, executing a software load optimization scheme for the real-time rendering of the sub-area; performing software load optimization scheme on the real-time rendering of the sub-area. Comparison of the load assessment coefficient and the software load assessment threshold. If the software load assessment coefficient is less than the software load assessment threshold, the sub-area is rendered in real time. If the software load assessment coefficient is greater than or equal to the software load assessment threshold, a rendering load optimization solution is implemented for the real-time rendering of the sub-area. Comparison of the rendering load assessment coefficient and the rendering load assessment threshold. If the rendering load assessment coefficient is less than the rendering load assessment threshold, the sub-area is rendered in real time. If the rendering load assessment coefficient is greater than or equal to the rendering load assessment threshold, a balanced rendering quality optimization solution is implemented for the real-time rendering of the sub-area. The hardware load optimization solution includes dynamic resolution adjustment. The software load optimization solution includes interactive event priority scheduling and multi-threaded parallel processing. The rendering load optimization solution includes cone-based GPU culling technology.

[0070] In this embodiment, the hardware load is evaluated to obtain a hardware load evaluation coefficient, and a hardware load evaluation threshold is obtained from a database. If the hardware load evaluation coefficient is less than the hardware load evaluation threshold, the sub-area is directly rendered in real time. If the hardware load evaluation coefficient is greater than or equal to the hardware load evaluation threshold, a hardware load optimization scheme is executed for the real-time rendering of the sub-area, and the computer automatically reduces the resolution according to a preset multiple. If the hardware load evaluation coefficient is still greater than or equal to the hardware load evaluation threshold at this time, the computer performs interactive event priority scheduling and multi-threaded parallel processing on the system. The software load is evaluated to obtain a software load evaluation coefficient. If the software load evaluation coefficient is less than the software load evaluation threshold, the sub-area is rendered in real time. If the software load evaluation coefficient is greater than or equal to the software load evaluation threshold, the computer performs a cone-based GPU culling technology on the system. The rendering load is evaluated to obtain a rendering load evaluation coefficient. If the rendering load evaluation coefficient is less than the rendering load evaluation threshold, the sub-area is rendered in real time. If the rendering load evaluation coefficient is greater than or equal to the rendering load evaluation threshold, a balanced rendering quality optimization scheme is executed for the real-time rendering of the sub-area.

[0071] In a specific embodiment, the system loads the intersection sub-area (building model, vehicles, pedestrians, dynamic lights) with an initial resolution of 1080P. The hardware load optimization solution is that the computer automatically reduces the resolution according to a preset multiple, for example, reducing the resolution from 1080P to 720P. The software load optimization solution is to suspend non-critical background data loading tasks and assign shadow calculations to dedicated computing threads. The rendering load optimization solution is to use the viewing cone to cull 30% of the back of the building and the underground pipeline model outside the field of view. The drawing calls are reduced to 8,500 times and the number of triangles is reduced to 5.6 million.

[0072] Furthermore, the specific process of the balanced rendering quality optimization plan is: if after executing the hardware load optimization plan, the software load optimization plan and the rendering load optimization plan, if the rendering load evaluation coefficient is still greater than or equal to the rendering load evaluation threshold, the balanced rendering quality optimization plan is executed for the real-time rendering of the sub-area, including performing polygon merging on the real-time rendering of the sub-area, reducing the number of polygons and not rendering the polygons facing away from the camera. If the rendering load evaluation coefficient is less than the rendering load evaluation threshold, the balanced rendering quality optimization plan will no longer be executed.

[0073] In this embodiment, the model within the sub-region is rendered in real time. After the hardware load optimization scheme, software load optimization scheme, and rendering load optimization scheme are implemented, the rendering load evaluation coefficient is still greater than or equal to the rendering load evaluation threshold. The glass curtain wall meshes of adjacent office buildings within the sub-region are merged from 20,000 polygons to 8,000 using the buffer method. The specific steps of the buffer method are to generate a buffer for the polygons (with the buffer distance set to 0) instead of the traditional Un i on operation to merge a large number of polygons. Rendering is stopped for structures such as air conditioner outdoor units and pipes that face away from the user's viewpoint, reducing 1,200 invisible faces. The degradation optimization is immediately stopped, and the current rendering quality is retained (the merged building model still displays the complete appearance, but the back details are no longer updated).

