Rice blast monitoring and diagnosing method based on captive hot-air balloon and unmanned aerial vehicle system

Abstract

The invention belongs to the technical field of agricultural disease monitoring, prevention and control, and particularly relates to a rice blast monitoring and diagnosing method based on a mooring hot-air balloon and an unmanned aerial vehicle system. By constructing a dynamic observation platform of a tethered hot-air balloon and an unmanned aerial vehicle group, multi-variety, multi-temporal and multi-dimensional rice blast diseased plant multi-spectral / hyperspectral data and multi-information such as geography, soil, climate and field management are synchronously obtained, and a spectral database is constructed. An early diagnosis inversion model is constructed based on spectral characteristics of diseased rice plants and data characteristics of an unmanned aerial vehicle and a captive hot-air balloon by using self-supervised learning and deep learning technologies. According to the method, the cooperative advantages of the tethered hot-air balloon and the unmanned aerial vehicle are fully played, comprehensive acquisition and deep fusion of data are realized, the accuracy and timeliness of rice blast monitoring and diagnosis are improved, powerful technical support is provided for healthy development of the rice industry, and agricultural intelligence and sustainable development are assisted.

Description

Technical Field

[0001] The present invention relates to the technical field of agricultural disease monitoring and prevention, and in particular to a rice blast monitoring and diagnosis method based on a tethered hot air balloon and an unmanned aerial vehicle system. Background Art

[0002] The statements in this section merely provide background technology related to the present invention and do not necessarily constitute prior art.

[0003] As one of the world's major food crops, rice plays a pivotal role in global food security. As a major rice producer and consumer, my country has a wide rice planting area and huge yields. However, the rice industry faces many challenges, among which rice blast is a key factor that seriously affects rice yield and quality. Rice blast causes a significant reduction in rice production every year, seriously threatening food security.

[0004] At present, although certain achievements have been made in the field of rice disease prevention and control, there are still many problems that need to be solved. In terms of monitoring methods, the existing methods are relatively single, and there is a lack of effective technology and platforms to achieve spatiotemporal coordinated observation of rice diseases, resulting in a lack of sufficient information support for rice disease detection and prevention and control management. Traditional monitoring methods are difficult to fully and real-time grasp the occurrence and development of rice blast in different regions and different growth stages. In terms of remote sensing data integration, the existing data acquisition system has an island phenomenon, the data fusion mechanism is imperfect, and the information chain is easy to break, which seriously hinders the deep use of data. This makes it difficult to extract valuable information from massive remote sensing data for accurate diagnosis of rice blast. In the field of diagnosis and prediction, there are trade-offs in the spectral, spatial, and temporal resolution of historical data, and the existing model relies too much on big data + manual training mode, making it difficult to achieve fully automated high-precision disease diagnosis and trend prediction. This leads to the inability to timely and accurately predict the outbreak trend of rice blast in practical applications, and it is difficult to take effective prevention and control measures in advance.

[0005] In addition, although UAV technology has been applied in the agricultural field, it still has limitations in payload and endurance, and the operating radius and efficiency need to be improved. The application of tethered hot air balloons in agricultural remote sensing is relatively rare, especially in the exploration of coordinated monitoring of rice diseases with UAVs is still in its infancy. At the same time, although major research teams in my country have their own advantages in rice blast research, the lack of effective cooperation and data sharing is not conducive to the construction of a comprehensive and accurate rice blast diagnosis model. An innovative and comprehensive rice blast monitoring and diagnosis technology is urgently needed to improve the information level and disease prevention and control capabilities of the rice industry. Summary of the invention

[0006] In order to overcome the many deficiencies in the current rice blast monitoring and diagnosis technology, the present invention provides a rice blast monitoring and diagnosis method based on a tethered hot air balloon and an unmanned aerial vehicle system. The method constructs a dynamic observation platform based on a tethered hot air balloon + unmanned aerial vehicle, collects multi- / hyperspectral information of rice blast diseased plants of multiple varieties, multiple phases, and multiple dimensions, and integrates multivariate data such as geography, soil, climate, and field management, and uses a self-supervised learning model combined with deep learning technology to construct an early diagnosis inversion model for rice blast. This invention gives full play to the synergistic advantages of tethered hot air balloons and unmanned aerial vehicle systems, effectively improves the accuracy and timeliness of rice blast monitoring and diagnosis, provides a new technical approach for accurately preventing and controlling rice blast and ensuring rice yield and quality, and promotes the intelligentization and sustainable development of the rice industry.

