A weather forecast picture-text scene real-time generation method and device

Abstract

The application discloses a weather forecast picture-text scene real-time generation method and device, relates to the three-dimensional modeling technical field, and comprises the following steps: establishing a weather model and an industry model; acquiring real-time weather forecast data and real-time industry data, calling the weather model matched with the real-time weather forecast data and the industry model matched with the real-time industry data; embedding the called weather model and industry model in a preset studio model; simulating the influence of the weather model on the industry model to obtain a weather forecast effect of a scene. According to the real-time meteorological data, the matched model is called, the animation is generated in combination with the industry model, the audience is presented in a direct animation demonstration form, the virtual weather forecast can not only provide accurate weather forecast information, but also can automatically generate a virtual scene, greatly simplifies the operation difficulty, and enhances the program attraction and the viewing experience of the audience.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of three-dimensional modeling, in particular to a weather forecast graphic scene real-time generation method and device. BACKGROUND

[0002] UE (Unreal Engine) technology has been increasingly widely used in 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 scenes, allowing viewers to experience weather changes and their impacts in a first-person perspective.

[0003] The existing related technology has the following three problems:

[0004] 1. It is impossible to obtain weather telegraph messages and other weather information in real time, and automatically call or generate related weather models or components based on existing related information.

[0005] 2. It is impossible to connect weather information with related industries such as agriculture, transportation, and civil aviation, and automatically call or generate related industry models or components based on existing related information.

[0006] 3. The UE+XR+virtual studio technology used in weather forecasts still uses traditional modeling and manual data analysis methods, which have high requirements for various needs in the system (model construction, system shooting, real-time data analysis, etc.), resulting in the fact that it cannot be widely used due to high usage requirements. SUMMARY

[0007] The present application provides a weather forecast graphic scene real-time generation method and device to solve the problem of insufficient support for weather models and industry models in the prior art, and high usage requirements.

[0008] In one aspect, the present application provides a weather forecast graphic scene real-time generation method, which comprises:

[0009] establishing weather models and industry models;

[0010] obtaining real-time weather forecast data and real-time industry data, calling weather models matched with the real-time weather forecast data, and calling industry models matched with the real-time industry data;

[0011] embedding the called weather models and industry models in a preset studio model;

[0012] simulating the influence of the weather models on the industry models to obtain a scene weather forecast effect.

[0013] In a possible implementation, 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.

[0014] In a possible implementation, the real-time weather forecast data includes weather type and intensity data, when the weather model matching the real-time weather forecast data is called, a single weather model matching the weather type is first called, and then the number and size of the single weather model are set according to the intensity data, and a plurality of single weather models constitute the weather model.

[0015] 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 a specific scene related to the industry in the area where the user is located, and the object model is a three-dimensional model established for an object related to the industry, and after the industry model is called, the object model is embedded in the scene model.

[0016] In a possible implementation, the real-time industry data includes object type and quantity data, when the industry model matching the real-time industry data is called, an object model matching the object type is first called, and then the number of the object model is set according to the quantity data.

[0017] In a possible implementation, after the influence of the weather model on the industry model is simulated, corresponding video and text are generated by using AIGC (artificial intelligence generated content) technology, and the video and the text are displayed to achieve the effect of real-time generation of weather forecast scene.

[0018] In a possible implementation, a plurality of observation points are set in each scene model, each observation point displays the scene model from a different position and direction, and after a user selects an observation point, a video corresponding to the selected observation point is generated.

[0019] On the other hand, the embodiment of the present application also provides a weather forecast scene real-time generation device. The device comprises:

[0020] A model establishing module is configured to establish a weather model and an industry model;

[0021] A model calling module is configured to obtain real-time weather forecast data and real-time industry data, call a weather model matching the real-time weather forecast data, and call an industry model matching the real-time industry data;

[0022] A model embedding module is configured to embed the called weather model and the called industry model in a preset studio model;

[0023] An influence simulation module is used to simulate the influence of a weather model on an industry model, to obtain a scenario-based weather forecast effect.

[0024] The weather forecast graphic and text scenario real-time generation method and device provided in the application has the following advantages:

[0025] According to real-time meteorological texts, meteorological data, etc., a correlation data model (rainfall level, storm, cold wave) of a meteorological model is generated in real time, and a related model animation can be automatically generated according to real-time data, to be directly presented to the audience in the form of animation demonstration. If this technology is widely used, the virtual weather forecast system can not only provide accurate weather forecast information, but also can enhance the attraction of the program and the viewing experience of the audience by automatically designing a virtual scene, greatly simplifying the difficulty of use and operation, so that the audience can intuitively feel the impact of the weather, and can combine more use scenarios with AI (artificial intelligence) technology. BRIEF DESCRIPTION OF DRAWINGS

[0026] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or prior art description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can be obtained without creative labor on the basis of these drawings.

