Crop growth management method and system based on sky-air-ground integration

Abstract

The invention relates to the technical field of crop growth management, and discloses a crop growth management method and system based on sky-air-ground integration, and the method comprises the following steps: collecting preliminary monitoring data based on a target detection instruction; performing data cleaning and formatting on the preliminary monitoring data to obtain a standard data format, and performing space-time alignment on the preliminary monitoring data in the standard data format to obtain stage detection data; constructing a crop growth model, and performing fusion processing on the stage detection data based on the crop growth model to obtain crop growth variables; acquiring a crop growth strategy matched with the crop growth variable based on the crop growth variable; the target detection instruction is adjusted based on the crop growth strategy, preliminary monitoring data are collected again, precise irrigation, precise fertilization and precise pest control are achieved, resource waste is reduced, environmental pollution is reduced, and the sustainability of agricultural production is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of agricultural monitoring. More specifically, the present invention relates to a crop growth management method and system based on the integration of space-air-ground Background Art

[0002] The growth status of crops at different growth stages has a direct impact on the final yield. Traditional monitoring methods are difficult to comprehensively and timely capture the growth changes of crops at each growth stage. Especially in large-scale farmlands, the limitations of manual monitoring make it difficult to dynamically manage crop growth. Moreover, crop yield is affected by multiple factors and is difficult to accurately reflect the true growth status of the current-season crops and the future yield potential. Especially in the face of abnormal climate or environmental changes, the prediction results are often not accurate enough.

[0003] For example, the application publication number is CN118211730A, which discloses a method for predicting crop yield based on meteorological feature matching and crop growth models. By constructing a simulation method for unknown meteorological data combining meteorological data matching and yield abundance index, and combining crop growth models to predict regional yield, accurate predictions can be made for different crop yields, and the overall supply and demand situation can be judged.

[0004] However, during the process of crop planting, the growth environment of crops is variable, and it is relatively difficult to predict the yield. Among them: Although satellite remote sensing has a wide coverage range, its resolution is relatively low, making it difficult to capture subtle crop changes. While drones have high resolution, their coverage area is limited and they are greatly affected by weather conditions. Although ground sensors provide accurate micro-environment data, their deployment range is limited and it is difficult to reflect the overall situation of large-scale farmlands. None of these single data sources can provide comprehensive crop growth variables, resulting in insufficient decision-making basis in farmland management and making it difficult to accurately reflect the true situation of farmland.

[0005] When integrating the data of satellites, drones, and ground sensors, the correlation and complementarity between the data cannot be mined through simple superposition analysis, and the advantages of various types of data cannot be fully utilized, resulting in resource waste and low management efficiency. In the face of sudden climate change or pests and diseases, single data sources or delayed data analysis cannot provide effective response strategies in a timely manner, increasing the risk of crop losses and affecting the final yield and quality.

[0006] In view of this, the present invention proposes a crop growth management method and system based on the integration of space-air-ground. Summary of the Invention

[0007] In order to overcome the deficiencies of the prior art, the present invention provides a crop growth management method and system based on the integration of space-air-ground, which has the advantage of more precise operation.

[0008] In a first aspect, the present invention provides a crop growth management method based on space-air-ground integration, comprising the following steps:

[0009] Adjust the frequency of data acquisition nodes within a preset period based on the target detection instruction, and record the timestamps and preliminary monitoring data corresponding to the data acquisition nodes;

[0010] Perform data cleaning and formatting on the preliminary monitoring data to obtain a standard data format, and perform spatio-temporal alignment on the preliminary monitoring data in the standard data format within a preset period to obtain stage detection data;

[0011] Construct a crop growth model, and perform fusion processing on the stage detection data based on the crop growth model to obtain crop growth variables of the crop within a preset period;

[0012] Obtain the yield prediction probability distribution based on the crop growth variables and generate a list of risk sources corresponding to the current probability distribution, and match the corresponding crop growth strategies based on the list of risk sources; adjust the target detection instruction through the crop growth strategy, adjust the frequency of data acquisition nodes within a preset period, and improve and update the preliminary monitoring data.

[0013] As a preferred technical solution of the first aspect of the present invention, the target detection instruction is an operation instruction issued based on different application scenarios during the space-air-ground integration operation process, so as to obtain different preliminary monitoring data, and the preliminary monitoring data includes space-based remote sensing data, air-based detection data, and ground-based planting data;

[0014] The space-based remote sensing data is remote sensing data regularly collected by a satellite sensor according to the satellite orbit and the geographical location of the farmland;

[0015] The air-based detection data is to use an unmanned aerial vehicle equipped with an imaging device to obtain local image data of the farmland;

[0016] The ground-based planting data is crop growth data collected in real time according to a ground monitoring network, and the ground monitoring network includes soil humidity sensors, temperature sensors, nutrient sensors, and photosynthetically active radiation sensors arranged in the target planting area to monitor the planting environment data of soil humidity, temperature, nutrient content, and meteorological conditions in real time.

