A method and system for crop disease monitoring and damage assessment

Abstract

The present invention relates to the field of image recognition technology, and in particular to a method for crop disease monitoring and hazard assessment. The method comprises the following steps: obtaining a crop image set; screening the crop image set for growth stage images to obtain a crop growth stage image set; extracting crop growth morphological features from the crop growth stage image set to obtain crop growth morphological feature data; visualizing virtual crop health based on the crop growth morphological feature data to obtain a virtual crop growth health image set; obtaining crop growth environment data; matching the crop image set for abnormal growth of crop stages based on the virtual crop growth health image set to generate abnormal crop growth stage data. The present invention improves the accuracy and effectiveness of disease identification through multi-dimensional information acquisition, phased feature extraction, combination of environmental and internal disease monitoring, hierarchical risk assessment, and dynamic adjustment of prevention and control strategies.

Description

Technical Field

[0001] The present invention relates to the technical field of image recognition, and in particular to a method for monitoring crop diseases and assessing their degree of harm. Background Art

[0002] The introduction of remote sensing technology has provided a new means for crop disease monitoring. Through satellite and aerial images, researchers can obtain information on crop growth conditions over a large area and analyze it in combination with meteorological data. During this period, despite significant technological advances, quantitative assessment of diseases still relied on manual analysis, which was inefficient. After entering the 21st century, the rapid development of information technology has promoted the modernization of crop disease monitoring. In particular, the application of sensor technology, the Internet of Things and big data has made real-time monitoring possible. Researchers can collect data such as soil moisture, temperature and plant growth conditions through sensors installed in the field, and quickly identify and evaluate diseases in combination with machine learning algorithms. In recent years, the rise of deep learning has further improved the accuracy and efficiency of disease monitoring. Through image recognition technology, researchers can use crop images taken by mobile phones or drones to automatically identify diseases and assess their degree of damage. This method not only saves time and labor costs, but also provides important support for precision agriculture. However, the current traditional methods often cannot accurately match the growth stage of crops with the corresponding disease characteristics, resulting in difficulties in early identification of diseases. At the same time, the monitoring of environmental diseases and internal diseases is carried out separately, making it difficult to form an effective monitoring system, which results in low accuracy and effectiveness of disease identification. Summary of the invention

[0003] Based on this, it is necessary to provide a crop disease monitoring and damage assessment method to solve at least one of the above technical problems.

[0004] To achieve the above object, a method for monitoring crop diseases and assessing the degree of damage is provided, the method comprising the following steps:

[0005] Step S1: Acquire a crop image set; perform crop growth stage image screening on the crop image set to obtain a crop growth stage image set; perform crop growth morphological feature extraction on the crop growth stage image set to obtain crop growth morphological feature data; perform virtual crop health visualization based on the crop growth morphological feature data to obtain a virtual crop growth health image set;

[0006] Step S2: Acquire crop growth environment data; match crop stage growth abnormalities with the crop image set according to the virtual crop growth health image set to generate crop abnormal growth stage data; perform external environmental disease monitoring on the crop abnormal growth stage data to generate crop environmental disease monitoring data; perform step-by-step internal disease monitoring on the crop abnormal growth stage data based on the crop environmental disease monitoring data to generate crop internal disease monitoring data;

[0007] Step S3: Marking the crop image set as high-risk crops through the crop environment disease monitoring data and the crop internal disease monitoring data to generate high-risk crop marking data; marking the diseased crop screening image as propagation path crops to generate medium-risk crop marking data; identifying and marking the crop image set as low-risk crops through the high-risk crop marking data and the medium-risk crop marking data to generate low-risk crop marking data; dynamically adjusting the monitoring frequency according to the high-risk crop marking data, the medium-risk crop marking data and the low-risk crop marking data, thereby generating crop disease monitoring frequency adjustment data;

[0008] Step S4: Define hierarchical risk rules for diseased crop screening images according to crop disease monitoring frequency adjustment data to generate hierarchical risk rule data; generate prevention and control strategies for high-risk crop marking data, medium-risk crop marking data and low-risk crop marking data through the hierarchical risk rule data, thereby obtaining a multi-level crop disease prevention and control strategy.

