34 results about "Weather map" patented technology

Three-dimensional target detection in complex weather is a key problem in intelligent driving, and extreme weather is easy to interfere with a sensor, so that the detection precision is reduced. The invention provides a three-dimensional target detection method in complex weather based on millimeter wave radar and vision fusion, and designs an image generation module for simulating complex environment images such as rainfall, snowfall and the like aiming at the problem of scarcity of bad weather image data in a public data set. And by adopting a self-distillation mechanism, multi-modal knowledge learned by the teacher model in normal weather is migrated to the student model trained in bad weather, so that the detection performance in a complex environment is improved. Besides, in order to enhance the multi-modal information fusion capability, gating fusion modules are introduced into the teacher model and the student model, and a learnable gating fusion mechanism is utilized to dynamically adjust weight distribution of multi-modal feature fusion. According to the method, the robustness and the detection precision of the model under the complex weather condition are effectively improved.

A system, method, and device for predicting a risk of wildfire are provided. The system includes a satellite imaging device for collecting data covering a target area and a processing server including an indices module for processing the data to generate key vegetation indices, a map generation module configured to generate weather and static maps, an analysis module configured to analyze historical wildfire data to identify past fire locations in the target area, an integration module configured to integrate the received, generated, and analyzed data to obtain a comprehensive dataset for the target area, and a risk prediction module configured to analyze the comprehensive dataset using a risk prediction model trained to predict the risk of wildfire for the targeted area, the risk prediction model including a machine-learning-based pattern detection model for receiving the comprehensive dataset as an input and generate prediction data describing a predicted risk as an output.

The invention discloses an image defogging method and system based on foggy weather depth estimation, and mainly relates to the technical field of foggy weather visual perception. Comprising the following steps: constructing a depth estimation network, training a depth estimation model based on a fogless weather image data set, and outputting a depth map of an input image; constructing a parameter estimation network, and estimating the global atmospheric light and attenuation coefficient of the foggy day image based on an atmospheric scattering model; constructing an optimization model based on a bidirectional feedback strategy to carry out bidirectional feedback optimization; and S3, constructing a quality evaluation system based on the generative adversarial network, and performing quality evaluation on the optimization result image in the step S3. The system has the beneficial effect that the defect that a traditional visual system is insufficient in sensing capability under the foggy weather condition is overcome.

This invention discloses a method and system for intelligent weather image acquisition and recognition, specifically relating to the field of weather image recognition technology. It involves continuous polarization imaging of the sky region while simultaneously acquiring spatial attitude and temporal information; preprocessing the polarization image data and extracting Stokes parameters to obtain standardized sky polarization light field data; then using a theoretical Rayleigh scattering model to analyze the polarization degree distribution pattern at different wavelengths to obtain theoretical sky polarization pattern characteristics; based on the polarization light field data, identifying the spatial distribution of actual clouds, aerosols, and haze in the sky to form actual atmospheric composition distribution characteristics; extracting anomalous polarization regions through difference analysis between theoretical and actual characteristics, and accurately classifying halo polarization anomalies and scattering polarization anomalies based on geometric and spectral characteristics; generating ultralocal weather change early warning signals based on the classification results; thus improving the accuracy and real-time performance of early warnings for ultralocal anomalous weather phenomena.

Data integration and distribution systems. A system includes a graphical user interface (GUI). Weather and market data are collected. A weather symbology including symbol elements linked to segments of the collected weather data and rules for generating weather symbology instructions are stored. The GUI is generated for display on a user device. A weather symbology instruction is determined based on at least one requested symbol element indicated in a weather data request and the rules. A weather forecast dataset is created from among the collected weather data based on the weather symbology instruction. A presentation package including the weather forecast dataset and the collected market data is generated such that the weather forecast dataset is integrated with the collected market data. The presentation package is presented on the GUI and updated concurrent with changes to at least one of the weather data, the market data and user input.

Data integration and distribution systems. A system includes a graphical user interface (GUI). Weather and market data are collected. A weather symbology including symbol elements linked to segments of the collected weather data and rules for generating weather symbology instructions are stored. The GUI is generated for display on a user device. A weather symbology instruction is determined based on at least one requested symbol element indicated in a weather data request and the rules. A weather forecast dataset is created from among the collected weather data based on the weather symbology instruction. A presentation package including the weather forecast dataset and the collected market data is generated such that the weather forecast dataset is integrated with the collected market data. The presentation package is presented on the GUI and updated concurrent with changes to at least one of the weather data, the market data and user input.