[0017] As a preferred technical solution of the first aspect of the present invention, the crops are set in a process flow, different growth periods are configured for different growth stages, and marked as preset periods; within each preset period, the data acquisition nodes are marked by the acquisition timestamps of the space-based remote sensing data, and the preliminary monitoring data obtained between adjacent data acquisition nodes is stored in the same time block, and the current time block is marked by the previous acquisition time node.

[0018] As a preferred technical solution of the first aspect of the present invention, the acquisition logic of the stage detection data is as follows:

[0019] Taking the target planting area identified in the space-based remote sensing data as a reference object, a space planting platform is constructed;

[0020] Mark the target planting characteristics identified in the ground-based planting data and match them proportionally to the space planting platform through the target planting area, so that the geographical space is consistent;

[0021] Overlap the contour line of the farmland local image data identified in the space-based detection data with the target planting area, and add local image characteristics within the overlapping target planting area;

[0022] Stack the target planting characteristics and local image characteristics on the space planting platform in sequence according to the time series, and the space planting platform superimposes and marks the detection marker icons collected within the target planting area based on the time stamp to mark the stage detection data of the current time block.

[0023] As a preferred technical solution of the first aspect of the present invention, the superimposing logic of the space planting platform is as follows:

[0024] The space planting platform arbitrarily selects a reference object from the current space-based remote sensing data as a reference point, takes the reference point as the central coordinate, and constructs a space coordinate system based on the reference point;

[0025] In the space coordinate system, classify and mark the reference object, and mark the area contour corresponding to the crop planting environment as the target planting area;

[0026] Completely overlap the area contours of the corresponding target planting areas in the space-based detection data and / or ground-based planting data with the area contours in the space-based remote sensing data, and determine the three-dimensional coordinate system of the target planting area based on the target planting area corresponding to the ground-based planting data;

[0027] Extract the detection marker icons marked at different coordinate positions in the space-based detection data and / or ground-based planting data, and use the detection marker icons to represent the target planting characteristics and local image characteristics;

[0028] Superimpose and display all the detection marker icons on the three-dimensional coordinate system of the target planting area, and determine the azimuth information of the detection marker icons in the target planting area according to the central coordinate.

[0029] As a preferred technical solution of the first aspect of the present invention, the acquisition logic of the crop growth variables is as follows:

[0030] Perform stage slicing analysis on the stage detection data to obtain stage analysis data. The stage analysis data includes target feature data containing various detection markers and icons. The target feature data includes the reference object in the target planting area, the target planting features, and the local image features;

[0031] Perform semantic alignment on the target feature data. For each target feature data, set a corresponding stacked autoencoder through a multi-modal autoencoder network, and convert the target feature data into target encoded features recognizable by the crop growth model;

[0032] Use the target encoded features as the input factors of the crop growth model. The output factor of the crop growth model is the growth anomaly probability;

[0033] According to the timestamp, reconstruct and label the target encoded features and the growth anomaly probability as the growth trajectory matrix. Perform growth anomaly statistics on the growth trajectory matrix, combine the pre-set reconstruction error threshold corresponding to each target encoded feature to determine the dimensionality reduction dimension, obtain the low-dimensional embedding of the current target encoded features, and splice the low-dimensional embedding representations to obtain the crop growth variables.

[0034] As a preferred technical solution of the first aspect of the present invention, search for a list of risk sources that match in the prior knowledge base based on the crop growth strategy. Based on prior knowledge, analyze the yield impact probability corresponding to each crop growth variable in the risk source list, and based on the analysis threshold of the yield impact probability in prior knowledge, determine the target detection instruction.

[0035] In the second aspect, the present invention provides a crop growth management system based on the integration of space-air-ground. Based on the implementation of the crop growth management method based on the integration of space-air-ground described in the first aspect, it includes a data acquisition module, a data processing module, a crop growth analysis module, and a decision feedback module. Each module is connected by wire or wireless;

[0036] The data acquisition module adjusts the frequency of the data acquisition nodes within the preset period based on the target detection instruction, records the timestamp corresponding to the data acquisition nodes and the preliminary monitoring data, and sends the preliminary monitoring data to the data processing module;

[0037] The data processing module performs data cleaning and formatting on the preliminary monitoring data to obtain the standard data format, performs spatio-temporal alignment on the preliminary monitoring data in the standard data format within the preset period to obtain the stage detection data; sends the stage detection data to the crop growth analysis module;

[0038] The crop growth analysis module constructs a crop growth model, performs fusion processing on the stage detection data based on the crop growth model, and obtains the crop growth variables of the crop within the preset period; sends the crop growth variables to the crop decision feedback module;

[0039] The decision feedback module obtains the yield prediction probability distribution based on crop growth variables, generates a list of risk sources corresponding to the current probability distribution, and matches the corresponding crop growth strategies based on the list of risk sources; adjusts the target detection instructions through the crop growth strategies, adjusts the frequency of data collection nodes within a preset period, and improves and updates the preliminary monitoring data.

