Intelligent agricultural risk map visualization method and device, equipment and medium

Abstract

The invention relates to the technical field of artificial intelligence, and discloses an intelligent agricultural risk map visualization method, device, equipment and medium, and the method comprises the steps: extracting a meteorological data set and a remote sensing image set in multi-dimensional recorded data according to the multi-dimensional recorded data of a target agricultural region; performing data space registration on the meteorological data set and the remote sensing image set to obtain a fusion data set; performing data block rendering according to the fused data set to obtain an agricultural map of the target agricultural area; performing agricultural risk prediction on the target agricultural region according to the agricultural map to obtain an agricultural prediction risk; and generating an agricultural insurance map according to the agricultural prediction risk and the agricultural map. Application scenes of the agricultural insurance map are increased, and a more accurate agricultural insurance map is generated.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to a method, device, equipment and medium for visualizing a smart agricultural risk map. Background Art

[0002] As an advanced stage of modern agriculture, smart agriculture is developing rapidly. Its core lies in the use of modern information technology, such as the Internet of Things, big data, cloud computing, and artificial intelligence, to achieve online monitoring, precision operations, and digital management of agricultural production. These technologies make agricultural production more accurate and efficient, while improving the ability to respond to natural environmental risks. WebGL technology can achieve 3D graphics rendering in web browsers, providing a new data visualization method for smart agriculture. Through WebGL, agricultural data can be displayed on risk maps in the form of 3D graphics, making agricultural risk management more intuitive and efficient.

[0003] However, the existing technology mainly integrates weather risk information with the map according to the geographic location information to obtain an agricultural insurance map that integrates weather risk information. There is no fusion of multi-source data to build a three-dimensional agricultural ecological model, which leads to problems in the interactivity, data fusion capability and application scenario limitations of the agricultural insurance map system. Therefore, how to generate a more accurate agricultural insurance map has become an urgent problem to be solved. Summary of the invention

[0004] The present invention provides an artificial intelligence-based smart agricultural risk map visualization method, device, computer equipment and medium to solve the technical problem of poor accuracy of agricultural insurance maps caused by limitations in agricultural insurance map data fusion capabilities and application scenarios.

[0005] In the first aspect, a method for visualizing a smart agricultural risk map is provided, including:

[0006] Acquire multidimensional record data of a target agricultural area, and extract a meteorological data set and a remote sensing image set from the multidimensional record data;

[0007] Performing data spatial registration on the meteorological data set and the remote sensing image set to obtain a fused data set;

[0008] Performing data block rendering according to the fused data set to obtain an agricultural map of the target agricultural area;

[0009] Performing agricultural risk prediction on the target agricultural area according to the agricultural map to obtain agricultural predicted risk;

[0010] An agricultural insurance map of the target agricultural area is generated according to the predicted agricultural risks and the agricultural map.

[0011] In a second aspect, a smart agricultural risk map visualization device is provided, comprising:

[0012] A data extraction module, used to obtain multi-dimensional record data of a target agricultural area, and extract a meteorological data set and a remote sensing image set from the multi-dimensional record data;

[0013] A data space registration module, used for performing data space registration on the meteorological data set and the remote sensing image set to obtain a fused data set;

[0014] A data block rendering module, used for performing data block rendering according to the fused data set to obtain an agricultural map of the target agricultural area;

[0015] An agricultural risk prediction module, used for predicting agricultural risks for the target agricultural area according to the agricultural map to obtain predicted agricultural risks;

[0016] The agricultural insurance map generation module is used to generate an agricultural insurance map of the target agricultural area based on the predicted agricultural risks and the agricultural map.

[0017] In a third aspect, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above-mentioned smart agriculture risk map visualization method when executing the computer program.

[0018] In a fourth aspect, a computer-readable storage medium is provided, which stores a computer program, and when the computer program is executed by a processor, the steps of the above-mentioned smart agriculture risk map visualization method are implemented.

[0019] In the scheme implemented by the above-mentioned smart agricultural risk map visualization method, device, computer equipment and storage medium, the multidimensional record data of the target agricultural area can be received by the client, and the meteorological data set and remote sensing image set in the multidimensional record data can be extracted; the meteorological data set and the remote sensing image set are spatially aligned to obtain a fused data set; the fused data set is rendered in blocks according to the fused data set to obtain an agricultural map of the target agricultural area; agricultural risk prediction is performed on the target agricultural area according to the agricultural map to obtain agricultural predicted risk; and an agricultural insurance map is generated according to the agricultural predicted risk and the agricultural map. The agricultural insurance map is fed back to the client. In the present invention, for the multidimensional record data of the agricultural area, the meteorological data set and the remote sensing image set in the multidimensional record data are extracted, and the meteorological data set and the remote sensing image set can be spatially aligned to obtain a fused data set; and then the fused data set is rendered in blocks to obtain an agricultural map; and agricultural risk prediction is performed according to the agricultural map; the agricultural insurance map is generated according to the agricultural predicted risk and the agricultural map of the agricultural risk prediction, which can realize the visualization of agricultural predicted risk and corresponding risk countermeasures, increase the application scenarios of the agricultural insurance map, and thus generate a more accurate agricultural insurance map. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative labor.