The application discloses a kind of based on feature fusion's severe weather target detection method and device, comprising: the target detection model is as detection branch, and the shallow feature map of real severe weather image is extracted, and the detection loss of detection branch is constructed;Increase severe weather processing model as severe weather processing enhancement branch, and the real severe weather image is removed severe weather interference factor processing and obtains image after processing;Shallow feature map and image after processing are respectively through mixed random pooling and obtain the feature after pooling, the feature of two features after pooling is fused to obtain fusion feature;The similarity loss of the feature after pooling corresponding to fusion feature and shallow feature map is calculated, and the parameter optimization of target detection model is carried out according to similarity loss and detection loss;The target detection model after parameter optimization is used to detect the severe weather image to be detected, so as to improve the precision of target detection in severe weather scene with lower calculation consumption.

The invention is suitable for the technical field of meteorological data processing, and provides a height field and wind field mutual smoothing method based on physical rule constraints, which comprises the following steps: preprocessing a two-dimensional height field lattice point matrix hgtgrid, a two-dimensional wind field horizontal component lattice point matrix ugrid and a two-dimensional wind field vertical component lattice point matrix vgrid to obtain a wind speed and a wind direction, and performing mutual smoothing processing based on a bidirectional coupling relationship between the height field and the wind field, and outputting the height field and the wind field which are subjected to mutual smoothing when the number of cycles reaches a preset maximum number of iterations maxter. Therefore, by establishing a bidirectional dynamic coupling mechanism of the wind field and the height field, visual drawing of the contour line of the weather map is supported, and the method has the advantages of being high in physical consistency and overall coordination.

A system, method, and device for predicting a risk of wildfire are provided. The system includes a satellite imaging device for collecting data covering a target area and a processing server including an indices module for processing the data to generate key vegetation indices, a map generation module configured to generate weather and static maps, an analysis module configured to analyze historical wildfire data to identify past fire locations in the target area, an integration module configured to integrate the received, generated, and analyzed data to obtain a comprehensive dataset for the target area, and a risk prediction module configured to analyze the comprehensive dataset using a risk prediction model trained to predict the risk of wildfire for the targeted area, the risk prediction model including a machine-learning-based pattern detection model for receiving the comprehensive dataset as an input and generate prediction data describing a predicted risk as an output.

The application discloses a weather broadcast and early warning method and system based on an intelligent door lock and an intelligent door lock, and the method comprises the following steps: acquiring current position information of the intelligent door lock, and acquiring local real-time weather data from a weather service provider according to the current position information; analyzing and processing the local real-time weather data to generate voice broadcast content, weather text content and weather icon content; setting multiple weather reminding scenes according to the voice broadcast content, the weather text content and the weather icon content, and triggering the corresponding weather reminding scenes to perform weather reminding according to a user's interactive scene selection instruction; and if the local real-time weather data belongs to a range of severe weather, then when it is detected that the user goes out, severe weather voice broadcast and early warning information voice broadcast are performed. The application sets multiple weather reminding modes on the intelligent door lock, timely reminds the user of the weather, especially the voice broadcast reminding of severe weather, and greatly improves user experience and life convenience.

The application discloses a short-term photovoltaic power generation capacity prediction method and device, comprising the following steps: constructing an attention-based hybrid model composed of a CNN-GRU-SE component for processing weather image data and an IPSO-GRU component based on IPSO optimization for processing meteorological data; training and testing the attention-based hybrid model by using a training data set; acquiring meteorological data and weather image data of the photovoltaic power plant at the current time, and predicting the short-term power generation capacity of the photovoltaic power plant in the next period of time by using the trained attention-based hybrid model. In the application, the spatial feature extraction capability of the convolutional neural network and the time sequence feature prediction capability of the GRU are fused, so that the spatial and time patterns can be effectively learned from the input data, and the performance is improved compared with a single model.

Example systems, methods, and non-transitory computer-readable media are directed to determining geographic sites for which lightning activity is to be evaluated over a period of time; obtaining weather information associated with the geographic site, the weather information being generated by a weather prediction model for the time period, where the weather information comprises a grid-based representation of the weather map; determining one or more grid points to assess lightning activity of the geographic site; evaluating a weather-related metric associated with the one or more grid points; determining one or more likelihood that the geographic site suffers from lightning activity during the time period based at least in part on the evaluation; and providing weather forecast information of the geographic site, the weather forecast information including at least one or more likelihood that the geographic site suffers from lightning activity during the time period.