[0040] In a third aspect, the present invention provides an electronic device, including: a processor and a memory, wherein a computer program callable by the processor is stored in the memory;

[0041] The processor executes the crop growth management method based on the integration of space-air-ground by calling the computer program stored in the memory in the first aspect.

[0042] In a fourth aspect, the present invention provides a computer-readable storage medium storing instructions, which when run on a computer, cause the computer to execute the crop growth management method based on the integration of space-air-ground in the first aspect.

[0043] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0044] Through the "space-air-ground" integration technology, the present invention integrates data from satellite remote sensing, unmanned aerial vehicle monitoring, and ground sensors, comprehensively covering all aspects of crop growth, converts data of different modalities into features recognizable by the crop growth model, improves the input quality of the model, ensures the comprehensiveness and accuracy of the data, makes up for the deficiencies of a single data source, can dynamically monitor the growth trajectory of crops through the construction and analysis of the growth trajectory matrix, timely discover potential risk sources, take measures in advance to reduce losses, dynamically adjust the target detection instructions, re-collect the preliminary monitoring data, form a closed-loop feedback mechanism, realize precise irrigation, precise fertilization, and precise pest control, reduce resource waste, reduce environmental pollution, and improve the sustainability of agricultural production. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 It is a schematic diagram of the framework of the crop growth management system of the present invention;

[0046] Figure 2 It is a flowchart of the crop growth management method of the present invention;

[0047] Figure 3 It is a schematic diagram of the structure of an electronic device of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0048] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0049] Embodiment 1: As Figure 1 shown, the present invention provides a crop growth management system based on sky-air-ground integration, including a data acquisition module 100, a data processing module 200, a crop growth analysis module 300, and a decision feedback module 400. Each module is connected by wire or wirelessly;

[0050] The data acquisition module 100 adjusts the frequency of data acquisition nodes within a preset period based on the target detection instruction, records the time stamps corresponding to the data acquisition nodes and the preliminary monitoring data; and sends the preliminary monitoring data to the data processing module 200;

[0051] Specifically, the target detection instruction is an operation instruction issued based on different application scenarios during the sky-air-ground integration operation process. Each operation instruction includes the acquisition frequency of the corresponding data acquisition node. Different data acquisition nodes obtain different preliminary monitoring data, and the preliminary monitoring data includes space-based remote sensing data, air-based detection data, and ground-based planting data.

[0052] More specifically, since the crop type and the crop planting time are not fixed, when using sky-air-ground integrated data monitoring, it is necessary to perform a process-based setting for the crops, configure different growth cycles for different growth stages, and mark them as preset periods; within each preset period, the data acquisition nodes are marked by the acquisition time stamps of the space-based remote sensing data, and the preliminary monitoring data obtained between adjacent data acquisition nodes is stored in the same time block, and the current time block is marked by the previous acquisition time node.

[0053] More specifically, the space-based remote sensing data is remote sensing data regularly collected by satellite sensors according to the satellite orbit and the geographical location of the farmland to capture the dynamic changes in crop growth. The satellite sensors are high-resolution, multi-spectral, and thermal infrared satellite sensors (such as Landsat, Sentinel-2, MODIS, etc.) that regularly obtain space-based remote sensing data covering a large area of farmland. The space-based remote sensing data includes vegetation cover indices (such as NDVI, DVI, PRI), as well as large-scale environmental variables such as soil moisture, surface temperature, and climate conditions.

[0054] The air-based detection data is obtained by using drones equipped with imaging equipment to obtain local image data of farmland; according to the preliminary analysis results of space-based remote sensing data or the predetermined crop monitoring plan, the flight path of the drone is planned, and the drone is equipped with high-resolution multi-spectral cameras, thermal imagers, RGB cameras and other sensor equipment to obtain detailed image data of local farmland, focusing on abnormal areas or crops in critical growth periods; the air-based detection data includes microscopic crop information such as crop canopy height, leaf area index (LAI), vegetation coverage, pest and disease status, water stress, etc. The air-based detection data is transmitted to the data processing center in real time via a wireless network, or uploaded in batches after the task is completed.

