Unmanned aerial vehicle visual scene simulation device and method based on programmed generation

Abstract

The invention discloses an unmanned aerial vehicle visual scene simulation device and method based on programmed generation, and relates to the field of man-machine testing and testing. The problems that an existing unmanned aerial vehicle visual simulation platform is single in simulation scene, complex in scene construction, lack of self-defined weather simulation and the like are solved. The device comprises a data reading and preprocessing module used for reading real scene data, preprocessing the data and outputting basic terrain model data and open source street map data; the terrain generation module is used for generating a virtual terrain; the city generation module is used for realizing the construction of a virtual city environment according to the city data and the city element composition method; the object generation module is used for acquiring a generation range and object type parameters specified by a user and generating a model required by the user; and the weather generation module is used for adding a self-defined weather simulation environment model in the constructed virtual city environment and setting a delay event to realize a self-defined weather simulation function. The method is also suitable for the field of complex unmanned aerial vehicle visual simulation scenes.

Description

Technical Field

[0001] The invention relates to the field of unmanned aerial vehicle experiments and tests. Background Art

[0002] Verifying drone-related functions in virtual simulation scenarios can accelerate intelligent drone testing. Rich and high-fidelity virtual simulation scenarios are the key to effective visual simulation of drones. However, existing drone visual simulation platforms have the following main problems:

[0003] (1) Single simulation scene. Currently, most UAV simulation platforms only provide a small number of fixed simulation scenes. Many simulation platforms have scenes with single elements and poor rendering effects.

[0004] (2) The process of scene construction is complicated. Some UAV simulation platforms provide virtual simulation scene editing or importing functions. However, the process of scene construction is very complicated and time-consuming, and requires users to have a certain understanding of the field of virtual scene construction.

[0005] (3) Lack of customizable weather environment simulation function. Most drone simulation platforms do not provide weather environment simulation or cannot adjust the weather type and weather intensity during the drone flight according to the user's settings.

[0006] In summary, the existing UAV simulation platform can hardly meet the needs of rapid and dynamic testing of UAVs. Summary of the invention

[0007] The present invention aims to solve the problems of the existing UAV visual simulation platform, such as single simulation scene, complex scene construction and lack of customized weather simulation. To solve the above technical problems, the present invention is implemented through the following technical solutions:

[0008] Solution 1: The present invention proposes a UAV visual scene simulation device based on programmatic generation, the device comprising:

[0009] Data reading and preprocessing module, used to read real scene data and preprocess the data, output basic terrain model data and point, line and surface geometric data corresponding to open source street map data;

[0010] The terrain generation module is used to add noise and water erosion simulation to the basic terrain model data output by the data reading and preprocessing module to generate the final virtual terrain;

[0011] A city generation module, used to construct a virtual city environment based on the point, line, and surface geometric data corresponding to the open source street map data output by the data reading and preprocessing module and the virtual terrain generated by the terrain generation module;

[0012] An object generation module, configured to obtain the generation range parameters and object type parameters of the user within the area, use the Poisson disk sampling algorithm to sample position points within the area for layout of the generation model, and generate a three-dimensional model at the coordinates of the sampled position points;

[0013] A weather generation module, configured to add a custom weather simulation environment model to the constructed virtual city environment and set delay events to implement the custom weather simulation function.

[0014] Furthermore, a preferred implementation is provided. Reading the real-scene data in the data reading and preprocessing module includes elevation model data and vector data of vegetation and water systems.

[0015] Furthermore, a preferred implementation is provided. The terrain generation module further includes the step of enriching terrain features using the Perlin noise algorithm.

[0016] Furthermore, a preferred implementation is provided. The terrain generation module using the water erosion algorithm further includes the step of simulating erosion of the terrain, and setting the precipitation, soil erodibility, and number of iterations as variable parameters for controlling the corrosion degree of the virtual terrain.