The invention discloses a short-term photovoltaic generating capacity prediction method. The method comprises the following steps: constructing an attention-based hybrid model consisting of a CNN-GRU-SE component for processing weather image data and an IPSO-GRU component for processing meteorological data and based on IPSO optimization; training and testing the attention-based hybrid model by adopting a training data set; and acquiring meteorological data and weather image data of the photovoltaic power plant at the current moment, and predicting the short-term power generation power of the photovoltaic power plant in the next period of time through the trained attention-based hybrid model. According to the method, the spatial feature extraction capability of the convolutional neural network and the time sequence feature prediction capability of the GRU are fused, so that space and time modes can be effectively learned from input data, and the performance is improved compared with that of a single model.

A system, method, and device for predicting a risk of wildfire are provided. The system includes a satellite imaging device for collecting data covering a target area and a processing server including an indices module for processing the data to generate key vegetation indices, a map generation module configured to generate weather and static maps, an analysis module configured to analyze historical wildfire data to identify past fire locations in the target area, an integration module configured to integrate the received, generated, and analyzed data to obtain a comprehensive dataset for the target area, and a risk prediction module configured to analyze the comprehensive dataset using a risk prediction model trained to predict the risk of wildfire for the targeted area, the risk prediction model including a machine-learning-based pattern detection model for receiving the comprehensive dataset as an input and generate prediction data describing a predicted risk as an output.

A system is provided.SOLUTION: A sensor that measures a water level of a river, a server that receives data from the sensor and determines whether the water level has reached a dangerous water area, means for the server to predict a damage assumed area using weather map data and water level data, and means for the server to generate a flight path of a drone based on the predicted damage assumed area, A system comprising: means for automatically creating a flight plan; means for causing the drone to fly, photographing a damage situation from the sky, and transmitting the data to the server; means for causing the drone to perform evacuation guidance by voice in a damage assumed area; and means for causing the server to transfer the received data to a terminal of a local government staff in real time.SELECTED DRAWING: Figure 1

The invention relates to the technical field of image generation, in particular to a severe weather image generation method based on a generative diffusion model, and the method comprises the steps: obtaining the multi-scale features of an original image through ViT; carrying out statistics on the real severe weather image set to obtain prompt basic information; performing instance sampling, scene construction processing and scene description processing on the prompt basic information to obtain scene information, and performing scene description processing on the scene information by adopting a large language model to generate a text prompt; converting the text prompt into text embedding, and enabling the text embedding to be consistent with the multi-scale feature dimension; performing multi-step diffusion on the text embedding and multi-scale features by adopting a diffusion model, updating a text prompt and updating the corresponding text embedding by adopting the LLM module in each step of diffusion process, and finally obtaining a severe weather image; according to the method, the controllability and the accuracy of the diffusion model generation process can be improved.

The invention provides an interactive weather mapping rendering method and device based on a grid and RTree, and relates to the field of rendering, and the method comprises the steps: building an RTree longitude and latitude spatial index of each meteorological station based on meteorological observation data; establishing a multi-level slice grid system through Web Mercator projection based on the target meteorological observation data; in response to an interaction operation for the current map, performing spatial range query on the current map according to the RTree longitude and latitude spatial index, and determining a to-be-displayed target grid slice from the plurality of grid slices according to a spatial range query result; and performing hierarchical rendering on the target grid slice, and performing icon collision detection aiming at the existence condition of a spatial intersection between a meteorological icon of a meteorological station to be drawn and a drawn icon in the target grid slice through the RTree pixel spatial index of the target grid slice in the rendering process. And generating a meteorological icon corresponding to the meteorological observation data according to the collision detection result.

This invention discloses an intelligent weather phenomenon recognition method based on AI recognition algorithms, belonging to the fields of image processing and meteorological monitoring technology. The method includes: acquiring weather image data through edge devices; separating the foreground and background of the weather image data to obtain the target area; determining whether the target area triggers a weather rule based on a preset weather rule scenario; when a weather rule is triggered, inputting the target area into a pre-trained AI recognition model for recognition and analysis to obtain the weather recognition result. This invention can achieve accurate recognition of various weather phenomena such as snowfall and frost, improving the accuracy and consistency of weather phenomenon recognition, and overcoming the problems of strong subjectivity and inconsistent judgment standards in manual observation methods.

Preemptively charging rechargeable energy storage systems (RESSs) in advance of an electrical grid outage. The preemptive charging may include generating a weather map of a geographical area anticipated to be impacted with a storm or a natural event of a severity predicted to induce a power outage, which may include generating a distribution map of an electrical power distribution network within the geographical area, generating an outage map by overlaying the weather map relative to the distribution map to identify one or more intersections where the power outage is predicted, and scheduling or otherwise implementing preemptively charging of one or more the RESSs downstream of the intersections in advance of the power outage.