[0021] Figure 1 This is a schematic diagram of an application environment of a smart agricultural risk map visualization method in one embodiment of the present invention;

[0022] Figure 2 It is a flow chart of a method for visualizing a smart agricultural risk map in one embodiment of the present invention;

[0023] Figure 3 It is a schematic diagram of a process for performing data spatial registration on a meteorological data set and a remote sensing image set provided by an embodiment of the present invention;

[0024] Figure 4 A schematic diagram of a process for performing data block rendering provided by an embodiment of the present invention;

[0025] Figure 5 It is a structural schematic diagram of a smart agricultural risk map visualization device in one embodiment of the present invention;

[0026] Figure 6 is a schematic diagram of a structure of a computer device in one embodiment of the present invention;

[0027] Figure 7 It is another structural schematic diagram of a computer device in one embodiment of the present invention. DETAILED DESCRIPTION

[0028] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0029] The smart agricultural risk map visualization method provided by the embodiment of the present invention can be applied in the following aspects: Figure 1 In the application environment, the client communicates with the server through the network. The server can receive the multidimensional record data of the target agricultural area through the client, and extract the meteorological data set and remote sensing image set in the multidimensional record data; perform data space registration on the meteorological data set and the remote sensing image set to obtain a fused data set; perform data block rendering according to the fused data set to obtain an agricultural map of the target agricultural area; perform agricultural risk prediction on the target agricultural area according to the agricultural map to obtain agricultural predicted risks; and then generate an agricultural insurance map according to the agricultural predicted risks and the agricultural map. The agricultural insurance map is fed back to the client. In the present invention, for the multidimensional record data of the agricultural area, the meteorological data set and the remote sensing image set in the multidimensional record data are extracted, and data space registration can be performed on the meteorological data set and the remote sensing image set to obtain a fused data set; then the fused data set is rendered in blocks to obtain an agricultural map; and agricultural risk prediction is performed according to the agricultural map; the agricultural insurance map is generated according to the agricultural predicted risks and the agricultural map of the agricultural risk prediction, which can realize the visualization of agricultural predicted risks and corresponding risk countermeasures, increase the application scenarios of agricultural insurance maps, and thus generate more accurate agricultural insurance maps. The client may be, but is not limited to, various personal computers, laptops, smart phones, tablet computers, and portable wearable devices. The server may be implemented by an independent server or a server cluster consisting of multiple servers. The present invention is described in detail below through specific embodiments.

[0030] See also Figure 2 As shown, Figure 2 A flow chart of a method for visualizing a smart agricultural risk map provided by an embodiment of the present invention includes the following steps:

[0031] S1. Acquire multidimensional record data of a target agricultural area, and extract a meteorological data set and a remote sensing image set from the multidimensional record data.

[0032] In the embodiment of the present invention, the target agricultural area is a geographical area where smart agriculture is carried out and agricultural insurance maps are generated. In the target agricultural area, agricultural production is the main land use mode and has a significant impact on the local economy, society and culture. These areas may include farmland, pastures, orchards, greenhouses, breeding facilities, etc., as well as infrastructure and services directly related to agricultural production, such as irrigation systems, storage facilities, markets and research institutions.

[0033] Among them, multi-dimensional record data is GIS (geographic information system) data in multiple dimensions of the target agricultural area, such as agricultural insurance meteorological data, remote sensing images, planting distribution, historical compensation records and other mainstream GIS data. GIS data can come from a variety of sources, including government agencies, commercial suppliers, scientific research institutions and personal collection. The quality, accuracy and resolution of the data will vary depending on the source and collection method. Therefore, according to different data sources, multi-dimensional record data of the target agricultural area in multiple data dimensions is obtained.

[0034] In the embodiment of the present invention, the meteorological data set includes various observation and analysis data about weather and climate at multiple historical time points and in real time. These data are usually collected by meteorological stations, satellites, radars and other observation equipment and used for weather forecasting, climate research and environmental monitoring. The meteorological data set usually includes temperature data, precipitation data, wind speed and direction data, humidity data, air pressure data, radiation data, climate index and weather phenomenon records. The remote sensing image set is a collection of images of the earth's surface obtained by remote sensing technology. These images can be taken by sensors on satellites, aircraft or drones, including visible light, infrared, radar and other types of images. Through the remote sensing image set, different types of surface cover can be identified and classified, vegetation health can be assessed, vegetation growth and biomass changes can be monitored, and crop growth conditions, pests and diseases and yield estimation can be monitored. The remote sensing image set is a time series image (images taken continuously over a period of time) used to analyze the dynamic changes of surface characteristics in the target agricultural area.

[0035] In the embodiment of the present invention, the step of extracting the meteorological data set and the remote sensing image set from the multidimensional recorded data includes:

[0036] Performing data cleaning and data standardization processing on the multidimensional record data according to the data format of each record data in the multidimensional record data to obtain target data;

[0037] Extracting data features of each of the target data;

[0038] The target multidimensional data is classified according to the data features to obtain a meteorological data set and a remote sensing image set.