[0017] Furthermore, a preferred implementation is provided. The city generation module includes a road network generation sub-module, a building generation sub-module, and a city accessory generation sub-module;

[0018] The road network generation sub-module is configured to construct a three-dimensional road model according to road data;

[0019] The building generation sub-module is configured to construct buildings in the UAV vision scene in a modular manner;

[0020] The city accessory generation sub-module is configured to generate street accessories and other city accessories.

[0021] Solution 2: A UAV vision scene simulation method based on procedural generation. The method is implemented based on the device described in Solution 1. The method includes the following steps:

[0022] Step 1: Read real-scene data and preprocess the data, and output the basic terrain model data and the point, line, and surface geometric data corresponding to the open-source street map data;

[0023] Step 2: Add noise and water erosion simulation to the basic terrain model data output in Step 1 to generate the final virtual terrain;

[0024] Step 3: Construct a virtual city environment model based on the point, line, and surface geometric data corresponding to the open-source street map data output in Step 1 and the virtual terrain generated in Step 2;

[0025] Step 4: Based on the virtual city environment model constructed in Step 3, obtain the types and range parameters of the generated objects input by the user, use the Poisson disk sampling algorithm to sample position points in the area for laying out the generated model, and generate a 3D model at the coordinates of the sampled position points.

[0026] Step 5: Add a custom weather simulation environment model to the virtual city environment constructed in Step 3, and set a delay event to implement the custom weather simulation function.

[0027] Furthermore, a preferred implementation is provided. Step 1 further includes the step of converting the input data coordinate system into a plane rectangular coordinate system using the haversine formula.

[0028] Furthermore, a preferred implementation is provided. The method of converting the input data coordinate system into a plane rectangular coordinate system using the haversine formula is as follows:

[0029]

[0030] d = R·c (1)

[0031] where d is the great circle distance between two points on the sphere, R is the radius of the sphere, and here the equatorial radius of the earth, 6371 km, is taken. λ1 and λ2 are the latitudes and longitudes of the two points respectively.

[0032] Solution 3: A computer device includes a memory and a processor. A computer program is stored in the memory. When the processor runs the computer program stored in the memory, the processor executes the method described in any one of Solution 2.

[0033] Solution 4: A computer-readable storage medium stores a computer program. When the computer program is executed by a processor, the steps of the method described in any one of Solution 2 are implemented.

[0034] The beneficial effects of the present invention are as follows:

[0035] The drone vision scene simulation device and method based on procedural generation according to the present invention realize terrain generation with rich details and real corrosion simulation through input data and the Berlin noise and water erosion algorithms; automatically generate geomorphic features for the terrain using information such as height and slope; and realize the generation of a virtual city with a real urban layout based on the user-input OSM data by combining data with rules and the rules of urban component elements.

[0036] The present invention realizes the function of generating models within a specified range based on the Poisson disk sampling algorithm, and sets a delay event to realize the function of custom weather simulation, ensuring the flexibility of the generated scene.

[0037] This application can quickly generate a UAV virtual simulation scenario with the characteristics of the input data according to the elevation model data, open source street map data, and vector data of vegetation and water systems input by the user, and combines a random generation algorithm, and provides an object generation function and a weather generation function to ensure that the user can generate the required objects in a specified area in the simulation scenario according to their own needs and set the weather transformation process during the simulation.

[0038] The present invention can quickly generate a virtual simulation scenario with high fidelity with less input to meet the needs of UAV visual simulation for high-quality virtual scenarios.

[0039] The method described in this application further includes generating a diverse, highly realistic virtual simulation scenario with a real urban layout according to the geospatial data and parameters input by the user to meet the requirements of complex UAV visual simulation scenarios.

[0040] The present invention is also applicable to a method and platform for generating a UAV virtual simulation scenario based on procedural generation. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 It is a schematic diagram of the overall framework of the platform of the UAV vision scenario simulation device based on procedural generation described in Embodiment 1.

[0042] Figure 2 It is a schematic diagram of the composition of the UAV simulation scenario described in Embodiment 11.