[0007] In order to achieve the above object, the present invention adopts the following technical solution: The first aspect of the present invention provides a method for monitoring and diagnosing rice blast based on a tethered hot air balloon and an unmanned aerial vehicle system: Build a dynamic observation platform based on tethered hot air balloons + drone groups, determine the coordination framework of tethered hot air balloons and drone systems and the sensor equipment they carry, carry mobile signal base stations, build multi-channel data interaction methods, and achieve fast and stable communication connections and data transmission between operating systems, ensuring that drones and tethered hot air balloons can simultaneously acquire multi / hyperspectral data in near-low-altitude composite airspaces; Based on the dynamic observation platform, we collected multi- / hyperspectral information of rice blast diseased plants of multiple varieties, multiple phases, and multiple dimensions, integrated geographic, soil, climate, and field management data, optimized the data preprocessing process, including denoising, radiation correction, and image enhancement, developed data fusion tools, realized the complementarity of UAV and tethered hot air balloon data, extracted common key bands and disease indices that are sensitive to rice blast diseased plants, and constructed a spectral database of rice blast disease of multiple varieties; Based on the spectral characteristics of rice diseased plants and the data features of drones and tethered hot air balloons, a self-supervised learning model is combined with deep learning technology to enhance the ability to recognize the hallmark features of rice blast. The continuous data stream provided by drones and tethered hot air balloons is used to enhance the dynamic change tracking ability of the rice blast monitoring model and construct an early diagnosis inversion model for rice blast.

[0008] As a further limitation of the first aspect of the present invention, when constructing the dynamic observation platform and the internal data transmission channel, the following details are added: Comprehensively investigate and analyze the monitoring needs of rice blast and rice growth characteristics, clarify the key growth stages of rice that need to be monitored (tillering stage, heading stage), common types of rice blast (leaf blast, panicle blast), causes of disease (climatic conditions, soil fertility, variety resistance) and key indicators of disease development (lesion area, disease index), to ensure that the platform system design closely fits the actual needs of agriculture; The flight altitude of the near-low altitude unmanned system is accurately determined based on factors such as the experimental rice field area, topography, expected experimental results, and local meteorological conditions. The flight altitude of the drone is set between 30 and 60 meters, and is flexibly adjusted according to different monitoring stages and rice varieties. It is equipped with a high-resolution multispectral camera and a high-precision temperature and humidity sensor to quickly collect detailed rice field crop information. The KX55 series tethered hot air balloons of the Chinese Academy of Sciences' Institute of Space Information Innovation are used as continuous monitoring equipment for rice fields. They fly at an altitude of 2000-3000m, stay in the air for 7-10 days, cover a radius of 160-200km, and are equipped with wind direction and speed sensors, atmospheric pressure sensors, and high-resolution hyperspectral sensors. The tethered hot air balloon uses a specially made high-strength, lightweight mounting structure to stably install a high-performance mobile signal base station on its hanging basket or specific bearing part. The base station has powerful signal transmission and reception capabilities and uses multi-band antenna technology. It can communicate stably with different types of drones and ground control terminals. Its transmission power can be dynamically adjusted to meet the communication needs of different distances and environments. The base station has a built-in high-performance signal processor that can perform real-time encoding, decoding and error correction processing on the transmission signal, effectively reducing the bit error rate during signal transmission and ensuring the accuracy of data transmission between it and the drone. In the process of multi-channel data interaction, the Paxos protocol is fully introduced to assign a unique identifier to each communication channel and establish a distributed node status management mechanism. When the drone sends a data request or status update to the tethered hot air balloon base station, the base station first parses the request and extracts key information (request type, source node identifier, data content, etc.), and then initiates a multi-node voting process based on the Paxos protocol to ensure that a consensus is reached on the same operation on multiple communication channels. That is, when the drone uploads rice blast monitoring data, the base station coordinates each channel through the Paxos protocol to ensure that all relevant nodes receive, store and process data in the same order to avoid data conflicts and inconsistencies. Through strict consistency assurance, the high reliability and stability of multi-channel interaction is achieved, effectively preventing monitoring interruptions or misdiagnoses caused by communication failures or data errors.