[0055] The ground-based planting data is crop growth data collected in real time based on a ground monitoring network, which includes soil moisture sensors, temperature sensors, nutrient sensors, and photosynthetically active radiation sensors deployed in the target planting area, real-time monitoring of soil moisture, temperature, nutrient content, and meteorological conditions of the planting environment data, and provides detailed information on crop growth. Ground sensors generally collect data every few minutes to hours, and transmit it to the central processing unit in real time through the Internet of Things (IoT) network to ensure the timeliness and continuity of the data.

[0056] The data processing module 200 cleans and formats the preliminary monitoring data to obtain a standard data format, performs spatiotemporal alignment on the preliminary monitoring data in the standard data format within a preset period to obtain stage detection data; and sends the stage detection data to the crop growth analysis module 300;

[0057] It should be noted that: Data cleaning: removes noise, invalid data and outliers to ensure the quality and reliability of input data. Use algorithms to automatically identify and remove abnormal data points, such as outliers caused by sensor failure or data transmission errors.

[0058] Data formatting: Unify the data formats from different sources and convert them into a standard data format for analysis, so as to facilitate subsequent fusion and analysis. This includes rasterization of image data and structured processing of sensor data.

[0059] The logic for acquiring the detection data in this stage is:

[0060] Using the target planting areas identified in space-based remote sensing data as references, a spatial planting platform is constructed;

[0061] Mark the target planting features identified in the ground planting data, and match them to the spatial planting platform in proportion to the target planting area to keep the geographic space consistent;

[0062] Overlay the contour lines of the local farmland image data identified in the space-based detection data with the contour lines of the local farmland image data and the target planting area, and add local image features within the overlapped target planting area;

[0063] Overlay the target planting features and local image features on the space planting platform in sequence according to the time series. The space planting platform superimposes and marks the detection marker icons collected within the target planting area based on the timestamp to mark the stage detection data of the current time block.

[0064] Further explanation: The overlay logic of the space planting platform is as follows:

[0065] The space planting platform arbitrarily selects a reference object from the current space-based remote sensing data as a reference point, uses the reference point as the center coordinate, and constructs a space coordinate system based on the reference point;

[0066] In the space coordinate system, classify and mark the reference object, and mark the area contour corresponding to the crop planting environment as the target planting area;

[0067] Completely overlap the area contours of the target planting areas corresponding to the space-based detection data and / or ground-based planting data with the area contours in the space-based remote sensing data, and determine the target planting area three-dimensional coordinate system with the target planting area corresponding to the ground-based planting data;

[0068] Extract the detection marker icons marked at different coordinate positions in the space-based detection data and / or ground-based planting data, and use the detection marker icons to represent the target planting features and local image features;

[0069] Superimpose and display all the detection marker icons on the target planting area three-dimensional coordinate system, and determine the azimuth information of the detection marker icons in the target planting area according to the center coordinate.

[0070] The crop growth analysis module 300 constructs a crop growth model, performs fusion processing on the stage detection data based on the crop growth model, and obtains the crop growth variables of the crop within a preset period; send the crop growth variables to the crop decision feedback module 400.

[0071] Specifically, the acquisition logic of the crop growth variables is as follows:

[0072] Perform stage slicing analysis on the stage detection data to obtain stage analysis data. The stage analysis data includes target feature data containing each detection marker icon, and the target feature data includes the reference object of the target planting area, target planting features, and local image features;

[0073] Semantically align the target feature data. For each target feature data, set a corresponding stacked autoencoder through a multi-modal autoencoder network, and convert the target feature data into target encoded features recognizable by the crop growth model;

[0074] Use the target encoded features as the input factors of the crop growth model, and the output factor of the crop growth model is the growth anomaly probability;

[0075] According to the timestamp, reconstruct the target encoded features and the growth anomaly probability and mark them as the growth trajectory matrix. Conduct growth anomaly statistics on the growth trajectory matrix, combine the pre-set reconstruction error threshold corresponding to each target encoded feature to determine the dimensionality reduction dimension, obtain the low-dimensional embedding of the current target encoded feature, and splice the low-dimensional embedding representations to obtain the crop growth variables.

[0076] Further explanation: Build a crop growth model. Use the target encoded features as the input factors of the crop growth model and the growth anomaly probability as the output factor. Use historical data and existing data to train the crop growth model through various machine learning algorithms such as random forest, logistic regression, and deep learning. The model training process includes steps such as feature selection, parameter optimization, and model verification, and generates prediction models for different crop types and growth stages.

[0077] Based on the real-time data of the current season, the model predicts the growth trend, yield, and health status of the crop at different growth stages. The model outputs include key indicators such as the growth trend level, potential yield, and maturity time; through model analysis, identify anomalies in the crop growth process, such as pests and diseases, insufficient water, nutrient deficiency, environmental stress, etc. Train based on the anomaly samples in the historical data, which can effectively capture the anomaly patterns, so that the growth anomaly probability that may cause anomalies is used as the output factor. After detecting an anomaly, the system automatically analyzes to obtain the crop growth variables and pushes them to farmers or agricultural managers in real time through mobile devices or management platforms, providing decision support for quick response.