[0074] like Figure 3 As shown, it is a structural diagram of the city three-dimensional visualization interaction system based on virtual reality provided by an embodiment of the present application. The city three-dimensional visualization interaction system based on virtual reality provided by an embodiment of the present application includes: an interaction and data collection module, a data preprocessing module, a data analysis module and an optimization module: the interaction and data collection module is used to divide the city three-dimensional model into multiple sub-areas, and the user selects a sub-area and performs visual interaction on the model in the sub-area. During the user's visual interaction, the virtual reality computer renders the sub-area in real time and collects virtual reality three-dimensional hardware load related data and virtual reality three-dimensional software load related data; the data preprocessing module is used to preprocess the virtual reality three-dimensional hardware load related data and virtual reality three-dimensional software load related data to obtain virtual reality three-dimensional hardware load preprocessing data and virtual reality three-dimensional software load preprocessing data; the data analysis module is used to analyze the virtual reality three-dimensional hardware load Load preprocessing data to obtain a hardware load assessment coefficient, analyze the virtual reality three-dimensional software load preprocessing data to obtain a software load assessment coefficient, comprehensively evaluate the hardware load assessment coefficient and the software load assessment coefficient to obtain a rendering load assessment coefficient; an optimization module: used to perform threshold comparison on the hardware load assessment coefficient, the software load assessment coefficient and the rendering load assessment coefficient with the hardware load assessment threshold, the software load assessment threshold and the rendering load assessment threshold respectively, and according to the threshold comparison result of the hardware load assessment coefficient, execute a hardware load optimization scheme for the real-time rendering of the sub-area; if the hardware load assessment coefficient is still greater than or equal to the hardware load assessment threshold, execute a software load optimization scheme for the real-time rendering of the sub-area; according to the threshold comparison result of the software load assessment coefficient, execute a rendering load optimization scheme for the real-time rendering of the sub-area; according to the threshold comparison result of the rendering load assessment coefficient, execute a balanced rendering quality optimization scheme for the real-time rendering of the sub-area.

[0075] It will be understood by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0076] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0077] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0078] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0079] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0080] Obviously, those skilled in the art may make various modifications and variations to the present invention without departing from the spirit and scope of the present invention. Thus, if such modifications and variations fall within the scope of the claims and their equivalents, the present invention is intended to include such modifications and variations.

Claims

1. A virtual reality-based three-dimensional urban visualization interactive method, characterized in that: The following steps are involved: The 3D city model is divided into multiple sub-areas. The user selects a sub-area and performs visual interaction on the model in the sub-area. During the user's visual interaction, the virtual reality computer renders the sub-area in real time and collects data related to the virtual reality 3D hardware load and the virtual reality 3D software load. Preprocessing the virtual reality three-dimensional hardware load related data and the virtual reality three-dimensional software load related data to obtain virtual reality three-dimensional hardware load preprocessing data and virtual reality three-dimensional software load preprocessing data; Analyze virtual reality 3D hardware load preprocessing data to obtain a hardware load evaluation coefficient, analyze virtual reality 3D software load preprocessing data to obtain a software load evaluation coefficient, and comprehensively evaluate the hardware load evaluation coefficient and the software load evaluation coefficient to obtain a rendering load evaluation coefficient; Perform threshold comparisons on the hardware load assessment coefficient, the software load assessment coefficient, and the rendering load assessment coefficient with the hardware load assessment threshold, the software load assessment threshold, and the rendering load assessment threshold, respectively; based on the threshold comparison results of the hardware load assessment coefficient, execute a hardware load optimization solution for the real-time rendering of the sub-area; if the hardware load assessment coefficient is still greater than or equal to the hardware load assessment threshold, execute a software load optimization solution for the real-time rendering of the sub-area; based on the threshold comparison results of the software load assessment coefficient, execute a rendering load optimization solution for the real-time rendering of the sub-area; based on the threshold comparison results of the rendering load assessment coefficient, execute a balanced rendering quality optimization solution for the real-time rendering of the sub-area; Hardware load optimization solutions include dynamic resolution adjustment; Software load optimization solutions include interactive event priority scheduling and multi-threaded parallel processing; Rendering load optimization solutions include frustum-based GPU culling technology; The specific process of the balanced rendering quality optimization solution is as follows: If after executing the hardware load optimization plan, the software load optimization plan and the rendering load optimization plan, if the rendering load evaluation coefficient is still greater than or equal to the rendering load evaluation threshold, the balanced rendering quality optimization plan is executed for the real-time rendering of the sub-area, including performing polygon merging on the real-time rendering of the sub-area, reducing the number of polygons and not rendering the polygons facing away from the camera. If the rendering load evaluation coefficient is less than the rendering load evaluation threshold, the balanced rendering quality optimization plan will no longer be executed.