[0009] As a further limitation of the first aspect of the present invention, when processing and analyzing a variety of remote sensing data and ground monitoring data, the following details are added: According to the diversity of rice varieties and the complexity of rice blast disease, multiple sampling time points are set for different rice varieties (common varieties such as indica rice, japonica rice, glutinous rice and local specialty varieties) throughout their growth cycle, covering key stages of rice from seeding, tillering, jointing, booting, heading to maturity, etc., to ensure the collection of multi-temporal data. In the spatial dimension, different monitoring areas are divided according to the layout and terrain characteristics of the rice fields. Multiple sampling points are set in each area to form a multi-dimensional monitoring network. The advantage of UAVs in flexible low-altitude flight is used to conduct close-range and refined multi / hyperspectral image collection of rice in different regions and varieties, and obtain spectral information of diseased plants at different growth stages and different parts (leaves, stems, rice ears, etc.). The tethered hot air balloon conducts macroscopic monitoring of the entire rice field at high altitude to obtain overall information on rice growth status and disease distribution over a large area. The two work together to achieve all-round and multi-level data collection. A unified data integration framework was established to correlate and integrate the collected geographic information, soil data, climate data, and field management data with the multi- / hyperspectral information of rice blast diseased plants. To address the noise problem in multi- / hyperspectral images, an adaptive filtering denoising algorithm was used to automatically adjust the filtering parameters according to the local features of the image to effectively remove noise interference. A radiation correction method based on a physical model was used in combination with field-measured atmospheric parameters (aerosol optical thickness, water vapor content) to accurately correct the radiation values ​​received by the sensor and restore the true spectral reflectance characteristics of rice. For image enhancement, a combination of nonlinear stretching and histogram equalization was used to enhance the contrast between the diseased plant features and the background, highlight the disease details, improve data quality, and provide a reliable basis for subsequent analysis. Based on the principle of multi-scale analysis and feature fusion, the wavelet transform is used to decompose the image into sub-bands of different scales and directions, and the high-resolution detail information collected by the UAV and the macro information obtained by the tethered hot air balloon are layered and fused to retain the global features and local details of the image. The gray-level co-occurrence matrix in the UAV image is extracted to calculate the texture parameters, the shape characteristics of the spectral curve, and the vegetation index and color characteristics in the tethered hot air balloon image. The principal component analysis (PCA) method is used for feature fusion to construct a comprehensive feature vector, giving full play to the advantages of the UAV and tethered hot air balloon data. By comparing and analyzing the reflectance differences between healthy rice plants and diseased plants with different degrees of disease in each spectral band, combined with statistical methods such as correlation analysis and stepwise discriminant analysis, the bands with significant differences were screened out. Through verification and optimization of a large amount of sample data, the disease index that can most effectively characterize the rice blast condition was determined, and a MySQL database was established to store rice variety information, spectral data, geographical soil and climate field management data, disease information, and extracted key bands and disease index information.

[0010] As a further limitation of the first aspect of the present invention, when constructing the rice blast disease monitoring model, the following details are added: Representative spectral data of diseased rice plants and corresponding spectral data of healthy plants were selected from the MySQL database as the basic data set for model construction. The diseased plant data were finely annotated, and the location, shape, size, and severity of the lesions were recorded in detail. The comprehensive feature vectors obtained in S23 were used as the input parameters of the model. Construct a deep learning model architecture based on self-supervised learning, using an encoder-decoder structure. The encoder part consists of multiple convolutional layers and pooling layers, which are used to encode the input multimodal feature vector, gradually extract abstract features in the data, and reduce the data dimension. The decoder part maps the encoded features back to the original data space through upsampling layers and deconvolution layers to achieve data reconstruction. Jump connections are introduced between the encoder and decoder to retain feature information at different levels and enhance the model's ability to capture detailed features. The model adopts a self-supervised learning strategy for model training, and uses the intrinsic structural information of the data itself to design pre-training tasks. By randomly cropping, rotating, flipping and other transformations on the input image, the model is allowed to predict the relationship between the transformed image and the original image, thereby learning the invariant characteristics of the data. In the pre-training stage, a large amount of unlabeled data is used for training, so that the model can automatically discover the potential patterns and feature representations in the data; After pre-training, the model is fine-tuned using labeled data to optimize model parameters to adapt to the rice blast diagnosis task. During the fine-tuning process, the cross entropy loss function is used as the optimization target of the model to measure the difference between the model prediction results and the true labels. At the same time, the stochastic gradient descent (SGD) algorithm is combined to dynamically adjust the learning rate according to the distribution characteristics of the data to accelerate model convergence and improve the generalization ability of the model. The model is evaluated using the validation set to monitor the performance of the model in terms of accuracy, recall, F1 value and other indicators. When the model performance no longer improves, training is stopped to prevent overfitting. Establish a real-time data stream processing system to receive new data continuously collected by drones and tethered hot air balloons, perform real-time preprocessing on the new data, including data cleaning, format conversion, and feature extraction in the same training phase, and input the extracted dynamic features into the trained model for real-time rice blast monitoring and diagnosis; In order to adapt to the dynamic changes in the rice growth process and the evolution of rice blast, a dynamic feature update mechanism is designed. As time goes by and data accumulates, the model is dynamically updated using an incremental learning method, allowing the model to continuously learn new data based on existing knowledge without retraining the entire model, thus avoiding the "catastrophic forgetting" problem. When new data accumulates to a certain extent, the model is fine-tuned using the new data and some parameters of the model are updated, so that the model can promptly adapt to new changes and trends in rice blast. In the process of model updating, a reward mechanism is introduced in combination with reinforcement learning ideas. Corresponding rewards are given according to the performance of the model in actual monitoring (accurately predicting the time of disease outbreak and accurately judging the severity of the disease). This encourages the model to continuously optimize its own parameters and improve monitoring accuracy and reliability. At the same time, a model performance evaluation index system is established. In addition to traditional indicators such as accuracy and recall rate, the timeliness and stability factors of the model are additionally considered to comprehensively evaluate the performance of the model. Targeted optimization and improvement are carried out based on the evaluation results to ensure that the constructed rice blast early diagnosis inversion model always maintains a good performance state and provides timely and effective disease warning and prevention and control decision support for rice production.