[0039] In an embodiment of the present invention, data cleaning and data standardization can be performed on multidimensional record data according to different data formats. For example, numerical record data, text record data, and image record data, etc., the corresponding algorithm can be selected according to the data format to perform data cleaning set standardization. For example, numerical record data can be cleaned by missing value processing and outlier detection, and data can be standardized by Min-Max standardization. Text record data can be cleaned by methods such as noise removal and format unification, and data standardization can be performed using a vocabulary standardization method. Text record data can be cleaned by image processing techniques such as denoising and sharpening, and then data standardization can be performed by pixel value scaling and normalization. Through data cleaning and data standardization, errors, duplications, incomplete or inconsistent information in multidimensional records can be identified and corrected, and the record data can be converted into a unified format or scale for comparison and analysis.

[0040] In an embodiment of the present invention, the data features of each target data can be extracted respectively through a pre-trained convolutional neural network. For example, the convolutional neural network includes an input layer for receiving the target data input, and convolution kernels of multiple convolution scales. Multiple convolution kernels (or filters) are applied to slide on the input data, and each convolution kernel is responsible for extracting a specific feature (such as edge, texture, etc.). The convolution operation will generate a set of feature maps, each feature map corresponds to a convolution kernel, and then the spatial size of the feature map is reduced through the pooling layer, which reduces the amount of calculation and makes the feature detection more robust. Finally, the target data is classified through the output layer, wherein the number of neurons in the output layer is usually equal to the number of categories, and the Softmax activation function is used for multi-category classification, thereby obtaining the meteorological data set and remote sensing image set in the multidimensional recorded data.

[0041] In the embodiment of the present invention, by extracting meteorological data sets and remote sensing image sets from multidimensional recorded data, meteorological, soil, remote sensing and other data of different time series can be integrated, a three-dimensional agricultural ecological model can be constructed, and the accuracy of subsequent agricultural insurance map generation can be improved.

[0042] S2. Performing data space registration on the meteorological dataset and the remote sensing image dataset to obtain a fused dataset.

[0043] In the embodiment of the present invention, since the data sources of the meteorological dataset and the remote sensing image dataset are different, it is necessary to spatially align the data in the meteorological dataset and the remote sensing image dataset so that they can be accurately matched geometrically to obtain a more accurate fused dataset.

[0044] In the embodiment of the present invention, refer to Figure 3As shown, the data space registration of the meteorological data set and the remote sensing image set is performed to obtain a fused data set, including:

[0045] S31, extracting a remote sensing data set according to the remote sensing image set;

[0046] S32, extracting target data points from the remote sensing data set and the meteorological data set;

[0047] S33, performing projection transformation on the target data points to obtain a fused data set.

[0048] In the embodiments of the present invention, remote sensing data refers to characteristic information of the earth's surface captured by remote sensing technology in remote sensing images. Sensors are used to capture electromagnetic radiation or reflection information of the earth's surface from a long distance (such as aircraft, satellites, drones and other platforms). Remote sensing data includes electromagnetic wave information reflected or emitted by surface objects. This information is received and recorded by sensors for identification and analysis of surface objects. Remote sensing data includes, but is not limited to, optical remote sensing data, radar remote sensing data, hyperspectral remote sensing data, infrared remote sensing data, microwave remote sensing data and laser radar (LiDAR) data.

[0049] In the embodiment of the present invention, the step of extracting the remote sensing data set according to the remote sensing image set includes:

[0050] Performing image denoising on each remote sensing image in the remote sensing image set to obtain a denoised remote sensing image;

[0051] Performing geometric correction on the denoised remote sensing image to obtain a corrected image;

[0052] The remote sensing spectral feature of each pixel in the corrected image is extracted to obtain a remote sensing data set.

[0053] Specifically, image denoising is to eliminate the errors of the sensor itself and the influence of external factors such as atmospheric conditions on remote sensing images. Among them, the spike noise can be filtered through Fourier transform, and the strip noise (bad lines) can be weakened by Fourier transform and low-pass filtering to obtain a denoised remote sensing image; geometric correction is to ensure that the image data is aligned with the geographic coordinate system, including image registration, geometric precision correction and orthorectification. Control points and digital elevation model (DEM) data can be used for geometric correction.

[0054] Furthermore, remote sensing spectral features are key information extracted from remote sensing images. They are quantitative descriptions of the electromagnetic radiation characteristics reflected or emitted by objects in different bands. Spectral features are crucial for the identification and classification of objects because different objects have significant differences in reflectivity or emissivity in specific bands.

[0055] In detail, remote sensing spectral features can include spectral reflectance, spectral index and spectral curve. By analyzing the spectral reflectance of each pixel in the image in different bands, different ground objects can be distinguished; spectral indexes include normalized difference vegetation index (NDVI) and water body index (NDWI), etc. These indices are calculated by specific band combinations and are used to characterize certain specific attributes of ground objects, such as vegetation health or water body existence;

[0056] The spectral curve, also known as the spectral signature, is a curve composed of the spectral reflectance of the object in all bands and can be used as the "fingerprint" of the object.