[0078] The decision feedback module 400 obtains the yield prediction probability distribution based on the crop growth variables and generates a list of risk sources corresponding to the current probability distribution, and matches the corresponding crop growth strategies based on the list of risk sources; adjusts the target detection instructions through the crop growth strategies, adjusts the frequency of data collection nodes within the preset period, and improves and updates the preliminary monitoring data.

[0079] It should be noted that the crop growth strategy includes an irrigation strategy plan, a fertilization strategy plan, and a pest and disease control strategy plan, which optimize the use of water resources and fertilizers, improve the yield and quality of crops. The strategy plan takes into account the different growth stages of crops, soil conditions, and environmental factors. During the process of providing pest and disease control, the appropriate application time, dosage, and method of pesticides are considered. By reducing the overuse of pesticides, the system helps to reduce costs and environmental impacts, and controls the drone to conduct targeted inspections according to the preset path and preset time based on the target detection instruction.

[0080] Farmers carry out field management according to the crop growth strategy provided by the system, and the actual data during the management process is input into the system again for verifying and adjusting the prediction model. Through iterative optimization, the model gradually improves its prediction accuracy and adaptability.

[0081] Using the feedback data and new data sources, continuously adjust the model parameters and algorithm structure to enhance the model's adaptability to different environmental conditions and crop varieties. The model optimization process includes techniques such as cross-validation, incremental learning, and online learning.

[0082] Specifically, based on the crop growth strategy, look up the list of risk sources that match in the prior knowledge base, based on prior knowledge, the probability of yield impact corresponding to each crop growth variable in the risk source list, and based on the analysis threshold of the yield impact probability in prior knowledge, determine the target detection instruction.

[0083] Example 2: Based on Example 1, in the intelligent agriculture project in a certain area, the refined management of crop growth is achieved through the above process. The specific steps are as follows:

[0084] In the intelligent agriculture system, satellite remote sensing, drones, and ground sensors are used to collect preliminary monitoring data of crops. The preliminary monitoring data includes spectral information, growth height, leaf area index, soil moisture, and meteorological data, etc.

[0085] Clean and format the collected preliminary monitoring data. For example, use filtering algorithms to remove noise, fill in missing values, and convert data from different sources into a unified standard data format. Then perform spatio-temporal alignment. Use the dynamic time warping (DTW) method to align time series data to ensure that data at different time scales are on the same time basis. Spatial alignment is achieved through coordinate system transformation and geometric correction.

[0086] Conduct stage slicing analysis on the stage detection data to extract key features. Then perform semantic alignment to ensure that the features of different data sources are comparable at the semantic level. Use the crop growth model to fuse multi-source heterogeneous data to generate comprehensive crop growth variables. Crop growth variables include the growth status, health status, and yield prediction of crops, etc.

[0087] Build crop growth models, such as the WOFOST or CERES models. Use the crop growth variables generated by the crop growth analysis module as the input factors of the model, and the yield prediction probability distribution as the output factor. The model analyzes the input environmental data to predict crop yields in real time and adjusts the planting strategy according to the prediction results. For example, adjust irrigation and fertilization strategies based on the predicted yield probability distribution to optimize crop growth conditions. Feed the adjusted target detection instructions back to the data acquisition module to re-collect the preliminary monitoring data, forming a closed-loop feedback mechanism. This not only improves the accuracy of crop yield prediction but also optimizes the utilization of agricultural resources, enhancing the efficiency and sustainability of agricultural production.

[0088] Example 3: Refer to Figure 2 As shown, for the parts not described in detail in this example, refer to the description in Example 1. This example provides a method for monitoring and warning crop diseases and pests based on satellite remote sensing and AI algorithms, including the following steps:

[0089] Adjust the frequency of data acquisition nodes within a preset period based on the target detection instructions, and record the timestamps and preliminary monitoring data corresponding to the data acquisition nodes;

[0090] Clean and format the preliminary monitoring data to obtain the standard data format, and align the preliminary monitoring data in the standard data format spatiotemporally within a preset period to obtain the stage detection data;

[0091] Build a crop growth model, and perform fusion processing on the stage detection data based on the crop growth model to obtain the crop growth variables of the crop within a preset period;

[0092] Obtain the yield prediction probability distribution based on the crop growth variables and generate a list of risk sources corresponding to the current probability distribution. Match the corresponding crop growth strategies based on the risk source list; adjust the target detection instructions through the crop growth strategies, adjust the frequency of data acquisition nodes within a preset period, and improve and update the preliminary monitoring data.