[0043] Figure 3 It is a schematic diagram of the composition of each module in the UAV vision scenario simulation device based on procedural generation described in Embodiment 1.

[0044] Figure 4 It is a schematic diagram of the composition of the roads in the road network generation sub-module described in Embodiment 11.

[0045] Figure 5 It is a schematic diagram of the virtual simulation scenario corresponding to the generation area described in Embodiment 11.

[0046] Among them, (a) is a schematic diagram of road generation, (b) is a schematic diagram of building generation, (c) is a schematic diagram of virtual scene generation, (d) is a schematic diagram of vehicle generation, and (e) is a schematic diagram of urban accessory generation. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0047] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the following will clearly and completely describe the technical solutions in the embodiments of this application in conjunction with the accompanying drawings in the embodiments of this application. Obviously, the described embodiments are part of the embodiments of this application, rather than all of them.

[0048] Embodiment 1. This embodiment proposes an unmanned aerial vehicle (UAV) visual scene simulation device based on procedural generation. The device includes:

[0049] A data reading and preprocessing module, configured to read real-scene data and preprocess the data, and output point, line, and surface geometric data corresponding to the basic terrain model data and the open-source street map data;

[0050] A terrain generation module, configured to add noise and water erosion simulation to the basic terrain model data output by the data reading and preprocessing module to generate a final virtual terrain;

[0051] A city generation module, configured to construct a virtual city environment according to the point, line, and surface geometric data corresponding to the open-source street map data output by the data reading and preprocessing module and the virtual terrain generated by the terrain generation module;

[0052] An object generation module, configured to obtain the generation range parameters and object type parameters of the user in the area, use the Poisson disk sampling algorithm to sample position points in the area to layout the generation model, and generate a three-dimensional model at the coordinates of the sampled position points;

[0053] A weather generation module, configured to add a custom weather simulation environment model to the constructed virtual city environment and set delay events to implement the custom weather simulation function.

[0054] Embodiment 2. This embodiment further limits the unmanned aerial vehicle (UAV) visual scene simulation device based on procedural generation described in Embodiment 1. Reading the real-scene data in the data reading and preprocessing module includes elevation model data and vector data of vegetation and water systems.

[0055] Embodiment 3. This embodiment further limits the unmanned aerial vehicle (UAV) visual scene simulation device based on procedural generation described in Embodiment 1. The terrain generation module further includes a step of enriching terrain features using the Perlin noise algorithm.

[0056] Embodiment 4. This embodiment further limits the unmanned aerial vehicle (UAV) visual scene simulation device based on procedural generation described in Embodiment 1. The water erosion algorithm terrain generation module further includes a step of simulating the erosion of the terrain, and setting the precipitation, soil corrosivity, and number of iterations as variable parameters to control the corrosion degree of the virtual terrain.

[0057] Embodiment 5. This embodiment further defines the drone vision scene simulation device based on procedural generation described in Embodiment 1. The city generation module includes a road network generation sub-module, a building generation sub-module, and a city accessory generation sub-module;

[0058] The road network generation sub-module is used to construct a 3D road model according to road data;

[0059] The building generation sub-module is used to construct buildings in the drone vision scene in a modular manner;

[0060] The city accessory generation sub-module is used to generate street accessories and other city accessories.

[0061] Embodiment 6. This embodiment proposes a method for simulating a drone vision scene based on procedural generation. The method is implemented based on the device described in Embodiment 1. The method includes the following steps:

[0062] Step 1: Read real scene data and preprocess the data, and output the basic terrain model data and the point, line, and surface geometric data corresponding to the open-source street map data;

[0063] Step 2: Add noise and water erosion simulation to the basic terrain model data output in Step 1 to generate the final virtual terrain;

[0064] Step 3: Construct a virtual city environment model based on the point, line, and surface geometric data corresponding to the open-source street map data output in Step 1 and the virtual terrain generated in Step 2;

[0065] Step 4: Based on the virtual city environment model constructed in Step 3 and obtain the type and range parameters of the generated objects input by the user. Use the Poisson disk sampling algorithm to sample position points in the area to layout the generated model, and generate a 3D model at the coordinates of the sampled position points;

[0066] Step 5: Add a custom weather simulation environment model to the virtual city environment constructed in Step 3, and set delay events to implement the custom weather simulation function.