The invention discloses a multi-modal flight prediction method based on an instrument panel. The method comprises the following steps: acquiring an aircraft instrument panel image and a weather image through image acquisition equipment; performing parameter identification on the instrument panel image to obtain instrument panel parameters and weather type results; the identified parameters and the weather type result are integrated according to a timestamp, and a file of time sequence flight data is generated; and constructing a multi-modal fusion model based on a Transform encoder, an LSTM network and an attention mechanism to analyze the time sequence flight data, and outputting a flight state prediction result. Therefore, on the basis of the method, flight parameters and state classification (rising / falling / deceleration / keeping) of the five steps in the future are synchronously output through the multi-task prediction head finally, and end-to-end accurate mapping from multi-dimensional time sequence data to flight evolution is achieved. The prediction result of the embodiment of the invention is obviously superior to the results of other prediction methods in the prior art.

1. Name of the product in this design: Voice-activated graphical user interface for weather query on electronic devices. 2. Purpose of this design: An electronic device. 3. The key design feature of this product is its graphical user interface. 4. The picture or photo that best illustrates the key design points: Design 1 front view. 5. Design 1 is designated as the basic design. 6. Purpose of the graphical user interface: For querying weather via voice. 7. Human-computer interaction method of graphical user interface: Ask questions by voice and display the answer information through the interface pop-up window. Design 1 and Design 2's main view shows the day's weather query via voice, with a pop-up window on the left side of the interface displaying the day's weather information. Design 3's main view is an interface for querying weather indices via voice, with a pop-up window on the left side of the interface displaying weather index information. Design 4's main view is an interface for querying the air quality index for multiple days via voice, with an air quality index information displayed in a pop-up window on the left side of the interface. Design 5's main view is an interface for querying the weather conditions of multiple locations on the same day via voice. A pop-up window on the left side of the interface displays the weather information for multiple locations on the same day. Design 6's main view is an interface for querying the weather for a certain location for the next week via voice. A pop-up window on the left side of the interface displays the weather information for that location for the next week. 8. Other situations requiring explanation: The red semi-transparent coating area is not fixed content and is not the part for which patent protection is sought; the design for which protection is sought does not include color.

The invention discloses a weather forecast visualization method and device based on multi-mode comparison and linkage analysis, and the method comprises the steps: determining a plurality of weather forecast modes which are selected by a user and need to be compared in response to a selection instruction, and reading the forecast data of each weather forecast mode; according to the forecast data, generating and displaying a multi-mode comparison view; performing linkage analysis on the multi-mode comparison view to obtain a meteorological chart linkage group needing linkage; in response to an operation instruction of a user for a target weather chart in the weather chart linkage group, a current operation event is captured, and the operation event carries view state change parameters; and carrying out weather forecast visualization on the weather chart linkage group according to the view state change parameters to obtain a displayed weather forecast visualization result. According to the invention, the working efficiency is improved, and a predictor can carry out weather forecast analysis and adjustment more efficiently. Meanwhile, the risk of inaccurate analysis results caused by manual misoperation is reduced, and the reliability of the analysis results is improved.

The invention provides a severe weather image restoration method and device and a medium, and relates to the technical field of computer vision. The objective of the invention is to solve the problem that real-time deployment is difficult due to poor generalization ability of an existing image restoration model to changeable weather and high model complexity. The method comprises the following steps: based on a clear image, generating a severe weather image by adopting a parameterized simulation scheme based on a physical model so as to form paired training data; using the paired training data to train a generative adversarial network image restoration model for restoring the severe weather image into a clear image; and implementing a structured pruning technology on the image restoration model after training convergence so as to reduce the parameter quantity and the calculation quantity of the image restoration model. The problem that training data is difficult to obtain is solved by simulating generated data, and the generalization ability of the model is improved; the high efficiency of the model is realized through model pruning, the real-time deployment requirement is met, and the performance of downstream tasks can be enhanced.

The invention provides an image generation method and device, electronic equipment, a storage medium and a product. The method comprises the following steps: acquiring a weather state needing to be displayed from input data of a user; determining a plurality of weather elements included in the weather state and a display parameter corresponding to each weather element; generating a display image corresponding to each weather element based on the display parameter corresponding to each weather element; according to the incidence relation among the multiple display images, the weather image corresponding to the weather state is generated, the display parameter setting process of the weather state needing to be displayed is simplified, and the parameter setting efficiency is improved.