[0057] In detail, various feature extraction algorithms, such as principal component analysis (PCA), minimum noise factorization (MNF), tasseled cap transform, etc., can be used to reduce data dimensions and extract useful features from remote sensing images. At the same time, the remote sensing spectral features of each pixel can be extracted using the constructed deep learning network to obtain the remote sensing data set of each remote sensing image. The remote sensing data set can be used to identify and analyze various natural and man-made features of the surface, providing a basis for subsequent agricultural insurance maps.

[0058] For example, when using deep learning methods to extract remote sensing spectral features for each pixel, deep learning models such as convolutional neural networks (CNNs) can be used to automatically learn complex feature representations from remote sensing data. These features are crucial for tasks such as remote sensing image classification, target detection, and scene understanding.

[0059] For example, in remote sensing data processing, deep learning networks can directly extract useful features from raw pixel-level data through an end-to-end learning process. These features include not only spectral information, but may also include texture, shape, and other spatial information. For example, CNN can automatically learn feature representations from shallow to deep layers through its multi-layer structure, and these features are very useful for distinguishing different types of objects.

[0060] Furthermore, deep learning is particularly useful in hyperspectral remote sensing image processing, because hyperspectral images have rich spectral features, but are also accompanied by problems of high-dimensional data and information redundancy. Through deep learning, feature dimensionality reduction and selection can be effectively performed to extract the most representative and discriminative features, thereby improving the accuracy and efficiency of classification.

[0061] In addition, deep learning models can also combine data from different sensors through data fusion technology, such as combining hyperspectral images with high spatial resolution images or LiDAR data to improve the spatial resolution and classification accuracy of remote sensing images.

[0062] In practical applications, deep learning models such as CNN have been used for various remote sensing tasks, including but not limited to object classification, target detection, scene understanding, and image fusion. Through training, these models can identify and extract key features that help complete the task, and they can perform well even when the training samples are limited. Therefore, the accuracy of remote sensing data extraction can be effectively improved.

[0063] In detail, the remote sensing dataset and the meteorological dataset can be specially matched to obtain matching feature points, and the matching feature points can be combined as target data points. A common coordinate system can be selected to project the target data points into the same system. Pre-built GIS software can be used for projection transformation. For example, in ArcGIS, the "Define Projection" and "Georeference" functions in the "Georeference" toolbar can be used to convert the coordinate system of the data, so that the remote sensing dataset and the meteorological dataset can be accurately matched geometrically to obtain a more accurate fused dataset.

[0064] S3. Rendering the data in blocks according to the fused data set to obtain an agricultural map of the target agricultural area.

[0065] In the embodiment of the present invention, block rendering is for large-scale GIS data, such as image pyramids, vector tiles, etc., to achieve progressive block loading. For example, only necessary core data can be loaded during the first rendering, and detailed data can be dynamically loaded on demand later. Spatial index and prediction algorithms are used to cache data blocks within the user's field of view in advance to reduce secondary loading delays, and to support background asynchronous data loading without blocking the main thread, thereby ensuring the smoothness of user interaction. A hierarchical caching mechanism for rendering resources is established to automatically identify commonly used resources for priority caching. Resources that users may access are intelligently predicted and preloaded during idle time. Progressive streaming loading and precaching technology are used to optimize the data loading process and reduce initial rendering delays.

[0066] In the embodiment of the present invention, refer to Figure 4 As shown, the data is rendered in blocks according to the fused data set to obtain an agricultural map of the target agricultural area, including:

[0067] S41, performing resource splitting on the fused data set to obtain split data sets;

[0068] S42, performing business classification on the split data set to obtain classified business data;

[0069] S43: Calling a tool with a rendering function to render the classified business data to obtain an agricultural map of the target agricultural area.

[0070] In an embodiment of the present invention, resource splitting is to divide a large fused data set (such as a high-resolution remote sensing image) into small blocks or tiles, and load the corresponding tiles according to the user's view and zoom level, pre-render the data on the server side, and provide it as tiles, so that the client only needs to request and load the required tiles instead of processing the original data.

[0071] Specifically, WebGL 2.0 with rendering function can be called according to the extended characteristics of classified business data to offload computing tasks such as spatial analysis and map projection to the GPU for asynchronous execution, and Web Workers can be used to implement multi-threaded parallel rendering. The CPU is responsible for managing scene data and user interaction, and the GPU focuses on high-performance rendering and computing tasks. The dynamic load balancing algorithm schedules rendering tasks according to the real-time load conditions of each working thread to obtain an agricultural map of the target agricultural area.

[0072] Preferably, the cross-browser compatibility of the core rendering engine can be achieved based on WebAssembly technology, and the key rendering and analysis algorithm modules can be compiled into the WebAssembly format to achieve high-performance operation across browsers and devices.