[0067] Embodiment 7. This embodiment further defines the method for simulating a drone vision scene based on procedural generation described in Embodiment 6. Step 1 further includes the step of converting the input data coordinate system into a plane rectangular coordinate system using the haversine formula.

[0068] Embodiment 8. This embodiment further defines the method for simulating a drone vision scene based on procedural generation described in Embodiment 7. The method of converting the input data coordinate system into a plane rectangular coordinate system using the haversine formula is as follows:

[0069]

[0070] d = R·c (2)

[0071] Wherein, d is the great circle distance between two points on the spherical surface, R is the radius of the sphere, and here the equatorial radius of the earth is taken as 6,371 km, λ1 and λ2 are the latitudes and longitudes of the two points respectively.

[0072] Embodiment Nine. This embodiment provides a computer device, including a memory and a processor. A computer program is stored in the memory. When the processor runs the computer program stored in the memory, the processor executes the method described in any one of Embodiments Six to Eight.

[0073] Embodiment Ten. This embodiment provides a computer-readable storage medium. The computer-readable storage medium stores a computer program. When the computer program is executed by a processor, the steps of the method described in any one of Embodiments Six to Eight are implemented.

[0074] Embodiment Eleven. This embodiment provides an example, which is used to explain the above Embodiments One to Ten. The specific example is as follows:

[0075] See Figures 1 to 5 To describe this embodiment, this embodiment provides an unmanned aerial vehicle (UAV) vision scene simulation device and method based on procedural generation, establishes a scene construction method driven jointly by data and knowledge, and combines user data and the procedural generation method used in virtual scene generation to realize the on-demand automatic generation of a high-fidelity simulation scene. It solves the problems of single simulation scenes, complex scene construction, and lack of custom weather simulation in existing UAV vision simulation platforms. See Figure 1 To describe this embodiment, the device described in this embodiment can generate a high-fidelity virtual simulation scene with the urban layout of the input data according to the geographical data input by the user and a small number of parameter settings, meeting the requirements of UAV vision simulation for high-quality virtual scenes.

[0076] See Figure 2 To describe this embodiment, this embodiment splits the UAV flight scene into four parts: terrain, ground surface, ground objects and cities, and weather according to the logical relationship of scene elements, and performs procedural generation and combination on the four parts respectively to form the final virtual simulation environment. As Figure 3 shown, this device is divided into five modules: data reading and preprocessing module, terrain generation module, city generation module, object generation module, and weather generation module, which jointly implement the function of generating the UAV virtual simulation scene.

[0077] Among them, the data reading and preprocessing module is used to receive the real-scenario data input by the user and perform preprocessing operations on the data. To ensure that the input data is easily obtainable and the generated scenario has a high fidelity compared to the real scenario. This device uses Digital Elevation Model (DEM) data, OpenStreetMap (OSM) data, and vector data of vegetation and water systems to generate virtual scenarios. The input data records the spatial position characteristics of geographical features through spatial coordinate information. To ensure that the scenarios generated based on multi-source data maintain consistency in scale and spatial relationships.

[0078] The present invention uses the haversine formula to transform the input data coordinate system into a local plane rectangular coordinate system. Specifically, the present invention determines the point with the smallest latitude and longitude in the dataset as the origin of the local coordinate system, and then calculates the spherical distance between each data point and the origin through the haversine formula and calculates the relative coordinates of the data points in the local coordinate system based on this. The haversine formula is calculated as follows:

[0079]

[0080] d = R·c (3)

[0081] Among them, d is the great circle distance between two points on the sphere, R is the radius of the sphere, and here the equatorial radius of the earth is taken as 6371 km. λ1 and λ2 are the latitudes and longitudes of the two points respectively.