[0093] The target detection instructions are operation instructions issued based on different application scenarios during the integrated sky-ground-space operation process to obtain different preliminary monitoring data. The preliminary monitoring data includes space-based remote sensing data, air-based detection data, and ground-based planting data;

[0094] The space-based remote sensing data is remote sensing data regularly collected by satellite sensors according to the satellite orbit and the geographical location of the farmland;

[0095] The air-based detection data is the local image data of the farmland obtained by using an unmanned aerial vehicle equipped with imaging equipment;

[0096] The foundation planting data is crop growth data collected in real time according to the ground monitoring network. The ground monitoring network includes soil moisture sensors, temperature sensors, nutrient sensors, and photosynthetically active radiation sensors deployed in the target planting area to monitor the planting environment data of soil moisture, temperature, nutrient content, and meteorological conditions in real time.

[0097] Divide the acquisition time nodes based on the acquisition timestamps of the space-based remote sensing data, store the preliminary monitoring data obtained between adjacent acquisition time nodes in the same time block, and mark the current time block with the previous acquisition time node.

[0098] The acquisition logic of the stage detection data is as follows:

[0099] Taking the target planting area identified in the space-based remote sensing data as a reference object, construct a spatial planting platform;

[0100] Mark the target planting characteristics identified in the foundation planting data and match them to the spatial planting platform in equal proportion through the target planting area to keep the geographical space consistent;

[0101] For the contour line of the farmland local image data identified in the space-based detection data, overlap the contour line of the farmland local image data with the target planting area, and add local image characteristics within the overlapped target planting area;

[0102] Overlay the target planting characteristics and local image characteristics on the spatial planting platform in sequence according to the time series. The spatial planting platform superimposes and marks the detection marker icons collected within the target planting area based on the timestamp to mark the stage detection data of the current time block.

[0103] The overlay logic of the spatial planting platform is as follows:

[0104] The spatial planting platform arbitrarily selects a reference object from the current space-based remote sensing data as a reference point, takes the reference point as the center coordinate, and constructs a spatial coordinate system based on the reference point;

[0105] In the spatial coordinate system, classify and mark the reference object, and mark the area contour corresponding to the crop planting environment as the target planting area;

[0106] Completely overlap the area contours of the target planting areas corresponding to the space-based detection data and / or the foundation planting data with the area contour in the space-based remote sensing data, and determine the target planting area three-dimensional coordinate system based on the target planting area corresponding to the foundation planting data;

[0107] Extract the detection marker icons marked at different coordinate positions in the space-based detection data and / or the foundation planting data, and represent the target planting characteristics and local image characteristics based on the detection marker icons;

[0108] All the detected marker icons are superposed and displayed on the three-dimensional coordinate system of the target planting area, and the azimuth information of the detected marker icons in the target planting area is determined according to the central coordinates.

[0109] The acquisition logic of the crop growth variables is as follows:

[0110] Perform stage slicing analysis on the stage detection data to obtain stage analysis data, where the stage analysis data includes target feature data containing each detected marker icon, and the target feature data includes the reference object, target planting features, and local image features of the target planting area;

[0111] Semantically align the target feature data, set a corresponding stacked autoencoder for each target feature data through a multi-modal autoencoder network, and convert the target feature data into target encoded features recognizable by the crop growth model;

[0112] Use the target encoded features as input factors of the crop growth model, and the output factor of the crop growth model is the growth anomaly probability;

[0113] Reconstruct and label the target encoded features and the growth anomaly probability according to the timestamp as the growth trajectory matrix, perform growth anomaly statistics on the growth trajectory matrix, combine the pre-set reconstruction error threshold corresponding to each target encoded feature to determine the dimensionality reduction dimension, obtain the low-dimensional embedding of the current target encoded feature, and splice the low-dimensional embedding representations to obtain the crop growth variables.

[0114] Based on the crop growth strategy, search for a list of risk sources that match in the prior knowledge base, and based on prior knowledge, determine the target detection instruction based on the probability of yield impact corresponding to each crop growth variable in the risk source list and the analysis threshold of the probability of yield impact.

[0115] Example 4: An electronic device shown according to an exemplary embodiment includes: a processor and a memory, where a computer program that can be called by the processor is stored in the memory;

[0116] The processor executes the above-mentioned crop growth management method based on the integration of sky-air-ground by calling the computer program stored in the memory.

[0117] Figure 3It is a schematic structural diagram of an electronic device provided by an embodiment of the present application. The electronic device may vary greatly due to different configurations or performances, and can include one or more processors (Central Processing Units, CPUs) and one or more memories. Among them, at least one computer program is stored in the memory, and the at least one computer program is loaded and executed by the processor to implement the sky-air-ground integrated crop growth management method provided by each of the above method embodiments. The electronic device can also include other components for implementing the functions of the device. For example, the electronic device can also have components such as wired or wireless network interfaces and input / output interfaces for input / output. Details are not described herein in the embodiments of the present application.