[0011] Compared with the prior art, the present invention has the following beneficial effects: 1. The present invention innovatively combines a tethered hot air balloon with an unmanned aerial vehicle system, realizing the simultaneous acquisition of multi- / hyperspectral data in near- and low-altitude composite airspaces, greatly expanding the monitoring range and data collection dimensions. The tethered hot air balloon stays at high altitudes for a long time, providing a large-area macroscopic monitoring field of view, while the unmanned aerial vehicle can perform detailed detection of key areas at low altitudes. The collaborative operation of the two effectively makes up for the limitations of a single monitoring platform.

[0012] 2. By comprehensively integrating multi-source data such as geography, soil, climate, and field management, and deeply integrating them with the multi- / hyperspectral information of rice blast diseased plants, a spectral database of rice blast disease in multiple varieties of rice was constructed, covering all-round information on rice growth environment and disease characteristics, forming a comprehensive data resource library, which provides solid data support for subsequent in-depth analysis of the pathogenesis, transmission patterns, and precise prevention and control of rice blast disease.

[0013] 3. Based on the spectral characteristics of rice diseased plants and the data characteristics of drones and tethered hot air balloons, combined with the continuous data streams provided by drones and tethered hot air balloons, an early diagnosis inversion model for rice blast is constructed using a self-supervised learning model combined with deep learning technology. It has powerful feature learning and pattern recognition capabilities, and realizes real-time tracking of the dynamic changes of rice blast.

[0014] 4. The monitoring and diagnosis system constructed by the present invention forms a complete set of technical processes, from the construction of a dynamic observation platform, data collection and integration, model construction and optimization to the final real-time monitoring and early warning, and each link is closely linked and mutually supportive.

[0015] Advantages of additional aspects of the present invention will be given in part in the following description, and in part will become obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0017] Figure 1 A technical roadmap for the rice blast disease monitoring and diagnosis method based on a tethered hot air balloon and an unmanned aerial vehicle system provided by the present invention; DETAILED DESCRIPTION

[0018] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0019] It should be noted that the following detailed descriptions are exemplary and are intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.

[0020] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0021] In the absence of conflict, the embodiments of the present invention and the features of the embodiments may be combined with each other.

[0022] The technical solution of the embodiment of the present application will be described below in conjunction with the drawings in the embodiment of the present application.

[0023] A rice blast monitoring and diagnosis method based on a tethered hot air balloon and an unmanned aerial vehicle system comprises the following steps: S10. Build a dynamic observation platform based on tethered hot air balloons + drone groups, determine the coordination framework of tethered hot air balloons and drone systems and the sensor equipment they carry, carry mobile signal base stations, build a multi-channel data interaction method, and achieve fast and stable communication connections and data transmission between operating systems, ensuring that drones and tethered hot air balloons acquire multi / hyperspectral data in near-low-altitude composite airspaces synchronously; S20. Based on the dynamic observation platform, collect multi- / hyperspectral information of rice blast diseased plants of multiple varieties, multiple phases, and multiple dimensions, integrate geographic, soil, climate, and field management data, optimize data preprocessing processes, including denoising, radiation correction, and image enhancement, develop data fusion tools, realize the complementarity of UAV and tethered hot air balloon data, extract common key bands and disease indices that are sensitive to rice blast diseased plants, and build a spectral database of rice blast disease of multiple varieties; S30. Based on the spectral characteristics of diseased rice plants and the data features of drones and tethered hot air balloons, a self-supervised learning model is used in combination with deep learning technology to enhance the ability to recognize the hallmark features of rice blast. The continuous data streams provided by drones and tethered hot air balloons are used to enhance the dynamic change tracking capability of the rice blast monitoring model, and to construct an early diagnosis inversion model for rice blast.