[0009] The present invention establishes a rich data foundation by acquiring a crop image set, which provides reliable data support for subsequent analysis. Screening the crop image set for growth stages can help farmers focus on specific growth stages and ensure effective monitoring and management during critical periods. By extracting the morphological characteristics of crop growth, the health status of crops can be quantified, providing an objective basis for health assessment. Through virtual crop health visualization, farmers can intuitively understand the health status of crops and take timely intervention measures to promote the healthy growth of crops. Collecting crop growth environment data provides an environmental background for subsequent analysis and helps to understand the impact of the environment on crop growth. According to the virtual crop growth health image set, stage growth abnormality matching can be performed to identify growth abnormalities early and provide a time window for disease prevention and control. External environmental disease monitoring can timely identify external factors that affect crop health and provide a basis for disease prevention and control. Based on the stepped internal disease monitoring data of environmental disease monitoring, diseases can be managed in layers, thereby improving the accuracy and timeliness of disease detection. By analyzing the environmental disease monitoring data and the internal disease monitoring data, accurate marking of high-risk crops is achieved, which helps farmers give priority to crops with serious diseases. Through the crop marking of the propagation path, medium-risk crops are identified to provide early warning for the further spread of the disease. Through the systematic classification of high, medium and low risk markers, farmers are helped to optimize resource allocation, concentrate on dealing with high-risk areas, and reduce overall disease losses. Dynamic adjustment of monitoring frequency makes monitoring more flexible and efficient, and the monitoring strategy can be adjusted in real time according to the risk level to improve monitoring efficiency. Define hierarchical risk rules according to the monitoring frequency adjustment data to ensure that risk management measures are targeted and reduce unnecessary waste of resources. The multi-level crop disease prevention and control strategy generated by the hierarchical risk rule data enables farmers to implement differentiated management and take corresponding prevention and control measures for crops with different risk levels. The multi-level disease prevention and control strategy makes the prevention and control work more systematic and targeted, thereby improving the overall prevention and control efficiency and reducing the risk of crop losses. The construction of the entire monitoring and prevention and control system can promote the development of sustainable agriculture, help farmers achieve higher crop yields and quality, and reduce dependence on chemical control methods. Therefore, the present invention improves the accuracy and effectiveness of disease identification through multi-dimensional information acquisition, staged feature extraction, combination of environmental and internal disease monitoring, hierarchical risk assessment and dynamic adjustment of prevention and control strategies.

[0010] Preferably, step S1 comprises the following steps:

[0011] Step S11: using a camera to acquire a crop image set;

[0012] Step S12: performing image preprocessing on the crop image set to generate a standard crop image set, wherein the image preprocessing includes image cleaning, image filtering and image brightness enhancement;

[0013] Step S13: screening the crop growth stage images of the standard crop image set to obtain a crop growth stage image set;

[0014] Step S14: extracting crop growth morphological features from the crop growth stage image set to obtain crop growth morphological feature data; and performing virtual crop health visualization based on the crop growth morphological feature data to obtain a virtual crop growth health image set.

[0015] The present invention can realize comprehensive monitoring of crops in farmland by acquiring crop image sets through cameras (step S11), and can save the details of crop growth in digital form. This provides basic data support for subsequent image processing and analysis. Through image preprocessing (step S12), such as image cleaning, image filtering and image brightness enhancement, noise can be effectively removed, image clarity and brightness can be improved, and the accuracy of subsequent analysis can be ensured. The image set after standardization can better reflect the real state of the crop, which is conducive to subsequent growth stage analysis and feature extraction. By screening the crop growth stage images of the standard crop image set (step S13), images of different growth stages (such as germination, reproduction, maturity, etc.) can be accurately identified and extracted. This process can help researchers or farm managers understand the current growth progress of crops and provide a basis for staged management decisions. Through crop growth morphological feature extraction (step S14), the growth characteristics of key parts such as leaves and stems can be extracted from the image. These morphological feature data play an important role in analyzing the health status of crops, predicting growth trends and discovering abnormalities. Through virtual crop health visualization (step S14), the generated virtual crop growth health image set provides an intuitive health status display, allowing farm managers to intuitively observe the crop growth status. This visualization helps to quickly discover potential diseases or growth abnormalities and provide support for timely prevention and control measures.