2. The method for interactive three-dimensional visualization of a city based on virtual reality according to claim 1, wherein: The specific interactive process of visually interacting with the models in the sub-areas is as follows: The model in the sub-area is acquired through laser scanning, imported into the VR development engine, and visual cues are added in the VR environment. The model in the sub-area is visually interacted with through handle buttons, gestures, and voice input. The visual interaction includes dragging and zooming, rotating to observe the internal structure, triggering animation demonstrations, day and night switching, and weather adjustment.

3. The method for interactive three-dimensional visualization of a city based on virtual reality according to claim 1, wherein: The specific process of collecting data related to virtual reality 3D hardware load and virtual reality 3D software load is as follows: Collect VR 3D hardware load and VR 3D software load data using the operating system's built-in tools and Unity's Profiler. VR 3D hardware load-related data includes the VR computer's CPU core utilization, the VR computer's GPU utilization, the VR computer's video memory usage percentage, and the VR computer's multi-threaded scheduling efficiency; The data related to the load of virtual reality 3D software includes the number of polygons in the 3D city during each frame rendering, the number of dynamic 3D objects in the city during each frame rendering, the interactive delay from input to screen update during each frame rendering, and the data stream bandwidth during each frame rendering.

4. The method for interactive three-dimensional visualization of a city based on virtual reality according to claim 1, wherein: The specific process of preprocessing the virtual reality 3D hardware load-related data and the virtual reality 3D software load-related data is as follows: The virtual reality 3D hardware load-related data and the virtual reality 3D software load-related data are associated through timestamps, and the short-term missing data in the virtual reality 3D hardware load-related data and the virtual reality 3D software load-related data are supplemented using the linear interpolation method. The time interval of the virtual reality 3D hardware load-related data and the virtual reality 3D software load-related data is detected, and the long-term missing data is marked and deleted. The virtual reality 3D hardware load-related data and the virtual reality 3D software load-related data are smoothed using a low-pass filter, and the virtual reality 3D hardware load-related data and the virtual reality 3D software load-related data are normalized using the Z-score normalization method.

5. The method for interactive three-dimensional visualization of a city based on virtual reality according to claim 1, wherein: The specific analysis process of analyzing the virtual reality 3D hardware load preprocessing data is as follows: Obtain a weight factor of the VR computer's CPU core utilization, a weight factor of the VR computer's GPU utilization, a weight factor of the VR computer's video memory occupancy percentage, and a weight factor of the VR computer's multi-threaded scheduling efficiency from a database, arrange the VR computer's CPU core utilization, the VR computer's GPU utilization, the VR computer's video memory occupancy percentage, and the VR computer's multi-threaded scheduling efficiency in a time series, assign weights to the VR computer's CPU core utilization, the VR computer's GPU utilization, the VR computer's video memory occupancy percentage, and the VR computer's multi-threaded scheduling efficiency, compare standard values of the VR computer's CPU core utilization, the VR computer's GPU utilization, the VR computer's video memory occupancy percentage, and the VR computer's multi-threaded scheduling efficiency with average values of the VR computer's CPU core utilization, the VR computer's GPU utilization, the VR computer's video memory occupancy percentage, and the VR computer's multi-threaded scheduling efficiency, and process them to obtain a hardware load evaluation coefficient.

6. The method for interactive three-dimensional visualization of a city based on virtual reality according to claim 1, wherein: The specific analysis process of analyzing the virtual reality 3D software load preprocessing data is as follows: The weight factor of the number of polygons in the three-dimensional city when rendering each frame, the weight factor of the number of dynamic objects in the three-dimensional city when rendering each frame, the weight factor of the interactive delay from input to screen update in each frame rendering, and the weight factor of the data stream bandwidth in each frame rendering are obtained from the database, the number of polygons in the three-dimensional city when rendering each frame, the number of dynamic objects in the three-dimensional city when rendering each frame, the interactive delay from input to screen update in each frame rendering, and the data stream bandwidth in each frame rendering are arranged in frame sequence, and weights are assigned to the number of polygons in the three-dimensional city when rendering each frame, the number of dynamic objects in the three-dimensional city when rendering each frame, the interactive delay from input to screen update in each frame rendering, and the data stream bandwidth in each frame rendering. The standard values of the number of polygons in the three-dimensional city when rendering each frame, the number of dynamic objects in the three-dimensional city when rendering each frame, the interactive delay from input to screen update in each frame rendering, and the data stream bandwidth in each frame rendering are compared with the average values of the number of polygons in the three-dimensional city when rendering each frame, the number of dynamic objects in the three-dimensional city when rendering each frame, the interactive delay from input to screen update in each frame rendering, and the data stream bandwidth in each frame rendering, and the software load assessment coefficient is obtained by processing.