[0024] Preferably, S10 also includes the following steps: S11. Comprehensively investigate and analyze the monitoring needs of rice blast and rice growth characteristics, clarify the key growth stages of rice that need to be monitored (tillering stage, booting stage), common types of rice blast (leaf blast, panicle blast), causes of disease (climatic conditions, soil fertility, variety resistance) and key indicators of disease development (lesion area, disease index), to ensure that the platform system design closely fits the actual needs of agriculture; S12. According to the experimental rice field area, topography, expected experimental results and local meteorological conditions, the flight altitude of the near-low altitude unmanned system is accurately determined. The flight altitude of the drone is set between 30 and 60 meters, and the altitude is flexibly adjusted according to different monitoring stages and rice varieties. It is equipped with a high-resolution multispectral camera and a high-precision temperature and humidity sensor to quickly collect detailed rice field crop information; S13. Use the KX55 series tethered hot air balloon of the Chinese Academy of Sciences' Institute of Space Information Innovation as a continuous monitoring device for rice fields, with a flight altitude of 2000-3000m, a stay time of 7-10 days, a coverage radius of 160-200km, and equipped with wind direction and speed sensors, atmospheric pressure sensors, and high-resolution hyperspectral sensors; S14. The tethered hot air balloon uses a specially made high-strength, lightweight mounting structure to stably install a high-performance mobile signal base station on its hanging basket or specific bearing part. The base station has powerful signal transmission and reception capabilities and uses multi-band antenna technology. It can communicate stably with different types of drones and ground control terminals. Its transmission power can be dynamically adjusted to meet the communication needs at different distances and environments. The base station has a built-in high-performance signal processor that can perform real-time encoding, decoding and error correction processing on the transmission signal, effectively reducing the bit error rate during signal transmission and ensuring the accuracy of data transmission between it and the drone. S15. In the process of multi-channel data interaction, the Paxos protocol is fully introduced to assign a unique identifier to each communication channel and establish a distributed node status management mechanism. When the drone sends a data request or status update to the tethered hot air balloon base station, the base station first parses the request and extracts key information (request type, source node identifier, data content, etc.), and then initiates a multi-node voting process based on the Paxos protocol to ensure that a consensus is reached on the same operation on multiple communication channels. That is, when the rice blast disease monitoring data is uploaded to the drone, the base station coordinates each channel through the Paxos protocol to ensure that all relevant nodes receive, store and process the data in the same order to avoid data conflicts and inconsistencies. Through strict consistency assurance, the high reliability and stability of multi-channel interaction is achieved, effectively preventing monitoring interruptions or misdiagnosis caused by communication failures or data errors.

[0025] Preferably, S20 also includes the following steps: S21. Based on the diversity of rice varieties and the complexity of rice blast disease, multiple sampling time points are set for different rice varieties (common varieties such as indica rice, japonica rice, glutinous rice and local specialty varieties) throughout their growth cycle, covering key stages of rice from seeding, tillering, jointing, booting, heading to maturity, etc., to ensure the collection of multi-temporal data. In the spatial dimension, different monitoring areas are divided according to the layout and terrain characteristics of the rice fields. Multiple sampling points are set in each area to form a multi-dimensional monitoring network. The advantage of UAVs in flexible low-altitude flight is used to conduct close-range and refined multi- / hyperspectral image acquisition of rice in different regions and varieties, and obtain spectral information of diseased plants at different growth stages and different parts (leaves, stems, rice ears, etc.). The tethered hot air balloon conducts macroscopic monitoring of the entire rice field at high altitude to obtain overall information on rice growth status and disease distribution over a large area. The two work together to achieve all-round and multi-level data collection. S22. Establish a unified data integration framework to associate and integrate the collected geographic information, soil data, climate data, and field management data with the multi-spectral / hyperspectral information of rice blast diseased plants. To address the noise problem in multi-spectral / hyperspectral images, use an adaptive filtering denoising algorithm to automatically adjust the filtering parameters according to the local characteristics of the image to effectively remove noise interference. Use a radiation correction method based on a physical model, combined with field-measured atmospheric parameters (aerosol optical thickness, water vapor content), accurately correct the radiation values ​​received by the sensor, and restore the true spectral reflectance characteristics of rice. For image enhancement, use a combination of nonlinear stretching and histogram equalization to enhance the contrast between the characteristics of the diseased plants and the background, highlight the disease details, improve data quality, and provide a reliable basis for subsequent analysis; S23. Based on the principle of multi-scale analysis and feature fusion, the wavelet transform is used to decompose the image into sub-bands of different scales and directions, and the high-resolution detail information collected by the UAV and the macro information obtained by the tethered hot air balloon are layered and fused to retain the global features and local details of the image. The gray-level co-occurrence matrix in the UAV image is extracted to calculate the texture parameters, the shape characteristics of the spectral curve, and the vegetation index and color characteristics in the tethered hot air balloon image. The principal component analysis (PCA) method is used for feature fusion to construct a comprehensive feature vector, giving full play to the advantages of the UAV and tethered hot air balloon data. S24. By comparing and analyzing the reflectance differences between healthy rice plants and diseased plants with different disease severity in each spectral band, combined with statistical methods such as correlation analysis and stepwise discriminant analysis, the bands with significant differences are screened out. Through verification and optimization of a large amount of sample data, the disease index that can most effectively characterize the condition of rice blast is determined, and a MySQL database is established to store rice variety information, spectral data, geographic soil climate field management data, disease information, and extracted key bands and disease index information.