[0016] Preferably, step S14 comprises the following steps:

[0017] Step S141: extracting three-dimensional morphological features from the crop growth stage image set to obtain three-dimensional morphological feature data of the crop, wherein the three-dimensional morphological feature extraction includes leaf spatial distribution extraction, stem three-dimensional structure extraction, and root underground development morphology extraction;

[0018] Step S142: Projecting the three-dimensional crop morphological feature data into a virtual three-dimensional space to generate a three-dimensional crop morphological feature matrix; performing morphological correction on the three-dimensional crop morphological feature matrix and a preset standard crop morphological feature matrix to generate a virtual crop growth morphological feature matrix;

[0019] Step S143: performing point cloud conversion on the virtual crop morphology feature matrix to generate virtual crop growth morphology point cloud data; constructing a crop growth three-dimensional model based on the virtual crop growth morphology point cloud data to generate a virtual crop growth three-dimensional model;

[0020] Step S144: Visualizing the growth of the virtual crop three-dimensional growth model at the healthy stage, thereby generating a set of virtual crop growth healthy images.

[0021] The present invention extracts three-dimensional morphological features of crop growth stage image sets, including the extraction of leaf spatial distribution, stem three-dimensional structure and root morphology, so that each key part of the crop growth morphology can be accurately modeled. This three-dimensional modeling is more comprehensive and detailed than traditional two-dimensional image analysis, and can reflect more subtle morphological changes in the crop growth process. The extracted three-dimensional morphological features are projected into a virtual three-dimensional space to generate a three-dimensional crop morphological feature matrix, and the matrix is ​​corrected with the standard crop morphological features. This step ensures a high degree of consistency between the virtual model and the actual growth state. The corrected virtual crop growth morphological feature matrix more accurately reflects the actual growth status of the crop, which is helpful for subsequent health monitoring and analysis. The virtual crop growth morphological feature matrix generates virtual three-dimensional point cloud data of the crop through point cloud conversion. This step enables complex crop morphological information to be expressed in the form of geometric points, which is convenient for subsequent three-dimensional model construction. Based on these point cloud data, a more accurate virtual crop growth three-dimensional model can be established, which further improves the ability to reproduce the details of the crop morphology. The generated virtual crop three-dimensional model can intuitively visualize the health status of the crop during growth. This function enables agricultural managers to observe the health status of crops at various growth stages in real time through crop models in a virtual environment, quickly discover potential diseases or growth abnormalities, and generate a set of virtual crop growth health images, which provides an important basis for the early detection of crop diseases. By visually displaying the health status of crops at different growth stages, farm managers can make response strategies faster and more effectively to reduce the spread of diseases and crop losses. The implementation of the entire step S14 provides a powerful digital means for crop health monitoring and management. Using three-dimensional models and visual displays of health status, farmers or agricultural experts can more comprehensively grasp the growth dynamics of crops, thereby optimizing planting and management strategies and improving yield and quality.