[0082] (2) Terrain generation module

[0083] The terrain generation module based on erosion simulation generates a virtual terrain with physical erosion characteristics according to the DEM data output by the data reading and preprocessing module. This module first generates a basic terrain model according to the DEM data, and then uses the Perlin noise algorithm to enrich the detailed features of the terrain to simulate the microscopic terrain undulations in the real world and improve the complexity of the terrain. In addition, to simulate the corrosion of the terrain surface in the natural world by various natural factors, this module introduces a water erosion algorithm to perform erosion simulation on the terrain and sets variables such as precipitation, soil corrosivity, and number of iterations to control the corrosion degree of the virtual terrain. Finally, during the terrain generation process, this module generates multiple layers for the terrain according to the slope, height, and occlusion information of the terrain, and generates different ground surfaces for the terrain accordingly.

[0084] (3) City generation module

[0085] The data- and knowledge-driven city generation module is responsible for constructing a virtual city environment based on the city data parsed by the data input and preprocessing module and the composition rules of city elements. To quickly construct a city in the virtual environment, this module splits city generation into three parts: road network generation, building generation, and city accessory generation, which respectively correspond to three sub-modules.

[0086] (a) Road network generation sub-module

[0087] The road network generation sub-module is responsible for constructing a 3D road model based on road data. As Figure 4 shown, the present invention splits the road into two parts: intersections and road segments and processes these two parts separately. The present invention first traverses all the points in the road line and marks the points as two types: intersection points and road segment points. For the intersection area, the present invention realizes the planarization of the intersection by setting the height values of all vertices in the area to the average height of the boundary points of the area. Then, the connection lines of the intersection points during the splitting of the intersection road segments are merged to generate the road surface of the intersection. For the road segment area, according to the road type, different widths are set for different types of roads, and then a plane section is scanned along the road line and a 3D model of the road surface of the road segment part is generated based on the width. Finally, the invention generates a plane model equal in size to the road boundary based on the road node coordinates and subtracts it from the road to obtain the 3D model of the road surface of the road segment part.

[0088] (b) Building generation sub-module

[0089] The building generation sub-module realizes the construction of buildings in the scene in a modular way. The present invention splits building generation into two parts: building geometry generation and building appearance generation. Building geometry generation converts the building contour line into a geometry based on the area contour and height information of the building in the OSM data, which serves as the basic part of the building. The appearance generation of the building generates diverse appearances for the building in a modular way. Since urban buildings often have common or similar geometric features and structures, most buildings can be attributed to the combination of some simple structures. The present invention splits the building into different layers, and each layer is further split into four modules: entrance, window, wall, and corner. For each building, this module realizes the generation of the building appearance according to the width and height of the building and the aspect ratio of the model assets used.

[0090] (c) City accessory generation sub-module

[0091] Urban accessories mainly refer to other components in the city except for the core elements such as buildings and roads that determine the urban layout. The generation of urban accessories in this device is mainly divided into two parts: one is the generation of street accessories. This part samples the road data output by the road network generation module to obtain the urban accessory generation points on both sides of the road, and places models such as green belts and street lamps at the positions of the generation points. The other is the generation of urban accessories built within the area. In addition to street accessories, there are also areas in the city where urban elements other than streets and buildings gather, such as woods, parks, and rivers. For the generation of this part of urban accessories, this embodiment divides these areas according to OSM data and vector data and combines different urban accessory models corresponding to the areas in different types of areas.

[0092] (4) Object generation module

[0093] The object generation module based on Poisson disk sampling enables the user to generate the required models within the specified range of the generated scene as needed to ensure the diversity of the generated scene. This module obtains the generation range parameters and object type parameters specified by the user, and then uses the Poisson disk sampling algorithm to sample position points within the specified range to ensure that the layout of the generated models is more realistic. Finally, the specified 3D model is generated at the position point coordinates.