[0073] Among them, WebAssembly (Wasm for short) is a new way of coding that can run in modern web browsers. It is a low-level assembly-like language with a compact binary format that can run with near-native performance and provide compilation targets for languages ​​such as C / C++, C#, and Rust so that they can run on the Web. WebAssembly is designed to coexist with JavaScript, allowing the two to work together. Through WebAssembly's JavaScript API, WebAssembly modules can be loaded into JavaScript applications and functions can be shared between the two. This allows developers to take advantage of the performance and capabilities of WebAssembly and the expressiveness and flexibility of JavaScript in the same application, even if you may not know how to write WebAssembly code. WebAssembly is an open standard of W3C and has received active participation from major browser manufacturers.

[0074] The working principle of WebAssembly includes the compilation process and the execution process. Developers first write source code in C, C++ or other languages, and then use tools such as Emscripten to compile the source code into WebAssembly bytecode (.wasm file). In the browser, the .wasm file can be loaded through the WebAssembly API of JavaScript and compiled into machine code for execution. WebAssembly and JavaScript can interoperate, and WebAssembly modules can call JavaScript functions and vice versa. Developers can use WebAssembly in a variety of ways, including porting C / C++ applications using Emscripten, directly writing or generating WebAssembly code, writing Rust programs and compiling them to WebAssembly, or using tools such as AssemblyScript. These methods provide developers with flexible options for innovation and development in the WebAssembly ecosystem.

[0075] In detail, WebAssembly can be seamlessly integrated with WebGL to provide more powerful GPU acceleration capabilities. It can dynamically adjust rendering details and image quality according to the hardware performance of different devices to ensure a smooth experience even on low-performance devices, and provide WebGL-based mobile rendering optimization strategies, such as touch interaction support and adaptive UI layout. Using progressive rendering technology, core content is loaded first to reduce the initial loading time. For the differences in WebGLAPI implementations of different browsers, unified encapsulation and compatibility processing are adopted. Using the WebGL Extension mechanism, extended functions are dynamically loaded and used according to browser support. A standardized API is provided to facilitate developers to integrate engine functions across platforms.

[0076] In detail, business classification is to use different methods to render according to different types of GIS data. For example, land parcel data uses geoJSON, and growth monitoring data uses WMTS (Web Map Tile Service, Web Map Tile Service Standard) image layer for rendering. Among them, WebGL 2.0 with rendering function can be called to render different categories according to business classification. WebGL 2.0 is a rendering engine for rendering 3D graphics in web browsers. Based on the WebGL rendering engine, it can realize smooth dynamic annotation, pop-up window, legend and other UI elements. Support data-driven dynamic style rendering, such as color, size, transparency, etc. that changes according to attributes, and provide rich chart components, such as line charts, bar charts, heat maps, etc., to realize data visualization. For example, through WebGL 2.0, corresponding data rendering styles can be provided according to different business data, such as administrative division layer is a line layer, and growth monitoring layer is a surface layer with a mask, so that rich rendering effects can be achieved according to needs.

[0077] S4. Predicting agricultural risks for the target agricultural area according to the agricultural map to obtain predicted agricultural risks.

[0078] In an embodiment of the present invention, agricultural risk prediction is to conduct disaster loss assessment, risk analysis, etc. on the target agricultural area, including analyzing meteorological factors that may cause agricultural losses, such as drought, flood, frost, hail, etc., as well as their temporal and spatial distribution, intensity, frequency and other characteristics, and evaluating the sensitivity and vulnerability of agricultural crops or agricultural systems to specific meteorological disasters, including factors such as the growth stage, health status, soil conditions, etc. of crops. It can also include quantifying the impact of disasters on agricultural production, including yield loss, quality decline and economic loss, comprehensively considering disaster-causing factors, vulnerability of disaster-bearing bodies and possible disaster losses, evaluating the overall risk of agricultural meteorological disasters under specific conditions, and formulating corresponding risk management measures based on the results of risk assessment, such as adjusting crop planting structure, improving agricultural technology, establishing disaster warning systems, etc.

[0079] In detail, agricultural maps can be used to visualize meteorological data and remote sensing data in target agricultural areas, and further agricultural risk prediction can be performed through agricultural maps to determine the possible risks in target agricultural areas.

[0080] In the embodiment of the present invention, the agricultural risk prediction of the target agricultural area according to the agricultural map to obtain the agricultural predicted risk includes:

[0081] Extracting multidimensional risk data of the target agricultural area according to the agricultural map;

[0082] Calculating a risk score for the target agricultural area based on the multidimensional risk data;

[0083] The agricultural forecast risk of the target agricultural area is determined based on the risk score.

[0084] In an embodiment of the present invention, the multidimensional risk data may be multidimensional meteorological data and remote sensing data visualized in an agricultural map, for example, meteorological data such as temperature data, precipitation data, wind speed and direction data, humidity data, air pressure data, radiation data, climate index, and weather phenomenon records in different regions in the agricultural map, as well as remote sensing data such as crop growth conditions, pests and diseases, and yield estimates. Risk parameters may be preset to extract risk data of multiple dimensions from the agricultural map to obtain multidimensional risk data.

