Weather forecast image-text scenarized real-time generation method and device

Abstract

The invention discloses a weather forecast image-text scenarized real-time generation method and device, and relates to the technical field of three-dimensional modeling, and the method comprises the steps: building a weather model and an industry model; acquiring real-time weather forecast data and real-time industry data, and calling a weather model matched with the real-time weather forecast data and an industry model matched with the real-time industry data; embedding the called weather model and industry model in a preset studio model; and simulating the influence of the weather model on the industry model to obtain a scenarized weather forecast effect. According to the method, the matched model is called according to the real-time meteorological data, and the animation can be generated in combination with the industry model and is presented to audiences in a direct animation demonstration form, so that the virtual weather forecast can provide accurate weather forecast information, and the virtual scene can be automatically generated, so that the user experience is improved. And the program attraction and the watching experience of audiences are enhanced while the use and operation difficulty is greatly simplified.

Description

Technical Field

[0001] This application relates to the field of 3D modeling technology, and particularly to a method and device for real-time generation of weather forecast graphic scenarios. Background Art

[0002] UE (Unreal Engine) technology has been increasingly widely used in studio weather forecasts in recent years, especially in creating immersive and interactive weather forecast programs. By combining virtual studio technology and XR (Extended Reality) technology, UE can create vivid and intuitive weather forecast scenarios, enabling viewers to feel the weather changes and their impacts as if they were on the scene.

[0003] The existing related technologies have the following three problems: 1. It is impossible to obtain weather telegram messages and other weather meteorological information in real time, and automatically call or generate relevant weather models or components based on the existing relevant information.

[0004] 2. It is impossible to connect weather information with related industries, such as agriculture, transportation, civil aviation, etc., and automatically call or generate relevant industry models or components based on the existing relevant information.

[0005] 3. The UE+XR+virtual studio technology used in weather forecasts still adopts the traditional method of modeling and manually analyzing data, which has high requirements for various requirements in the system (model construction, system shooting, real-time data analysis, etc.). As a result, it cannot be widely used due to high usage requirements. Summary of the Invention

[0006] Embodiments of this application provide a method and device for real-time generation of weather forecast graphic scenarios, aiming to solve the problems of insufficient support for weather models and industry models in the prior art, and high usage requirements.

[0007] On the one hand, embodiments of this application provide a method for real-time generation of weather forecast graphic scenarios, which includes: Establish weather models and industry models; Obtain real-time weather forecast data and real-time industry data, call the weather model matching the real-time weather forecast data, and the industry model matching the real-time industry data; Embed the called weather model and industry model into a preset studio model; Simulate the impact of the weather model on the industry model to obtain a scenario-based weather forecast effect.

[0008] In a possible implementation, if there is no weather model that matches the real-time weather forecast data, a matching model is generated in real time according to the real-time weather forecast data; if there is no industry model that matches the real-time industry data, a matching model is generated in real time according to the real-time industry data.

[0009] In a possible implementation, the real-time weather forecast data includes weather type and intensity data. When calling the weather model that matches the real-time weather forecast data, first call the single weather model that matches the weather type, and then set the quantity and size of the single weather model according to the intensity data. Multiple single weather models form the weather model.

[0010] In a possible implementation, the industry model includes a scene model and an object model. The scene model is a three-dimensional model established for specific industry-related scenes in the user's location area, and the object model is a three-dimensional model established for industry-related objects. After calling the industry model, the object model is embedded in the scene model.

[0011] In a possible implementation, the real-time industry data includes object type and quantity data. When calling the industry model that matches the real-time industry data, first call the object model that matches the object type, and then set the quantity of the object model according to the quantity data.

[0012] In a possible implementation, after simulating the impact of the weather model on the industry model, AIGC (Artificial Intelligence Generated Content) technology is used to generate corresponding videos and texts, and the videos and texts are displayed to achieve the effect of real-time generation of weather forecast graphic scenes.