[0094] (5) Weather generation module

[0095] The weather generation module based on delayed events needs to add a customizable weather simulation environment to the simulation environment to ensure the realism of the simulation environment. When the simulation starts running, the weather generation module reads the weather transformation time parameters set by the user and sets delayed events for each weather change time node. During the simulation process, when the simulation running time reaches the weather transformation time point set by the user, the delayed event will be triggered, and the weather in the simulation scene will be changed to the specified weather type and weather intensity.

[0096] As described above, in this embodiment, detailed terrain generation with real corrosion simulation is achieved through input data and the Berlin noise and water erosion algorithms; geomorphic features are automatically generated for the terrain using information such as height and slope; a virtual city with a real urban layout is generated based on the user-input OSM data by combining data and rules with the rules of urban composition elements. In addition, the function of generating models within a specified range is realized based on the Poisson disk sampling algorithm, and the function of customizing weather simulation is realized through delayed events, ensuring the flexibility of the generated scene. In summary, this application can generate diverse, highly realistic virtual simulation scenes with real urban layouts according to the geographical space data and parameters input by the user, meeting the requirements of complex UAV vision simulation scenes.

[0097] Example 1: Generate a virtual simulation scene corresponding to a region based on data

[0098] Step 1: First, obtain the digital elevation data, open-source street data, and vector data of vegetation and water systems corresponding to the region and import them as inputs into the platform.

[0099] Step 2: The data import module of the platform reads and parses the data, converts the spatial coordinate system information in the data to the local plane coordinate system, and converts the data into geometric data such as basic points, lines, and surfaces in the scene. Then, preprocess the data. According to the tags in the OSM data, extract the building data and road data, and perform topological checks on the road data.

[0100] Step 3: The terrain generation module receives the DEM data output by the data import module as input, generates the basic terrain according to the DEM data information, then calculates the two-dimensional noise map based on the Berlin noise and superimposes it on the terrain. Then, perform water erosion simulation on the terrain according to the iteration times and erosion intensity parameters set by the user. Finally, this module generates different layer information for the terrain according to the height and slope information, and generates the terrain surface appearance accordingly.

[0101] Step 4: The road generation part in the city generation module receives the road line data split by the data import module as input. It splits the road lines into intersections and road segments, and generates corresponding 3D models respectively. At the same time, sampling points will be used on both sides of the road segments to generate street accessories such as green belts and street lights.

[0102] Step 5: The city generation part in the city generation module receives the building data as input. It generates the basic building geometry according to the height information and range information of the buildings. Then, generate different appearances for different types of buildings according to the building type information.

[0103] Step 6: The city accessory generation part in the city generation module receives the range information of the data of vegetation, water systems, and basic infrastructure in the city such as parks. Generate city accessories of corresponding types in the corresponding areas.

[0104] Step 7: Merge the scenes generated in the above steps, and add collision information to the scenes.

[0105] Step 8: The user sets the type and range parameters of object generation, and the object generation module reads the parameters and generates objects in the scene when the simulation starts running.

[0106] Step 9: The user adds weather change time points and sets the weather type and weather intensity for each time point. The weather generation module reads the parameters and generates the corresponding weather simulation when the simulation starts running.

[0107] In summary, this embodiment can quickly generate a UAV virtual simulation scenario with the characteristics of the input data according to the DEM, OSM, vegetation, and water system data input by the user, and combines a random generation algorithm. It also provides an object generation function and a weather generation function, ensuring that the user can generate the required objects in a specified area in the simulation scenario according to their own needs and set the weather change process during the simulation. In summary, this platform can quickly generate a virtual simulation scenario with high fidelity with less input to meet the needs of UAV visual simulation for high-quality virtual scenarios.

[0108] Those skilled in the art can understand that the above is only the preferred embodiment of the present invention. The features described in each embodiment and / or claim of the present disclosure can be combined or combined in various ways, even if such combinations or combinations are not explicitly described in the present disclosure. It is not used to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

[0109] Although the preferred embodiments of the present invention have been described, those skilled in the art can make additional changes and modifications once they learn the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications falling within the scope of the present invention. Obviously, those skilled in the art can make various changes and variations to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention also intends to include these modifications and variations.