[0085] In detail, the risk score of the target agricultural area can be calculated by a pre-built target scoring model, wherein the target scoring model can be a prediction model trained based on the Xgboost algorithm and the K-Folds cross validation algorithm method. Specifically, XGBoost (eXtreme GradientBoosting) is an optimization machine learning algorithm based on the gradient boosting decision tree (GBDT), which performs well in dealing with regression, classification and ranking problems. The core idea of ​​XGBoost is to combine multiple weak prediction models (usually decision trees) into a strong prediction model, and improve the performance of the model by minimizing a regularized objective function that includes training loss and model complexity. K-Folds cross validation is a statistical method used to evaluate and compare the performance of machine learning models to ensure that the model has good generalization ability to unknown data. It divides the dataset into K equal-sized subsets (or "folds"), and for each fold (except one), the data of all other folds are used as training sets, and the data of the current fold is used as a validation set. This process is repeated K times, each time a different fold is selected as the validation set, and after each iteration, the performance of the model on the validation set is recorded. This can be indicators such as accuracy, recall, F1 score, mean square error, etc. After all K iterations are completed, the average value of the model performance indicators is calculated to obtain the overall performance estimate of the model, and then the trained scoring model is obtained, which can improve the accuracy of risk score calculation.

[0086] In detail, multiple historical risk data can be obtained, and the obtained historical risk data can be divided into K-1 training sets and 1 test set according to the K-Folds cross-validation algorithm. Based on the Xgboost algorithm, i regression tree training prediction models are established on the K-1 training sets. The prediction model is used to score and predict the size of agricultural prediction risks, where i is a positive integer greater than 1, that is, m regression trees are established on the training set for prediction: each regression tree is a weak learner, and new weak learners are added step by step to try to correct the residuals of all previous weak learners. Therefore, the final m learners added together can be used to predict the final agricultural prediction risk.

[0087] Specifically, a preset risk score range can be set to determine the agricultural forecast risk corresponding to the risk score. For example, if the score of the crop growth stage is 80 points, which is within the preset good growth risk score range, the agricultural forecast risk is good; if the score of the crop growth stage is 50 points, which is within the preset yield reduction risk score range, the agricultural forecast risk is yield reduction; and disaster loss assessment can be performed based on the risk score. Therefore, agricultural risk prediction can be used to conduct disaster loss assessment and risk analysis on the target agricultural areas.

[0088] S5. Generate an agricultural insurance map of the target agricultural area according to the predicted agricultural risks and the agricultural map.

[0089] In the embodiment of the present invention, the agricultural insurance map is a visualized agricultural map that further includes functions such as remote sensing monitoring, disaster loss assessment, and meteorological early warning.

[0090] In detail, the agricultural insurance map supports basic map navigation operations such as panning, zooming, and rotation to ensure that users can browse geographic information smoothly, and realizes multiple interaction methods such as mouse dragging, scroll wheel zooming, and keyboard navigation to meet the habits of different users. It provides a custom map perspective API, and developers can flexibly control the map status according to the application scenario. It supports map status saving and restoration, which is convenient for users to record and return to the perspective position of interest.

[0091] In an embodiment of the present invention, generating the agricultural insurance map of the target agricultural area according to the predicted agricultural risk and the agricultural map includes:

[0092] Calculating risk countermeasures corresponding to the agricultural forecast risk;

[0093] The agricultural map is annotated and rendered according to the agricultural predicted risks and the risk countermeasures corresponding to the agricultural predicted risks to obtain an agricultural insurance map of the target agricultural area.

[0094] Furthermore, risk countermeasures are a series of measures taken to deal with agricultural forecast risks. They can be a detailed emergency plan for possible natural disasters and other emergencies, including pre-disaster prevention, disaster response and post-disaster recovery, improving the stability of the ecosystem and the disaster resistance of the agricultural system, researching and applying agricultural technologies and management measures that adapt to climate change, and reducing the impact of climate change on agricultural production. The preset risk countermeasure set can be matched with the agricultural forecast risk to obtain the risk countermeasures corresponding to each agricultural forecast risk.

[0095] In detail, the above-mentioned WebGL rendering engine can be used to display the agricultural predicted risks and corresponding risk countermeasures that may appear in the agricultural map on the map, realize the visualization of agricultural predicted risks and corresponding risk countermeasures, increase the application scenarios of agricultural insurance maps, and generate more accurate agricultural insurance maps.

[0096] It should be noted that if software tools or components other than those of our company appear in the embodiments of the present application, they are only used for illustrative purposes and do not represent actual use.

[0097] It can be seen that in the above scheme, for the multidimensional record data in the agricultural area, the meteorological data set and the remote sensing image set in the multidimensional record data are extracted, and the meteorological data set and the remote sensing image set can be spatially aligned to obtain a fused data set; then the fused data set is rendered in blocks to obtain an agricultural map; and agricultural risk prediction is performed based on the agricultural map; agricultural insurance maps are generated based on the agricultural predicted risks and agricultural maps of the agricultural risk prediction, which can realize the visualization of agricultural predicted risks and corresponding risk countermeasures, and at the same time, meteorological, soil, remote sensing and other data of different time series can be integrated, so as to construct a three-dimensional agricultural ecological model, increase the application scenarios of agricultural insurance maps, and thus generate more accurate agricultural insurance maps.

[0098] It should be understood that the order of execution of the steps in the above embodiment does not necessarily mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present invention.