7. The method for interactive three-dimensional visualization of a city based on virtual reality according to claim 1, wherein: The specific analysis process of comprehensively evaluating the hardware load evaluation coefficient and the software load evaluation coefficient is as follows: The weight factors of the hardware load assessment coefficient and the software load assessment coefficient are obtained from the database, weights are assigned to the hardware load assessment coefficient and the software load assessment coefficient, the hardware load assessment coefficient and the software load assessment coefficient are combined, and processed to obtain a rendering load assessment coefficient.

8. The method for interactive three-dimensional visualization of a city based on virtual reality according to claim 1, wherein: The specific threshold comparison method for comparing the hardware load evaluation coefficient, the software load evaluation coefficient, and the rendering load evaluation coefficient with the hardware load evaluation threshold, the software load evaluation threshold, and the rendering load evaluation threshold, respectively, is as follows: The hardware load assessment coefficient is compared with the hardware load assessment threshold. If the hardware load assessment coefficient is less than the hardware load assessment threshold, the sub-area is directly rendered in real time. If the hardware load assessment coefficient is greater than or equal to the hardware load assessment threshold, the hardware load optimization plan is executed for the real-time rendering of the sub-area. After executing the hardware load optimization plan, the hardware load assessment coefficient is continued to be compared with the hardware load assessment threshold. If the hardware load assessment coefficient is still greater than or equal to the hardware load assessment threshold, the software load optimization plan is executed for the real-time rendering of the sub-area, and the software load assessment coefficient is compared with the software load assessment threshold. If the software load assessment coefficient is less than the software load assessment threshold, the sub-area is rendered in real time. If the software load assessment coefficient is greater than or equal to the software load assessment threshold, the rendering load optimization plan is executed for the real-time rendering of the sub-area, and the rendering load assessment coefficient is compared with the rendering load assessment threshold. If the rendering load assessment coefficient is less than the rendering load assessment threshold, the sub-area is rendered in real time. If the rendering load assessment coefficient is greater than or equal to the rendering load assessment threshold, the balanced rendering quality optimization plan is executed for the real-time rendering of the sub-area.

9. The city three-dimensional visualization interactive system based on virtual reality is characterized by: Including interaction and data collection module, data preprocessing module, data analysis module and optimization module: Interaction and data collection module: used to divide the urban 3D model into multiple sub-areas. Users select a sub-area and interact with the model in the sub-area visually. During the user's visual interaction, the virtual reality computer renders the sub-area in real time and collects data related to the virtual reality 3D hardware load and the virtual reality 3D software load. Data preprocessing module: used for preprocessing virtual reality 3D hardware load related data and virtual reality 3D software load related data to obtain virtual reality 3D hardware load preprocessing data and virtual reality 3D software load preprocessing data; Data analysis module: used to analyze virtual reality 3D hardware load preprocessing data to obtain hardware load evaluation coefficients, analyze virtual reality 3D software load preprocessing data to obtain software load evaluation coefficients, and conduct comprehensive evaluation of the hardware load evaluation coefficients and software load evaluation coefficients to obtain rendering load evaluation coefficients; Optimization module: used to perform threshold comparison on the hardware load assessment coefficient, the software load assessment coefficient and the rendering load assessment coefficient with the hardware load assessment threshold, the software load assessment threshold and the rendering load assessment threshold respectively; based on the threshold comparison result of the hardware load assessment coefficient, execute the hardware load optimization plan for the real-time rendering of the sub-area; if the hardware load assessment coefficient is still greater than or equal to the hardware load assessment threshold, execute the software load optimization plan for the real-time rendering of the sub-area; based on the threshold comparison result of the software load assessment coefficient, execute the rendering load optimization plan for the real-time rendering of the sub-area; based on the threshold comparison result of the rendering load assessment coefficient, execute the balanced rendering quality optimization plan for the real-time rendering of the sub-area; Hardware load optimization solutions include dynamic resolution adjustment; Software load optimization solutions include interactive event priority scheduling and multi-threaded parallel processing; Rendering load optimization solutions include frustum-based GPU culling technology; The specific process of the balanced rendering quality optimization solution is as follows: If after executing the hardware load optimization plan, the software load optimization plan and the rendering load optimization plan, if the rendering load evaluation coefficient is still greater than or equal to the rendering load evaluation threshold, the balanced rendering quality optimization plan is executed for the real-time rendering of the sub-area, including performing polygon merging on the real-time rendering of the sub-area, reducing the number of polygons and not rendering the polygons facing away from the camera. If the rendering load evaluation coefficient is less than the rendering load evaluation threshold, the balanced rendering quality optimization plan will no longer be executed.

Images