[0013] In a possible implementation, multiple observation points are set in each scene model, and each observation point displays the scene model in different positions and directions. After the user selects an observation point, a video corresponding to the selected observation point is generated.

[0014] On the other hand, the embodiment of the present application also provides a device for real-time generation of weather forecast graphic scenes. The device includes: A model establishment module for establishing a weather model and an industry model; A model calling module for obtaining real-time weather forecast data and real-time industry data, calling the weather model that matches the real-time weather forecast data, and the industry model that matches the real-time industry data; A model embedding module for embedding the called weather model and industry model in a preset studio model; An impact simulation module for simulating the impact of the weather model on the industry model to obtain a scene-based weather forecast effect.

[0015] A method and device for real-time generation of weather forecast graphic scenarios in this application have the following advantages: According to real-time meteorological telegrams, meteorological data, etc., a correlation data model (precipitation level, storm, cold wave) of the meteorological model is generated in real time, and relevant model animations can be automatically generated according to real-time data, and presented to the audience in the form of direct animation demonstrations. If this technology can be widely used, the virtual weather forecast system can not only provide accurate weather forecast information, but also enhance the attractiveness of the program and the viewing experience of the audience by automatically designing virtual scenes while greatly simplifying the operation difficulty, enabling the audience to intuitively feel the impact of the weather, and more importantly, associating more usage scenarios in combination with AI (artificial intelligence) technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present application, and those of ordinary skill in the art can obtain other drawings based on these drawings without creative efforts.

[0017] Figure 1 It is a flowchart of a method for real-time generation of weather forecast graphic scenarios provided by an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0018] The following will clearly and completely describe the technical solutions in the embodiments of the present application in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only some embodiments of the present application, rather than all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present application.

[0019] Figure 1 It is a flowchart of a method for real-time generation of weather forecast graphic scenarios provided by an embodiment of the present application. An embodiment of the present application provides a method for real-time generation of weather forecast graphic scenarios, including: S100, establishing a weather model and an industry model.

[0020] Exemplarily, the weather model is a three-dimensional model established for various weathers. Taking rain as an example, the weather model will include clouds, wind, and raindrops, and the clouds and raindrops among them can be displayed through visual effects. Therefore, the corresponding three-dimensional model can be established through three-dimensional modeling, while the wind cannot be visually displayed, so it can be used as implicit data, and its effect can be displayed through the industry model.

[0021] The industry model includes a scene model and an object model. The scene model is a three-dimensional model established for specific industry-related scenes in the area where the user is located, and the object model is a three-dimensional model established for industry-related objects. The industries in the embodiments of the present application mainly include agriculture, transportation, and civil aviation. Therefore, the scene models related to the air traffic management of these industries can be farmland, agricultural production bases, roads, waterways, airports, the sky, etc. In order to improve the affinity when the user uses it, a three-dimensional model can be established for the representative scenes in the area where the user is located and used as the scene model. It should be understood that when establishing the scene model, the relevant objects in the scene also need to be reflected, such as the trees and houses around the farmland, the buildings on both sides of the road, etc., and the real state of these representative scenes should be reflected as much as possible.

[0022] For the above industries, the object model can be crops, fruit trees, vehicles, ships, airplanes, etc. When modeling these objects, several representative objects can be selected. For example, wheat seedlings, apple trees, litchi trees, etc. in agriculture, small cars, freight cars, trains, cargo ships, etc. in the transportation industry, and passenger planes, shuttle buses, etc. in civil aviation. After establishing these object models, corresponding attribute parameters need to be set according to the real attributes of these objects, such as weight, size, elasticity, etc., in order to simulate the real impact of the weather model on these object models in subsequent calculations and improve the authenticity of the weather forecast effect.

[0023] S110, Obtain real-time weather forecast data and real-time industry data, call the weather model matching the real-time weather forecast data, and the industry model matching the real-time industry data.