[0099] In one embodiment, a smart agricultural risk map visualization device is provided, and the smart agricultural risk map visualization device corresponds to the smart agricultural risk map visualization method in the above embodiment. Figure 5 As shown, the smart agricultural risk map visualization device 100 may include a data extraction module 101, a data space registration module 102, a data block rendering module 103, an agricultural risk prediction module 104 and an agricultural risk map generation module 105. The functional modules are described in detail as follows:

[0100] The data extraction module 101 is used to obtain multi-dimensional record data of the target agricultural area and extract the meteorological data set and remote sensing image set from the multi-dimensional record data;

[0101] The data space registration module 102 is used to perform data space registration on the meteorological data set and the remote sensing image set to obtain a fused data set;

[0102] The data block rendering module 103 is used to perform data block rendering according to the fused data set to obtain an agricultural map of the target agricultural area;

[0103] The agricultural risk prediction module 104 is used to perform agricultural risk prediction on the target agricultural area according to the agricultural map to obtain agricultural predicted risk;

[0104] The agricultural insurance map generating module 105 is used to generate an agricultural insurance map of the target agricultural area according to the predicted agricultural risks and the agricultural map.

[0105] In one embodiment, the data extraction module 101 is specifically used for:

[0106] Performing data cleaning and data standardization processing on the multidimensional record data according to the data format of each record data in the multidimensional record data to obtain target data;

[0107] Extracting data features of each of the target data;

[0108] The target multidimensional data is classified according to the data features to obtain a meteorological data set and a remote sensing image set.

[0109] In one embodiment, the data space registration module 102 is specifically configured to:

[0110] extracting a remote sensing data set according to the remote sensing image set;

[0111] Extracting target data points from the remote sensing data set and the meteorological data set;

[0112] Projection transformation is performed on the target data points to obtain a fused data set.

[0113] In one embodiment, the data space registration module 102 is further configured to:

[0114] Performing image denoising on each remote sensing image in the remote sensing image set to obtain a denoised remote sensing image;

[0115] Performing geometric correction on the denoised remote sensing image to obtain a corrected image;

[0116] The remote sensing spectral feature of each pixel in the corrected image is extracted to obtain a remote sensing data set.

[0117] In one embodiment, the data block rendering module 103 is specifically used to:

[0118] Performing resource splitting on the fused data set to obtain split data sets;

[0119] Performing business classification on the split data set to obtain classified business data;

[0120] A tool with a rendering function is called to render the classified business data to obtain an agricultural map of the target agricultural area.

[0121] In one embodiment, the agricultural risk prediction module 104 is specifically used to:

[0122] Extracting multidimensional risk data of the target agricultural area according to the agricultural map;

[0123] Calculating a risk score for the target agricultural area based on the multidimensional risk data;

[0124] The agricultural forecast risk of the target agricultural area is determined based on the risk score.

[0125] In one embodiment, the agricultural insurance map generation module 105 is specifically used to:

[0126] Calculating risk countermeasures corresponding to the agricultural forecast risk;

[0127] The agricultural map is annotated and rendered according to the agricultural predicted risks and the risk countermeasures corresponding to the agricultural predicted risks to obtain an agricultural insurance map of the target agricultural area.

[0128] The present invention provides a smart agricultural risk map visualization device, which first extracts a meteorological data set and a remote sensing image set from multidimensional record data of an agricultural area, and can perform data spatial registration on the meteorological data set and the remote sensing image set to obtain a fused data set; then, the fused data set is rendered in blocks to obtain an agricultural map; and agricultural risk prediction is performed based on the agricultural map; an agricultural insurance map is generated based on the agricultural predicted risks and the agricultural map of the agricultural risk prediction, so that visualization of agricultural predicted risks and corresponding risk countermeasures can be achieved, and at the same time, meteorological, soil, remote sensing and other data of different time series can be integrated, so as to construct a three-dimensional agricultural ecological model, increase the application scenarios of the agricultural insurance map, and thus generate a more accurate agricultural insurance map.

[0129] For the specific definition of the smart agricultural risk map visualization device, please refer to the definition of the intelligent question and answer method above, which will not be repeated here. Each module in the above-mentioned smart agricultural risk map visualization device can be implemented in whole or in part by software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0130] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 6 As shown. The computer device includes a processor, a memory, a network interface and a database connected via a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile and / or volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external client via a network connection. When the computer program is executed by the processor, it realizes the functions or steps on the service side of a smart agricultural risk map visualization method.

[0131] In one embodiment, a computer device is provided. The computer device may be a client, and its internal structure diagram may be as follows: Figure 7 As shown. The computer device includes a processor, a memory, a network interface, a display screen and an input device connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external server via a network connection. When the computer program is executed by the processor, it implements the functions or steps of a client side of a smart agricultural risk map visualization method

[0132] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the following steps when executing the computer program:

[0133] Acquire multidimensional record data of a target agricultural area, and extract a meteorological data set and a remote sensing image set from the multidimensional record data;

[0134] Performing data spatial registration on the meteorological data set and the remote sensing image set to obtain a fused data set;

[0135] Performing data block rendering according to the fused data set to obtain an agricultural map of the target agricultural area;

[0136] Performing agricultural risk prediction on the target agricultural area according to the agricultural map to obtain agricultural predicted risk;

[0137] An agricultural insurance map of the target agricultural area is generated according to the predicted agricultural risks and the agricultural map.