[0118] Embodiment 5: A computer-readable storage medium shown according to an exemplary embodiment, on which a rewritable computer program is stored;

[0119] When the computer program runs on a computer device, the computer device is caused to execute the above-mentioned sky-air-ground integrated crop growth management method.

[0120] In an exemplary embodiment, a computer-readable storage medium is also provided, such as a memory including at least one computer program, and the at least one computer program can be executed by a processor to complete the sky-air-ground integrated crop growth management method in the above embodiment. For example, the computer-readable storage medium can be a read-only memory (Read-Only Memory, abbreviated as: ROM), a random access memory (Random Access Memory, abbreviated as: RAM), a compact disc read-only memory (Compact Disc Read-Only Memory, abbreviated as: CD-ROM), magnetic tape, floppy disk, and optical data storage device, etc.

[0121] In an exemplary embodiment, a computer program product or a computer program is also provided. The computer program product or the computer program includes one or more program codes, and the one or more program codes are stored in a computer-readable storage medium. One or more processors of an electronic device can read the one or more program codes from the computer-readable storage medium, and the one or more processors execute the one or more program codes, so that the electronic device can execute the above-mentioned sky-air-ground integrated crop growth management method.

[0122] It should be understood that in various embodiments of the present application, the magnitudes of the serial numbers of the above processes do not mean the order of execution. The execution order of each process should be determined according to its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present application.

[0123] It should be understood that determining B based on A does not mean determining B solely based on A, and B can also be determined based on A and / or other information.

[0124] Those of ordinary skill in the art can understand that all or part of the steps for implementing the above embodiments can be completed by hardware, or can be completed by instructing relevant hardware through a program. This program can be stored in a computer-readable storage medium, and the above-mentioned storage medium can be a read-only memory, a magnetic disk, an optical disc, or the like.

[0125] The above description is only an optional embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present application shall be included within the protection scope of the present application.

[0126] In several embodiments provided by the present invention, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are only illustrative. For example, the division of the units is only one type, and there can be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces, and the indirect couplings or communication connections of the devices or units can be in electrical, mechanical, or other forms.

[0127] The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they can be located in one place, or can be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0128] In addition, the functional units in each embodiment of the present invention can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit.

[0129] As described above, this is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or replacements within the technical scope disclosed in the present application, and all should be covered by the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A crop growth management method based on sky-air-ground integration, characterized by: The following steps are involved: Adjust the frequency of data collection nodes within a preset period based on target detection instructions, and record the timestamps and preliminary monitoring data corresponding to the data collection nodes; Clean and format the preliminary monitoring data to obtain a standard data format, and align the preliminary monitoring data in the standard data format in time and space within a preset period to obtain phase detection data; Construct a crop growth model, fuse the stage detection data based on the crop growth model, and obtain the crop growth variables within a preset period; Based on crop growth variables, the yield prediction probability distribution is obtained and a list of risk sources corresponding to the current probability distribution is generated. The corresponding crop growth strategy is matched based on the list of risk sources. The target detection instructions are adjusted through the crop growth strategy, the frequency of data collection nodes within the preset period is adjusted, and the preliminary monitoring data is improved and updated.

2. The crop growth management method based on sky-air-ground integration according to claim 1 is characterized in that: The target detection instruction is an operation instruction issued based on different application scenarios during the space-air-ground integrated operation process, so as to obtain different preliminary monitoring data, which includes space-based remote sensing data, air-based detection data and ground-based planting data; The space-based remote sensing data is remote sensing data collected regularly by satellite sensors based on satellite orbits and farmland geographical locations; The air-based detection data is obtained by using a drone equipped with imaging equipment to obtain local image data of farmland; The ground-based planting data is crop growth data collected in real time based on a ground monitoring network. The ground monitoring network includes soil moisture sensors, temperature sensors, nutrient sensors and photosynthetically active radiation sensors deployed in the target planting area to monitor the planting environment data of soil moisture, temperature, nutrient content and meteorological conditions in real time.

3. The crop growth management method based on sky-air-ground integration according to claim 2 is characterized in that: The crops are set up in a process-based manner, and different growth cycles are configured for different growth stages and marked as preset cycles. In each preset cycle, the data collection nodes are benchmarked by the collection timestamp of the space-based remote sensing data, and the preliminary monitoring data obtained between adjacent data collection nodes are stored in the same time block, and the current time block is marked with the previous collection time node.