[0024] Exemplarily, the real-time weather forecast data includes weather type and intensity data. When calling the weather model matching the real-time weather forecast data, first call the single weather model matching the weather type, and then set the number and size of the single weather models according to the intensity data. Multiple single weather models form the weather model.

[0025] The weather forecast data can be obtained from a specific API (Application Programming Interface) after receiving the generation instruction to obtain the latest and relatively authoritative weather forecast data. Generally speaking, the weather forecast data will include various data such as weather type, intensity data, travel suggestions, etc. In the present application, only the weather type and intensity data are retained, and the rest of the data will be discarded. The weather type among them is the specific weather state, such as sunny, rainy, snowy, cloudy, etc., which is one of the data that people are most concerned about, while the intensity data is the intensity of the weather state. For example, the intensities of sunny and cloudy are the same, rainy is divided into light rain, moderate rain, heavy rain, etc., and snowy is divided into light snow, moderate snow, and heavy snow, etc.

[0026] The single weather model is used to reflect the weather changes with the change of intensity data. For example, in rainy weather, the single weather model is raindrops. In light rain weather, the raindrops are not only few in number but also small in volume. In moderate rain weather, the raindrops are appropriately larger in number and size compared to light rain. In the embodiments of the present application, corresponding values are set for each intensity data according to the actual situation. After obtaining the intensity data, setting the number and size of the single weather model according to its corresponding value can achieve the simulation of the weather with different intensity data.

[0027] Further, the real-time industry data includes object type and quantity data. When calling the industry model matching the real-time industry data, first call the object model matching the object type, and then set the quantity of the object model according to the quantity data.

[0028] Taking the transportation industry as an example, after receiving the generation instruction, real-time vehicle flow and type data in the real scene corresponding to the scenario model can be obtained from a specific API. Call the corresponding vehicle model according to the vehicle type, and then set the corresponding number of vehicles on the road according to the vehicle flow, thus completing the call of the industry model.

[0029] It should be understood that since the industry model includes the scenario model and the object model, when calling the industry model, the object model needs to be embedded into the scenario model to form a complete industry model.

[0030] Further, if there is no weather model matching the real-time weather forecast data, a matching model is generated in real time according to the real-time weather forecast data; if there is no industry model matching the real-time industry data, a matching model is generated in real time according to the real-time industry data.

[0031] Specifically, the process of generating the model is as follows: 1. Real-time generation of the weather model.

[0032] When there is no matching model for the real-time weather forecast data, the following process is executed: 1.1: Data parsing. Extract the weather type, such as rainfall, snowfall, strong wind, and intensity data, such as light rain, moderate snow, 10-level wind.

[0033] 1.2: Model library matching check. Retrieve in the pre-stored weather model library whether there is a model with the same weather type and parameter range covering the current intensity data.

[0034] 1.3: Parameterized dynamic generation. If there is a basic model template, such as the "rainfall" general template, adjust the parameters according to the intensity data: Rain model: Adjust the raindrop density (number of raindrops per unit area), size (raindrop radius), and falling speed.

[0035] Snow Model: Adjust the snowflake distribution density, falling trajectory, and accumulation rate.

[0036] Wind Model: Adjust the particle movement trajectories corresponding to wind force levels, such as the swing amplitude of leaves and the intensity of water surface ripples.

[0037] If the basic model template is missing, call the AI generation module, such as a 3D model generator based on GAN (Generative Adversarial Network), and input weather type and intensity data to automatically generate a dynamic model that conforms to physical characteristics.

[0038] 1.4: Model Optimization and Storage. Lightweight optimize the newly generated model, such as reducing the number of polygon faces, and store it in the model library for subsequent calls.

[0039] 2. Real-time Generation of Industry Models.

[0040] When there is no matching model for real-time industry data, execute the following process: 2.1: Data Parsing. Extract industry types (such as transportation, agriculture), object types (such as vehicles, crops), and quantity data.

[0041] 2.2: Model Library Matching Check. Check in the industry model library whether there is a benchmark model for the same type of object, such as the basic model of a "small car".