The invention discloses an intelligent vehicle window control method and system based on multi-source data fusion, relates to the technical field of vehicle window intelligentization, and dynamically evaluates the weather condition of the position where a vehicle is located and the air quality in the vehicle by integrating multi-source data such as a satellite weather chart, vehicle longitude and latitude coordinates, a millimeter wave radar and an infrared optical sensor. Therefore, intelligent vehicle window regulation and control are realized; the system calculates weather state indexes according to real-time meteorological elements, and triggers a window closing instruction when encountering extreme weather or foreign matter blocking, so that the driving safety is guaranteed; after the vehicle window is completely closed, continuously monitoring and calculating an air quality state evaluation value in the vehicle, and if the air quality state evaluation value deviates from a preset interval, automatically deciding or transmitting abnormal information to passengers in combination with external weather conditions to form an environment closed-loop regulation mechanism so as to ensure intelligent balance of comfort and safety of the environment in the vehicle.

A method is provided for supporting flight operations of vertical takeoff and landing (VTOL) aircraft in an urban air mobility (UAM) environment that includes a vertiport. The method includes accessing first observed weather data obtained by sensors located on-site at the vertiport and second observed weather data obtained by sensors onboard VTOL aircraft. A data fusion is performed to integrate the first observed weather data and the second observed weather data to produce a collection of weather data that describes the weather in an environment of the vertiport, and an environment of the VTOL aircraft. And a weather map is produced based on the collection of weather data, the weather map illustrating one or more meteorological features of the weather in the environment of the vertiport and / or the VTOL aircraft, the weather map produced for use in supporting flight operations of the VTOL aircraft.

The invention discloses an intelligent weather image acquisition and weather image recognition method and system, and particularly relates to the technical field of weather image recognition. The method comprises the following steps: performing continuous polarization imaging on a sky area, and synchronously acquiring space attitude and time information; carrying out preprocessing and Stokes parameter extraction on the polarized image data to obtain standardized sky polarized light field data; analyzing polarization degree distribution modes under different wavelengths by using a theoretical Rayleigh scattering model to obtain theoretical sky polarization mode characteristics; based on the polarized light field data, spatial distribution of actual clouds, aerosols and haze in the sky is identified, and actual atmospheric component distribution characteristics are formed; an abnormal polarization region is extracted through difference analysis of theoretical and actual characteristics, and halation polarization abnormity and scattering polarization abnormity are accurately classified based on geometric and spectral characteristics; generating a super local weather change early warning signal according to the classification result; the early warning accuracy and real-time performance of the super-local abnormal weather phenomenon are improved.

The invention discloses a short-term photovoltaic generating capacity prediction method. The method comprises the following steps: constructing an attention-based hybrid model consisting of a CNN-GRU-SE component for processing weather image data and an IPSO-GRU component for processing meteorological data and based on IPSO optimization; training and testing the attention-based hybrid model by adopting a training data set; and acquiring meteorological data and weather image data of the photovoltaic power plant at the current moment, and predicting the short-term power generation power of the photovoltaic power plant in the next period of time through the trained attention-based hybrid model. According to the method, the spatial feature extraction capability of the convolutional neural network and the time sequence feature prediction capability of the GRU are fused, so that space and time modes can be effectively learned from input data, and the performance is improved compared with that of a single model.

The invention relates to the technical field of image processing, in particular to an image restoration method in a multi-weather scene, which comprises the following steps: S1, acquiring a real or synthetic weather picture set including an input picture and a label picture; s2, normalizing the input picture, adjusting the pixel mean value and variance of the input picture to 0.5, and forming an image set with the tag picture; s3, constructing a multi-weather removal model, and training the model through the training image set to obtain a trained multi-weather removal model; s4, acquiring a weather picture needing to be repaired, and normalizing the weather picture; and S5, abnormal weather removal is carried out on the weather picture to be recovered in the step S4 through the trained multi-weather removal model, and a reconstructed and recovered picture is obtained. According to the method, feature extraction can be enhanced, the performance overhead of attention score calculation can be reduced, the most critical information learning can be adaptively selected, and excellent image restoration capability can be realized while high performance is maintained.

An object of a system according to an embodiment is to automate monitoring of a water level of a river to evacuation guidance and to enable quick and accurate response.SOLUTION: A system according to an embodiment includes a sensor unit, an alarm unit, an analysis unit, a route creation unit, a control unit, a sound broadcasting unit, and an image transfer unit. The sensor unit monitors a water level of a river. The alarm unit issues an alarm based on the data collected by the sensor unit. The analysis unit analyzes the weather map and predicts a damage assumed area. The route creation unit creates a flight route of the drone based on the damage assumed area predicted by the analysis unit. The control unit causes the drone to fly automatically based on the flight route created by the route creation unit. In the voice broadcasting unit, the drone flying by the control unit performs evacuation guidance to residents in the damage assumed area by voice. The image transfer unit transfers an image captured by the drone flying by the control unit to the local government staff.SELECTED DRAWING: Figure 1