[0026] Preferably, S30 also includes the following steps: S31, selecting representative spectral data of diseased rice plants and corresponding spectral data of healthy plants from the MySQL database as the basic data set for model construction, finely annotating the diseased plant data, and recording in detail information such as the location, shape, size, and severity of the diseased spots, and combining them with the comprehensive feature vector obtained in S23 as the input parameters of the model; S32. Construct a deep learning model architecture based on self-supervised learning, using an encoder-decoder structure. The encoder part consists of multiple convolutional layers and pooling layers, which are used to encode the input multimodal feature vector, gradually extract abstract features in the data, and reduce the data dimension. The decoder part maps the encoded features back to the original data space through upsampling layers and deconvolution layers to achieve data reconstruction. Jump connections are introduced between the encoder and decoder to retain feature information at different levels and enhance the model's ability to capture detailed features. S33. The model adopts a self-supervised learning strategy for model training, and uses the intrinsic structural information of the data itself to design pre-training tasks. By randomly cropping, rotating, flipping and other transformations on the input image, the model is allowed to predict the relationship between the transformed image and the original image, thereby learning the invariant characteristics of the data. In the pre-training stage, a large amount of unlabeled data is used for training, so that the model can automatically discover the potential patterns and feature representations in the data; S34. After pre-training, the model is fine-tuned using labeled data to optimize model parameters to adapt to the rice blast diagnosis task. During the fine-tuning process, the cross entropy loss function is used as the optimization target of the model to measure the difference between the model prediction results and the true labels. At the same time, the stochastic gradient descent (SGD) algorithm is combined to dynamically adjust the learning rate according to the distribution characteristics of the data to accelerate model convergence and improve the generalization ability of the model. The model is evaluated using the validation set to monitor the performance of the model in terms of accuracy, recall, F1 value and other indicators. When the model performance no longer improves, the training is stopped to prevent overfitting. S35. Establish a real-time data stream processing system to receive new data continuously collected by drones and tethered hot air balloons, perform real-time preprocessing on the new data, including data cleaning, format conversion, and feature extraction in the same training phase, and input the extracted dynamic features into the trained model to perform real-time rice blast monitoring and diagnosis; S36. In order to adapt to the dynamic changes in the rice growth process and the evolution of rice blast, a dynamic feature update mechanism is designed. As time goes by and data accumulates, the model is dynamically updated using an incremental learning method, allowing the model to continuously learn new data based on existing knowledge without retraining the entire model, thus avoiding the "catastrophic forgetting" problem. When new data accumulates to a certain extent, the model is fine-tuned using the new data and some parameters of the model are updated, so that the model can promptly adapt to new changes and trends in rice blast; S37. In the process of model updating, combined with the idea of ​​reinforcement learning, a reward mechanism is introduced. According to the performance of the model in actual monitoring (accurately predicting the outbreak time of the disease, accurately judging the severity of the disease), corresponding rewards are given to encourage the model to continuously optimize its own parameters and improve the accuracy and reliability of monitoring. At the same time, a model performance evaluation index system is established. In addition to traditional indicators such as accuracy and recall rate, the timeliness and stability factors of the model are additionally considered to comprehensively evaluate the performance of the model. According to the evaluation results, targeted optimization and improvement are carried out to ensure that the constructed rice blast early diagnosis inversion model always maintains a good performance state, providing timely and effective disease warning and prevention and control decision support for rice production.