[0042] 2.3: AI-assisted Modeling. If there is a benchmark model, batch instantiate the model according to the quantity data, such as copying 100 car models, and adjust layout parameters, such as the vehicle spacing on the road.

[0043] If there is no benchmark model, generate it in the following ways: A: Parametric Generation. Input object attributes, such as size, color, function, and generate a simple 3D model through parametric modeling tools.

[0044] B: AIGC Generation: Call a pre-trained 3D generation model, such as Point-E or Shap-E, input a text description, such as "freight truck", output a preliminary model, and then optimize the details through manual or automated tools.

[0045] 2.4: Scene Fusion. Embed the generated object model into a predefined scene model, such as roads, farmlands, and adjust the position (such as vehicle distribution along the road) and posture (such as the tilt angle of crops) according to real-time data.

[0046] 3. Technical Guarantee for Real-time Generation.

[0047] Computing Optimization. Adopt edge computing or cloud computing resources to preferentially process key model generation tasks to ensure low latency.

[0048] Model lightweighting. The LOD (Level of Detail) technology is automatically applied during the generation process to reduce the rendering load.

[0049] Compatibility check. Use a script to check the physical interaction rules between the new model and the studio model, such as collision volume and light reflection, to ensure seamless embedding.

[0050] 4. Example process.

[0051] Taking the generation of the "extremely heavy rain" weather model as an example: Parse data: Weather type = heavy rain, intensity data = 100 mm / h.

[0052] There is no matching model in the model library, so call the rainfall basic template.

[0053] Adjust parameters: Increase the raindrop density to 5000 drops / m³, increase the raindrop radius to 0.5 cm, and add the ground splash particle effect.

[0054] After generation, save it in the model library and mark it as the "heavy rain - 100 mm / h" model for subsequent direct calls.

[0055] Through the above steps, the system can quickly generate a 3D model that conforms to real-time data when there is no pre-stored model, and ensure the dynamics and authenticity of the scenario-based weather forecast.

[0056] S120, Embed the called weather model and industry model into the preset studio model.

[0057] Exemplarily, the studio model is a 3D model with a certain degree of authenticity built according to a real studio. Usually, there is only one performance area in this studio model, and other devices are arranged around the performance area. Therefore, the performance area in the center of the studio model can be set as a blank area. When embedding the weather model and industry model, the industry model can be first fully covered in the blank area of the entire studio model, and then the weather model can be embedded in the industry model with a suitable position and size. At this time, the weather model is located on the outer layer of the industry model. Finally, a digital human model can be set on the outermost layer, and this digital human model can announce the content of the weather forecast through actions and language.

[0058] S130, Simulate the impact of the weather model on the industry model to obtain the scenario-based weather forecast effect.

[0059] Exemplarily, the weather model not only contains visual content, but also contains impact effect data representing weather types and intensity data. For example, rain will wet the road surface and accumulate water droplets on the leaves of crops. The water accumulation on the road surface caused by light rain and moderate rain is different, and the hitting frequency and force of light rain and moderate rain on the leaves of crops are also different. At the same time, the industry model also has a significant impact on the impact effect, which can be determined by the attributes of the object model in the industry model. For example, a vehicle will only be wet under the influence of raindrops and will not shake like the leaves of crops.

[0060] Based on the influence relationship between the weather model and the industry model, a more realistic three-dimensional model can be simulated using the rendering system. After simulating the influence of the weather model on the industry model, the embodiments of the present application also use AIGC technology to generate corresponding videos and texts from the three-dimensional model, and display the videos and texts to achieve the effect of real-time generation of weather forecast graphics and text scenes.

[0061] It should be understood that the content embedded in the blank area of the studio model is the above video, and the text can be displayed at a specific position in the studio model.

[0062] Furthermore, multiple observation points are set in each scene model, and each observation point displays the scene model at different positions and directions. After the user selects an observation point, a video corresponding to the selected observation point is generated.