[0027] Figure 1 A flowchart of a weather forecast graphic and text scenario real-time generation method provided for an embodiment of the present application. DETAILED DESCRIPTION

[0028] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some embodiments of the present application, not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0029] Figure 1 A flowchart of a weather forecast graphic and text scenario real-time generation method provided for an embodiment of the present application. The present application provides a weather forecast graphic and text scenario real-time generation method, which comprises:

[0030] S100, a weather model and an industry model are established.

[0031] 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, wherein the clouds and raindrops can be displayed through visual effects, and thus a corresponding three-dimensional model can be established through three-dimensional modeling. The wind cannot be visually displayed, and thus can be taken as implicit data, and effects thereof can be displayed through an industry model.

[0032] 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 an industry in a region where a user is located. The object model is a three-dimensional model established for an object related to an industry. Industries in the embodiments of the present application mainly include agriculture, transportation and civil aviation, and thus the scene model related to air traffic control of these industries can be farmland, an agricultural production base, a road, a waterway, an airport, the air, etc. In order to improve the affinity of the user in use, a representative scene in a region where the user is located can be established as a three-dimensional model, which is taken as the scene model. It should be understood that, when the scene model is established, related objects in the scene also need to be embodied, for example, trees and houses around the farmland, buildings on both sides of the road, etc., and the real state of these representative scenes is embodied as much as possible.

[0033] For the above industries, the object model can be crops, fruit trees, vehicles, ships and aircraft, etc. When modeling these objects, several representative objects can be selected, for example, wheat seedlings, apple trees and litchi trees in agriculture, small cars, freight cars, trains and cargo ships in the transportation industry, and passenger aircraft and shuttle buses in civil aviation. After the object model is established, corresponding attribute parameters need to be set according to the real properties of the objects, for example, weight, size, elasticity, etc., so as to simulate the real influence of the weather model on the object model in subsequent calculation, and improve the authenticity of the weather forecast effect.

[0034] In S110, real-time weather forecast data and real-time industry data are obtained, a weather model matched with the real-time weather forecast data is called, and an industry model matched with the real-time industry data is called.

[0035] Exemplarily, the real-time weather forecast data includes weather type and intensity data. When the weather model matched with the real-time weather forecast data is called, a single-body weather model matched with the weather type is called first, and then the number and size of the single-body weather model are set according to the intensity data, and multiple single-body weather models constitute the weather model.

[0036] The weather forecast data can be obtained from a specific API (application programming interface) after receiving the generation instruction to obtain the latest and more authoritative weather forecast data. Generally speaking, the weather forecast data will include weather type, intensity data, travel recommendations and other data. This application only retains the weather type and intensity data, and the rest of the data will be discarded. The weather type is a specific weather state, such as sunny, rainy, snowy, cloudy, etc. It is one of the data that people care most about. The intensity data is the intensity of the weather state, such as sunny and cloudy, which have the same intensity. Rain is divided into light rain, moderate rain, heavy rain, etc. Snow is divided into light snow, moderate snow and heavy snow, etc.

[0037] The single weather model is used to reflect the change of weather with the change of intensity data. For example, in the rainy weather, the single weather model is raindrop. In the light rain, the raindrop is not only small in quantity, but also small in size. In the moderate rain, the raindrop is appropriately increased in quantity and size compared with the light rain. In the embodiment of the present application, the corresponding value of each intensity data is set according to the actual situation. After obtaining the intensity data, the quantity and size of the single weather model are set according to the corresponding value, so that the weather of different intensity data can be simulated.

[0038] Further, the real-time industry data includes object type and quantity data. When calling the industry model matched with the real-time industry data, the object model matched with the object type is called first, and then the quantity of the object model is set according to the quantity data.

[0039] Taking the transportation industry as an example, the real-time vehicle flow and type data in the real scene corresponding to the scene model can be obtained from a specific API after receiving the generation instruction. The corresponding vehicle model is called according to the vehicle type, and then the corresponding number of vehicles is set on the road according to the vehicle flow, so as to complete the calling of the industry model.

[0040] It should be understood that since the industry model includes the scene model and the object model, when the industry model is called, the object model needs to be embedded into the scene model to form a complete industry model.