4. The crop growth management method based on sky-air-ground integration according to claim 3 is characterized in that: The acquisition logic of the detection data in this stage is: Using the target planting areas identified in space-based remote sensing data as references, a spatial planting platform is constructed; Mark the target planting features identified in the ground planting data, and match them to the spatial planting platform in proportion to the target planting area to keep the geographic space consistent; Overlap the contour lines of the local image data of the farmland identified in the air-based detection data with the target planting area, and add local image features in the overlapped target planting area; The target planting features and local image features are superimposed on the spatial planting platform in time series. The spatial planting platform superimposes the collected detection mark icons in the target planting area based on the timestamp to mark the stage detection data of the current time block.

5. The crop growth management method based on sky-air-ground integration according to claim 4 is characterized in that: The superposition logic of the space planting platform is: The spatial planting platform selects any reference object from the current space-based remote sensing data as a reference point, uses the reference point as the central coordinate, and constructs a spatial coordinate system based on the reference point; In the spatial coordinate system, the reference objects are classified and marked, and the area contour corresponding to the crop planting environment is marked as the target planting area; Completely overlap the area outline of the target planting area corresponding to the air-based detection data and / or the ground-based planting data with the area outline in the space-based remote sensing data, and determine the target planting area corresponding to the ground-based planting data as the three-dimensional coordinate system of the target planting area; Extracting detection mark icons marked at different coordinate positions in the air-based detection data and / or the ground-based planting data, and characterizing the target planting features and local image features based on the detection mark icons; All detection mark icons are displayed in a superimposed manner on the three-dimensional coordinate system of the target planting area, and the position information of the detection mark icons in the target planting area is determined according to the central coordinates.

6. The crop growth management method based on sky-air-ground integration according to claim 5 is characterized in that: The acquisition logic of the crop growth variables is: Performing stage-by-stage slice analysis on the stage detection data to obtain stage analysis data, wherein the stage analysis data includes target feature data containing various detection mark icons, and the target feature data includes a reference object of a target planting area, target planting features, and local image features; Semantically align the target feature data, set a corresponding stacked autoencoder for each target feature data through a multimodal autoencoder network, and convert the target feature data into target coding features that can be recognized by the crop growth model; The target coding feature is used as the input factor of the crop growth model, and the output factor of the crop growth model is the probability of abnormal growth; According to the timestamp, the target coding features and the probability of growth anomaly are reconstructed and marked as a growth trajectory matrix, and growth anomaly statistics are performed on the growth trajectory matrix. The dimension reduction is determined by combining the pre-set reconstruction error threshold corresponding to each target coding feature to obtain a low-dimensional embedding of the current target coding feature. The low-dimensional embedding representation is spliced ​​to obtain the crop growth variable.

7. The crop growth management method based on sky-air-ground integration according to claim 6 is characterized in that: Based on the crop growth strategy, a matching list of risk sources is searched in the prior knowledge base. Based on the prior knowledge, the yield impact probability corresponding to each crop growth variable in the risk source list is determined. Based on the analysis threshold of the yield impact probability based on the prior knowledge, the target detection instruction is determined.

8. A crop growth management system based on sky-air-ground integration, based on the implementation of the crop growth management method based on sky-air-ground integration according to any one of claims 1 to 7, characterized in that: It comprises a data collection module (100), a data processing module (200), a crop growth analysis module (300) and a decision feedback module (400), wherein each module is connected via wired or wireless connections; A data collection module (100) adjusts the frequency of data collection nodes within a preset period based on the target detection instruction, records the timestamp and preliminary monitoring data corresponding to the data collection node, and sends the preliminary monitoring data to a data processing module (200); A data processing module (200) performs data cleaning and formatting on the preliminary monitoring data to obtain a standard data format, and performs spatiotemporal alignment on the preliminary monitoring data in the standard data format within a preset period to obtain phase detection data; Sending the stage detection data to the crop growth analysis module (300); The crop growth analysis module (300) constructs a crop growth model, performs fusion processing on the stage detection data based on the crop growth model, obtains crop growth variables of the crop within a preset period; and sends the crop growth variables to the crop decision feedback module (400); The decision feedback module (400) obtains the yield prediction probability distribution based on the crop growth variables and generates a risk source list corresponding to the current probability distribution, matches the corresponding crop growth strategy based on the risk source list; adjusts the target detection instructions through the crop growth strategy, adjusts the frequency of data collection nodes within a preset period, and improves and updates the preliminary monitoring data.

9. An electronic device, characterized in that: include: A processor and a memory, wherein the memory stores a computer program that can be called by the processor; The processor executes the sky-air-ground integrated crop growth management method according to any one of claims 1 to 7 by calling the computer program stored in the memory.

10. A computer-readable storage medium, characterized in that: Instructions are stored, and when the instructions are executed on a computer, the computer is caused to execute the crop growth management method based on sky-air-ground integration as described in any one of claims 1 to 7.

Images