[0063] The above observation points can be set on the ground or in the air in the scene model to display the influence of the weather model on the object model in the scene model from different perspectives. Specifically, the observation points can be set in the scene model according to the real positions. By default, a specific observation point will be used to generate the video. If the user clicks on other observation points, a video will be re-generated in real time at the clicked observation point. When selecting an observation point, it should be ensured that all other observation points at all positions can be displayed in the video of each observation point for the user to select.

[0064] The embodiments of the present application also provide a device for real-time generation of weather forecast graphics and text scenes, and the device includes the following functional modules: A model establishment module for establishing a weather model and an industry model; A model calling module for obtaining real-time weather forecast data and real-time industry data, calling a weather model matching the real-time weather forecast data, and an industry model matching the real-time industry data; A model embedding module for embedding the called weather model and industry model in a preset studio model; An influence simulation module for simulating the influence of the weather model on the industry model to obtain a scene-based weather forecast effect.

[0065] Although the preferred embodiments of the present application have been described, those skilled in the art can make additional changes and modifications once they learn the basic creative concepts. Therefore, the appended claims are intended to be construed as including the preferred embodiments as well as all changes and modifications falling within the scope of the present application.

[0066] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalent technologies, the present application is also intended to include these modifications and variations.

Claims

1. A method for generating weather forecast images and texts in real time based on scenarios, characterized in that: include: Build weather models and industry models; Acquire real-time weather forecast data and real-time industry data, and call the weather model that matches the real-time weather forecast data and the industry model that matches the real-time industry data; Embedding the called weather model and the industry model in a preset studio model; The impact of the weather model on the industry model is simulated to obtain a scenario-based weather forecast effect.

2. A method for generating weather forecast images and texts in real time according to claim 1, characterized in that: If there is no weather model that matches the real-time weather forecast data, a matching model is generated in real time based on the real-time weather forecast data; if there is no industry model that matches the real-time industry data, a matching model is generated in real time based on the real-time industry data.

3. The method for generating weather forecast images and texts in real time according to claim 1, characterized in that: The real-time weather forecast data includes weather type and intensity data. When calling the weather model that matches the real-time weather forecast data, the individual weather model that matches the weather type is called first, and then the number and size of the individual weather models are set according to the intensity data. Multiple individual weather models constitute the weather model.

4. The method for generating weather forecast images and texts in real time according to claim 1, characterized in that: The industry model includes a scene model and an object model. The scene model is a three-dimensional model established for a specific scene related to the industry in the user's area, and the object model is a three-dimensional model established for objects related to the industry. After calling the industry model, the object model is embedded in the scene model.

5. A method for generating weather forecast images and texts in real time according to claim 4, characterized in that: The real-time industry data includes object type and quantity data. When calling the industry model that matches the real-time industry data, the object model that matches the object type is first called, and then the quantity of the object model is set according to the quantity data.

6. A method for generating weather forecast images and texts in real time according to claim 4, characterized in that: After simulating the impact of the weather model on the industry model, AIGC technology is used to generate corresponding videos and texts, and the video and texts are displayed to achieve the effect of real-time generation of weather forecast graphics and text scenarios.

7. A method for generating weather forecast images and texts in real time according to claim 6, characterized in that: A plurality of observation points are arranged in each of the scene models, and each of the observation points displays the scene model at a different position and direction. After the user selects one of the observation points, the video corresponding to the selected observation point is generated.

8. A device using the method for generating weather forecast images and texts in real time according to any one of claims 1 to 7, characterized in that: include: Model building module, used to build weather models and industry models; A model calling module, used for acquiring real-time weather forecast data and real-time industry data, calling the weather model matching the real-time weather forecast data, and the industry model matching the real-time industry data; A model embedding module, used for embedding the called weather model and the industry model in a preset studio model; The impact simulation module is used to simulate the impact of the weather model on the industry model to obtain a scenario-based weather forecast effect.

Images