[0041] Further, if there is no weather model matched with the real-time weather forecast data, the matched model is generated in real time according to the real-time weather forecast data; if there is no industry model matched with the real-time industry data, the matched model is generated in real time according to the real-time industry data.

[0042] Specifically, the process of generating the model is as follows:

[0043] 1. Real-time weather model generation.

[0044] When there is no matched model for the real-time weather forecast data, the following process is performed:

[0045] 1.1: Data Parsing. Extract weather types such as rain, snow, and strong wind, and intensity data such as light rain, moderate snow, and 10-level wind.

[0046] 1.2: Model Library Matching Check. Search the pre-stored weather model library to see if there is a model with the same weather type and parameter range covering the current intensity data.

[0047] 1.3: Parameterized Dynamic Generation. If there is a basic model template, such as a "rain" general template, adjust the parameters according to the intensity data:

[0048] Rain Model: Adjust raindrop density (number of raindrops per unit area), size (raindrop radius), and falling speed.

[0049] Snow Model: Adjust snowflake distribution density, drift trajectory, and accumulation rate.

[0050] Wind Model: Adjust the particle motion trajectory corresponding to the wind level, such as leaf swing amplitude and water surface ripple intensity.

[0051] If the basic model template is missing, call the AI generation module, such as a GAN (Generative Adversarial Network) three-dimensional model generator, input the weather type and intensity data, and automatically generate a dynamic model that meets the physical characteristics.

[0052] 1.4: Model Optimization and Storage. Optimize the newly generated model, such as reducing the number of polygon faces, and store it in the model library for subsequent calls.

[0053] 2, Real-time Industry Model Generation.

[0054] When there is no matching model for real-time industry data, the following process is executed:

[0055] 2.1: Data Parsing. Extract industry type (such as transportation, agriculture), object type (such as vehicle, crop), and quantity data.

[0056] 2.2: Model Library Matching Check. Search the industry model library for a reference model of the same object, such as a "small car" basic model.

[0057] 2.3: AI-assisted modeling. If there is a reference model, instantiate the model in batches according to the quantity data, such as copying 100 car models, and adjust the layout parameters, such as the distance between vehicles on the road.

[0058] If there is no reference model, generate it in the following ways:

[0059] A: Parameterized Generation. Input object attributes such as size, color, and function, and generate a simple three-dimensional model through parameterized modeling tools.

[0060] B: AIGC Generation: Call pre-trained 3D generation models such as Point-E or Shap-E, input text description such as "cargo truck", output preliminary model, and then optimize details through manual or automated tools.

[0061] 2.4: Scene Fusion. Embed the generated object model into a predefined scene model such as a road or farmland, and adjust the position (such as vehicle distribution along the road) and attitude (such as crop inclination angle) according to real-time data.

[0062] 3. Real-time generation technical support.

[0063] Computational optimization. Use edge computing or cloud computing resources to prioritize critical model generation tasks and ensure low latency.

[0064] Model lightening. Automatically apply LOD (level of detail) technology during generation to reduce rendering load.

[0065] Compatibility verification. Check the physical interaction rules of the new model with the studio model, such as collision volume and light reflection, to ensure seamless embedding.

[0066] 4. Example process.

[0067] Take the generation of "heavy rain" weather model as an example:

[0068] Parse data: weather type = heavy rain, intensity data = 100mm / h.

[0069] No matching model in the model library, call the rainfall base template.

[0070] Adjust parameters: raindrop density increased to 5000 drops / m³, raindrop radius increased to 0.5cm, and ground splash particle effect added.

[0071] After generation, store in the model library and mark as "heavy rain-100mm / h" model for subsequent direct call.

[0072] Through the above steps, the system can quickly generate three-dimensional models that meet real-time data when there is no pre-stored model, and ensure the dynamic and authenticity of the scene weather forecast.

[0073] S120, embed the called weather model and industry model in the preset studio model.

[0074] Exemplarily, the studio model is a three-dimensional model with a certain degree of reality established according to a real studio. Generally, such a studio model has only one performance area, and other equipment surrounds 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 the industry model, the industry model can be first laid out in the entire blank area of the studio model, and then the weather model can be embedded in the industry model at a suitable position and size. At this time, the weather model is located outside the industry model. Finally, a digital human model can be set in the outermost layer. The digital human model can broadcast the content of the weather forecast through actions and language.

[0075] In S130, the influence of the weather model on the industry model is simulated to obtain a weather forecast effect of a scene.