Claims

1. A drone vision scene simulation device based on procedural generation, characterized in that, The device includes: A data reading and preprocessing module, which is used to read real-scene data and preprocess the data, and output the point, line, and surface geometric data corresponding to the basic terrain model data and the open-source street map data; A terrain generation module, which is used to add noise and water erosion simulation to the basic terrain model data output by the data reading and preprocessing module to generate the final virtual terrain; A city generation module, which is used to construct a virtual city environment according to the point, line, and surface geometric data corresponding to the open-source street map data output by the data reading and preprocessing module and the virtual terrain generated by the terrain generation module; An object generation module, which is used to obtain the generation range parameters and object type parameters of the user in the area, use the Poisson disk sampling algorithm to sample position points in the area to layout the generation model, and generate a 3D model at the coordinates of the sampled position points; A weather generation module, which is used to add a custom weather simulation environment model to the constructed virtual city environment and set delay events to implement the custom weather simulation function.

2. The drone vision scene simulation device based on programmed generation according to claim 1, characterized in that, The real-scene data read in the data reading and preprocessing module includes elevation model data and vector data of vegetation and water systems.

3. The drone vision scene simulation device based on programmed generation according to claim 1, wherein, The terrain generation module also includes the step of enriching terrain features using the Perlin noise algorithm.

4. The drone vision scene simulation device based on programmed generation according to claim 1, characterized in that The water erosion algorithm terrain generation module also includes the step of simulating the erosion of the terrain, and setting the precipitation, soil corrosivity, and number of iterations as variable parameters to control the corrosion degree of the virtual terrain.

5. The drone vision scene simulation device based on programmed generation according to claim 1, wherein The city generation module includes a road network generation sub-module, a building generation sub-module, and a city accessory generation sub-module; The road network generation sub-module is used to construct a 3D road model according to the road data; The building generation sub-module is used to construct buildings in the UAV vision scene in a modular manner; The city accessory generation sub-module is used to generate street accessories and other city accessories.

6. The method for simulating an unmanned aerial vehicle visual scene based on procedural generation, characterized in that The method is implemented based on the device described in claim 1, and the method includes the following steps: Step 1: Read real-scene data and preprocess the data, and output the point, line, and surface geometric data corresponding to the basic terrain model data and the open-source street map data; Step 2: Add noise and water erosion simulation to the basic terrain model data output in Step 1 to generate the final virtual terrain; Step 3: Construct a virtual city environment model based on the point, line, and surface geometric data corresponding to the open-source street map data output in Step 1 and the virtual terrain generated in Step 2; Step 4: Based on the virtual city environment model constructed in Step 3 and obtain the type and range parameters of the generated objects input by the user, use the Poisson disk sampling algorithm to sample position points in the area to layout the generation model, and generate a 3D model at the coordinates of the sampled position points; Step 5: Add a custom weather simulation environment model to the virtual city environment constructed in Step 3 and set delay events to implement the custom weather simulation function.

7. The method for simulating an unmanned aerial vehicle vision scene based on programmed generation according to claim 6, wherein Step 1 also includes the step of converting the input data coordinate system into a plane rectangular coordinate system using the haversine formula.

8. The method for simulating an unmanned aerial vehicle vision scene based on programmatic generation according to claim 7, characterized in that, The method of converting the input data coordinate system into a plane rectangular coordinate system using the haversine formula is: d = R·c (1) Among them, d is the great circle distance between two points on the spherical surface, and R is the radius of the sphere. Here, the equatorial radius of the Earth, 6371 km, is taken. λ1 and λ2 are the latitudes and longitudes of the two points respectively.

9. A computer device, comprising a memory and a processor, characterized in that, A computer program is stored in the memory. When the processor runs the computer program stored in the memory, the processor executes the method according to any one of claims 6 to 8.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program. When the computer program is executed by a processor, the steps of the method according to any one of claims 6 to 8 are implemented.

Images