[0138] In one embodiment, a computer readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

[0139] Acquire multidimensional record data of a target agricultural area, and extract a meteorological data set and a remote sensing image set from the multidimensional record data;

[0140] Performing data spatial registration on the meteorological data set and the remote sensing image set to obtain a fused data set;

[0141] Performing data block rendering according to the fused data set to obtain an agricultural map of the target agricultural area;

[0142] Performing agricultural risk prediction on the target agricultural area according to the agricultural map to obtain agricultural predicted risk;

[0143] An agricultural insurance map of the target agricultural area is generated according to the predicted agricultural risks and the agricultural map.

[0144] It should be noted that the above functions or steps that can be implemented by the computer-readable storage medium or computer device can refer to the relevant descriptions on the server side and the client side in the aforementioned method embodiment. To avoid repetition, they will not be described one by one here.

[0145] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0146] Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0147] The embodiments described above are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. A method for visualizing a smart agricultural risk map, characterized in that: include: Acquire multidimensional record data of a target agricultural area, and extract a meteorological data set and a remote sensing image set from the multidimensional record data; Performing data spatial registration on the meteorological data set and the remote sensing image set to obtain a fused data set; Performing data block rendering according to the fused data set to obtain an agricultural map of the target agricultural area; Performing agricultural risk prediction on the target agricultural area according to the agricultural map to obtain agricultural predicted risk; An agricultural insurance map of the target agricultural area is generated according to the predicted agricultural risks and the agricultural map.

2. The method for visualizing a smart agricultural risk map according to claim 1, characterized in that: The step of extracting the meteorological data set and the remote sensing image set from the multidimensional recorded data comprises: Performing data cleaning and data standardization processing on the multidimensional record data according to the data format of each record data in the multidimensional record data to obtain target data; Extracting data features of each of the target data; The target multidimensional data is classified according to the data features to obtain a meteorological data set and a remote sensing image set.

3. The method for visualizing a smart agricultural risk map according to claim 1, characterized in that: The step of performing data space registration on the meteorological data set and the remote sensing image set to obtain a fused data set includes: extracting a remote sensing data set according to the remote sensing image set; Extracting target data points from the remote sensing data set and the meteorological data set; Projection transformation is performed on the target data points to obtain a fused data set.

4. The method for visualizing a smart agricultural risk map according to claim 3, characterized in that: The step of extracting a remote sensing data set according to the remote sensing image set includes: Performing image denoising on each remote sensing image in the remote sensing image set to obtain a denoised remote sensing image; Performing geometric correction on the denoised remote sensing image to obtain a corrected image; The remote sensing spectral feature of each pixel in the corrected image is extracted to obtain a remote sensing data set.

5. The method for visualizing a smart agricultural risk map according to claim 1, characterized in that: The step of performing data block rendering according to the fused data set to obtain an agricultural map of the target agricultural area includes: Performing resource splitting on the fused data set to obtain split data sets; Performing business classification on the split data set to obtain classified business data; A tool with a rendering function is called to render the classified business data to obtain an agricultural map of the target agricultural area.

6. The method for visualizing a smart agricultural risk map according to any one of claims 1 to 5, characterized in that: The step of performing agricultural risk prediction on the target agricultural area according to the agricultural map to obtain agricultural predicted risk includes: Extracting multidimensional risk data of the target agricultural area according to the agricultural map; Calculating a risk score for the target agricultural area based on the multidimensional risk data; The agricultural forecast risk of the target agricultural area is determined based on the risk score.

7. The method for visualizing a smart agricultural risk map according to any one of claims 1 to 5, characterized in that: The step of generating the agricultural insurance map of the target agricultural area according to the predicted agricultural risk and the agricultural map comprises: Calculating risk countermeasures corresponding to the agricultural forecast risk; The agricultural map is annotated and rendered according to the agricultural predicted risks and the risk countermeasures corresponding to the agricultural predicted risks to obtain an agricultural insurance map of the target agricultural area.

8. A smart agricultural risk map visualization device, characterized in that: include: A data extraction module, used to obtain multi-dimensional record data of a target agricultural area, and extract a meteorological data set and a remote sensing image set from the multi-dimensional record data; A data space registration module, used for performing data space registration on the meteorological data set and the remote sensing image set to obtain a fused data set; A data block rendering module, used for performing data block rendering according to the fused data set to obtain an agricultural map of the target agricultural area; An agricultural risk prediction module, used for predicting agricultural risks for the target agricultural area according to the agricultural map to obtain predicted agricultural risks; The agricultural insurance map generation module is used to generate an agricultural insurance map of the target agricultural area according to the predicted agricultural risks and the agricultural map.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the smart agriculture risk map visualization method as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the smart agriculture risk map visualization method as described in any one of claims 1 to 7 are implemented.

Images