Claims

1. A method for monitoring and diagnosing rice blast based on a tethered hot air balloon and an unmanned aerial vehicle system, characterized in that: The following steps are involved: S10. Build a dynamic observation platform based on tethered hot air balloons and drone groups, determine the coordination framework of tethered hot air balloons and drone systems and the sensor equipment they carry, carry mobile signal base stations, build a multi-channel data interaction method, and achieve fast and stable data transmission between operating systems, ensuring that drones and tethered hot air balloons can simultaneously obtain multi / hyperspectral data in near-low-altitude composite airspaces; S20. Based on the dynamic observation platform, collect multi- / hyperspectral information of rice blast diseased plants of multiple varieties, multiple phases, and multiple dimensions, integrate geographic, soil, climate, and field management data, optimize data preprocessing processes, including denoising, radiation correction, and image enhancement, develop data fusion tools, realize the complementarity of UAV and tethered hot air balloon data, extract common key bands and disease indices that are sensitive to rice blast diseased plants, and build a spectral database of rice blast disease of multiple varieties; S30. Based on the spectral characteristics of diseased rice plants and the data features of drones and tethered hot air balloons, a self-supervised learning model is used in combination with deep learning technology to enhance the ability to recognize the hallmark features of rice blast. The continuous data streams provided by drones and tethered hot air balloons are used to enhance the dynamic change tracking capability of the rice blast monitoring model, and to construct an early diagnosis inversion model for rice blast.

2. The rice blast monitoring and diagnosis method according to claim 1, characterized in that: The step S10 also includes: S11. Comprehensively investigate and analyze the monitoring needs of rice blast and rice growth characteristics, clarify the key growth stages of rice that need to be monitored, common types of rice blast, causes of disease and key indicators of disease development, and ensure that the platform system design closely fits the actual needs of agriculture; S12. Determine the flight altitude of the near-low altitude unmanned system based on the experimental rice field area, topography, expected experimental results and local meteorological conditions. The flight altitude of the drone is set between 30m and 60m. The altitude is adjusted according to different monitoring stages and rice varieties. It is equipped with a high-resolution multispectral camera and a high-precision temperature and humidity sensor to quickly collect detailed rice field crop information. S13. Use tethered hot air balloons as continuous monitoring equipment for rice fields, with a flight altitude of 2km-3km, a stay time of 7-10 days, a coverage radius of 160-200km, and equipped with wind direction and speed sensors, atmospheric pressure sensors, and high-resolution hyperspectral sensors; S14. The tethered hot air balloon uses a special high-strength, lightweight mounting structure to stably install the mobile signal base station on its hanging basket or specific bearing part. The base station has powerful signal transmission and reception capabilities and adopts multi-band antenna technology. It can communicate stably with different types of drones and ground control terminals. Its transmission power can be dynamically adjusted to meet the communication needs at different distances and environments. The base station has a built-in high-performance signal processor that can perform real-time encoding, decoding and error correction processing on the transmission signal, effectively reducing the bit error rate during signal transmission and ensuring the accuracy of data transmission between it and the drone. S15. In the process of multi-channel data interaction, the Paxos protocol is fully introduced, a unique identifier is assigned to each communication channel, and a distributed node status management mechanism is established. When the UAV sends a data request or status update to the tethered hot air balloon base station, the tethered hot air balloon base station first parses the request and extracts key information, and then initiates a multi-node voting process based on the Paxos protocol to ensure that consensus is reached on the same operation on multiple communication channels. That is, when the UAV uploads rice blast monitoring data, the tethered hot air balloon base station coordinates each channel through the Paxos protocol to ensure that all relevant nodes receive, store and process data in a consistent order, avoid data conflicts and inconsistencies, and effectively prevent monitoring interruptions or misdiagnoses caused by communication failures or data errors.

3. The rice blast monitoring and diagnosis method according to claim 2, characterized in that: The step S20 also includes: S21. Based on the diversity of rice varieties and the complexity of rice blast disease, multiple sampling time points are set for different rice varieties throughout their growth cycle, covering key stages of rice from seeding, tillering, jointing, booting, heading to maturity, to ensure the collection of multi-temporal data. In the spatial dimension, different monitoring areas are divided according to the layout of rice fields and terrain characteristics. Multiple sampling points are set in each area to form a multi-dimensional monitoring network. The advantage of UAVs in flexible low-altitude flight is used to conduct close-range and refined multi / hyperspectral image collection of rice in different regions and varieties to obtain spectral information of diseased plants at different growth stages and different parts. Tethered hot air balloons conduct macroscopic monitoring of the entire rice field at high altitude to obtain overall information on rice growth status and disease distribution over a large area. The two work together to achieve all-round and multi-level data collection. S22. Establish a unified data integration framework to correlate and integrate the collected geographic information, soil data, climate data, and field management data with the multi- / hyperspectral information of rice blast diseased plants. To address the noise problem in multi- / hyperspectral images, use an adaptive filtering denoising algorithm to automatically adjust the filtering parameters according to the local features of the image to effectively remove noise interference. Use a radiation correction method based on a physical model, combined with field-measured atmospheric parameters, to accurately correct the radiation values ​​received by the sensor and restore the true spectral reflectance characteristics of rice. For image enhancement, use a combination of nonlinear stretching and histogram equalization to enhance the contrast between the characteristics of the diseased plants and the background, highlight the disease details, improve data quality, and provide a reliable basis for subsequent analysis. S23. Based on the principle of multi-scale analysis and feature fusion, the wavelet transform is used to decompose the image into sub-bands of different scales and directions, and the high-resolution detail information collected by the UAV and the macro information obtained by the tethered hot air balloon are layered and fused to retain the global features and local details of the image. The gray-level co-occurrence matrix in the UAV image is extracted to calculate the texture parameters, the shape characteristics of the spectral curve, and the vegetation index and color characteristics in the tethered hot air balloon image. The principal component analysis method is used for feature fusion to construct a comprehensive feature vector, giving full play to the advantages of the UAV and tethered hot air balloon data. S24. By comparing and analyzing the reflectance differences between healthy rice plants and diseased plants with different disease severity in each spectral band, combined with correlation analysis and stepwise discriminant analysis statistical methods, the bands with significant differences are screened out. Through verification and optimization of a large amount of sample data, the disease index that can most effectively characterize the condition of rice blast is determined, and a MySQL database is established to store rice variety information, spectral data, geographic soil climate field management data, disease information, and extracted key bands and disease index information.