[0076] Exemplarily, the weather model not only contains visual content, but also contains influence effect data representing the weather type and intensity data. For example, rain will wet the road surface and accumulate water droplets on the leaves of crops. The amount of water on the road caused by light rain and moderate rain is different, and the frequency and strength of the impact of light rain and moderate rain on the leaves of crops are also different. At the same time, the industry model has a significant influence on the influence 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, but will not sway like the leaves of crops.

[0077] Based on the influence relationship between the weather model and the industry model, a relatively realistic three-dimensional model can be simulated by using a rendering system. After simulating the influence of the weather model on the industry model, the AIGC technology is further used to generate corresponding video and text from the three-dimensional model, so as to achieve the effect of real-time generation of weather forecast picture-text scenes.

[0078] It should be understood that the content embedded in the blank area of the studio model is the above-mentioned video, and the text can be displayed at a specific position in the studio model.

[0079] Further, a plurality of observation points are set in each scene model, each observation point displays the scene model from a different position and direction. After a user selects an observation point, a video corresponding to the selected observation point is generated.

[0080] The observation points can be set on the ground or in the air in the scene model to show the user 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 real positions, and a video is generated by default using one of the specific observation points, and if the user clicks another observation point, the video is re-generated in real time at the clicked observation point. When selecting the observation points, it should be ensured that in the video at each observation point, all other positions of the observation points can be displayed for the user to select.

[0081] The embodiments of the present application also provide a weather forecast graphic scene real-time generation device, which comprises the following functional modules.

[0082] A model establishing module, configured to establish a weather model and an industry model;

[0083] A model calling module, configured to acquire real-time weather forecast data and real-time industry data, call a weather model matched with the real-time weather forecast data, and call an industry model matched with the real-time industry data;

[0084] A model embedding module, configured to embed the called weather model and industry model in a preset studio model;

[0085] An influence simulating module, configured to simulate the influence of the weather model on the industry model to obtain a weather forecast effect of a scene.

[0086] Although the preferred embodiments of the present application have been described, those skilled in the art who are familiar with the basic inventive concept can make further changes and modifications to the embodiments. 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 application.

[0087] Obviously, those skilled in the art can make various modifications and variations 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 equivalent technologies thereof, the present application also intends to include these modifications and variations.

Claims

1. A method for real-time generation of weather forecast images and text, characterized in that, include: Establish weather models and industry models; the industry models include scene models and object models, the scene models are three-dimensional models of specific scenes related to agriculture, transportation and civil aviation in the user's area, and the object models are three-dimensional models of objects related to agriculture, transportation and civil aviation. The system acquires real-time weather forecast data and real-time industry data, and calls the weather model matching the real-time weather forecast data and the industry model matching the real-time industry data. The real-time weather forecast data includes weather type and intensity data. When calling the weather model matching the real-time weather forecast data, it first calls the individual weather model matching the weather type, and then sets the number and size of the individual weather models according to the intensity data. Multiple individual weather models constitute the weather model. The real-time industry data includes object type and quantity data. When calling the industry model matching the real-time industry data, it first calls the object model matching the object type, and then sets the number of object models according to the quantity data. After calling the industry model, the object model is embedded into the scene model. The weather model and the industry model are embedded in the preset studio model; the studio model is a three-dimensional model built according to the real studio. By simulating the impact of the weather model on the industry model through the attributes of the object model in the industry model, a scenario-based weather forecast effect is obtained.

2. The method for real-time generation of weather forecast images and text according to claim 1, characterized in that, If no weather model matches the real-time weather forecast data, a matching model is generated in real time based on the real-time weather forecast data; if no industry model 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 real-time generation of weather forecast images and text according to claim 1, 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, demonstrating that the videos and texts achieve the effect of real-time generation of weather forecast graphics and text.

4. The method for real-time generation of weather forecast images and text according to claim 3, characterized in that, Each scene model has multiple observation points, and each observation point displays the scene model from a different position and direction. After the user selects an observation point, a video corresponding to the selected observation point is generated.

5. An apparatus for real-time generation of weather forecast maps and text according to any one of claims 1-4, characterized in that, include: The model building module is used to build weather models and industry models; The model invocation module is used to obtain real-time weather forecast data and real-time industry data, and to invoke the weather model that matches the real-time weather forecast data and the industry model that matches the real-time industry data. The model embedding module is used to embed the weather model and the industry model called into a preset studio model; The impact simulation module is used to simulate the impact of the weather model on the industry model to obtain scenario-based weather forecast results.

Images