4. The rice blast monitoring and diagnosis method according to claim 3, characterized in that: The step S30 also includes: S31, selecting representative spectral data of diseased rice plants and corresponding spectral data of healthy plants from the MySQL database as the basic data set for model construction, finely annotating the diseased plant data, and recording the location, shape, size and severity of the lesions in detail, and combining them with the comprehensive feature vector obtained in step S23 as the input parameters of the model; S32. Construct a deep learning model architecture based on self-supervised learning, using an encoder-decoder structure. The encoder part consists of multiple convolutional layers and pooling layers, which are used to encode the input multimodal feature vector, gradually extract abstract features in the data, and reduce the data dimension. The decoder part maps the encoded features back to the original data space through upsampling layers and deconvolution layers to achieve data reconstruction. Jump connections are introduced between the encoder and decoder to retain feature information at different levels and enhance the model's ability to capture detailed features. S33. The model adopts a self-supervised learning strategy for model training, and uses the intrinsic structural information of the data itself to design pre-training tasks. By randomly cropping, rotating, flipping and other transformations on the input image, the model is allowed to predict the relationship between the transformed image and the original image, thereby learning the invariant characteristics of the data. In the pre-training stage, a large amount of unlabeled data is used for training, so that the model can automatically discover the potential patterns and feature representations in the data; S34. After the pre-training is completed, the model is fine-tuned using the labeled data to optimize the model parameters to adapt to the rice blast disease diagnosis task. During the fine-tuning process, the cross entropy loss function is used as the optimization target of the model to measure the difference between the model prediction results and the true labels. At the same time, the stochastic gradient descent algorithm is combined to dynamically adjust the learning rate according to the distribution characteristics of the data to accelerate the model convergence and improve the generalization ability of the model. The model is evaluated using the validation set to monitor the performance of the model in terms of accuracy, recall rate, and F1 value indicators. When the model performance no longer improves, the training is stopped to prevent overfitting; S35. Establish a real-time data stream processing system to receive new data continuously collected by drones and tethered hot air balloons, perform real-time preprocessing on the new data, including data cleaning, format conversion, and feature extraction in the same training phase, and input the extracted dynamic features into the trained model to perform real-time rice blast monitoring and diagnosis; S36. In order to adapt to the dynamic changes in the rice growth process and the evolution of rice blast disease, a dynamic feature update mechanism is designed. As time goes by and data accumulates, the model is dynamically updated using an incremental learning method, allowing the model to continuously learn new data based on existing knowledge without retraining the entire model. When new data accumulates to a certain extent, the model is fine-tuned using the new data and some parameters of the model are updated, so that the model can promptly adapt to new changes and trends in rice blast disease; S37. In the process of model updating, combined with the idea of ​​reinforcement learning, a reward mechanism is introduced. Corresponding rewards are given according to the performance of the model in actual monitoring, which encourages the model to continuously optimize its own parameters and improve monitoring accuracy and reliability. At the same time, a model performance evaluation index system is established. In addition to the traditional accuracy and recall rate indicators, the timeliness and stability factors of the model are additionally considered to comprehensively evaluate the performance of the model. Targeted optimization and improvement are carried out based on the evaluation results to ensure that the constructed rice blast early diagnosis inversion model always maintains a good performance state and provides timely and effective disease warning and prevention and control decision support for rice production.

Images