[0022] Preferably, step S2 comprises the following steps:

[0023] Step S21: using environmental sensors to obtain crop growth environment data;

[0024] Step S22: matching the standard crop image set with the virtual crop healthy growth image set to find out the abnormal growth of the crop stage, and generating the abnormal growth stage data of the crop, wherein the abnormal growth stage data of the crop includes the abnormal germination stage data of the crop, the abnormal reproductive stage data of the crop and the abnormal maturity stage data of the crop;

[0025] Step S23: Performing external environmental disease monitoring on the crop abnormal growth stage data through the crop growth environment data to generate crop environmental disease monitoring data;

[0026] Step S24: Based on the crop environmental disease monitoring data, the crop abnormal germination stage data, the crop abnormal reproduction stage data and the crop abnormal maturity stage data are subjected to step-by-step internal disease monitoring to generate crop internal disease monitoring data.

[0027] The present invention collects crop growth environment data by using environmental sensors to provide a basis for subsequent analysis. These environmental data (such as temperature, humidity, soil moisture content, etc.) have a direct impact on the crop growth process and can provide an objective basis for crop growth abnormality and disease monitoring. By matching the virtual crop growth health image set with the standard crop image set for abnormal growth, the generated crop abnormal growth stage data covers the germination, reproduction and maturity stages. This step enables the abnormal conditions of crops at different growth stages to be clearly identified, which helps to detect diseases at an early stage and avoid the deterioration of the overall growth state of crops. Environmental disease monitoring is performed on abnormal growth stage data through crop growth environment data. This step can effectively identify crop diseases caused by external environmental factors, such as diseases caused by excessively high temperature or low humidity. This process can generate crop environmental disease monitoring data to help farmers take preventive measures as soon as possible. Based on the environmental disease monitoring data, internal disease monitoring is further performed on each growth stage of crops, and the generated crop internal disease monitoring data can provide more detailed disease analysis. Through the step-by-step monitoring method, the severity and scope of the disease can be accurately located, and the risk of mild, moderate or severe diseases can be distinguished, thereby guiding different prevention and control strategies. The combination of external environmental disease monitoring and S24's internal disease monitoring can conduct all-round disease monitoring of the entire crop growth process. This disease monitoring system from the outside to the inside can greatly improve the effectiveness of disease prevention and provide more health protection for crops. The data chain generated by the entire S2 step, from crop environmental data to abnormal growth stage data, to environmental disease monitoring and internal disease monitoring, provides comprehensive decision-making support for crop disease prevention and control. Farmers and agricultural experts can quickly formulate targeted prevention and control measures based on the monitoring data generated at different stages, greatly improving the efficiency and accuracy of disease prevention and control.

[0028] Preferably, step S23 includes the following steps:

[0029] Step S231: extracting environmental change parameters from the crop abnormal growth stage data through the crop growth environment data to obtain environmental change parameter data;

[0030] Step S232: sorting the environmental change parameter data and the crop abnormal growth stage data to generate sorted environmental change parameter data and crop abnormal growth stage data; performing rank calculation on the sorted environmental change parameter data and crop abnormal growth stage data to obtain an environmental change rank value table and a crop abnormal growth rank value table;

[0031] Step S233: performing rank difference calculation on the environmental change rank value table and the crop abnormal growth rank value table to obtain environment-growth stage rank difference data; performing Spearman rank correlation calculation on the environment-growth stage rank difference data to generate environment-growth abnormality correlation detection coefficient;

[0032] Step S234: performing correlation detection on the crop growth environment data and the crop abnormal growth stage data using the environment-growth abnormality correlation detection coefficient to generate environment-growth abnormality correlation detection data; performing external environmental disease monitoring on the crop growth environment data according to the environment-growth abnormality correlation detection data to generate crop environmental disease monitoring data;

[0033] The calculation formula for Spearman rank correlation is as follows:

[0034]

[0035] In the formula, Expressed as the environment-growth anomaly association detection coefficient, Expressed as the sum of squares of the environment-growth stage rank differences, Expressed as the total number of data points.

[0036] The present invention can effectively discover the impact of environmental changes on abnormal crop growth stages by extracting environmental change parameter data and sorting and ranking it with crop abnormal growth stage data. The generated environment-growth abnormality correlation detection coefficient can quantitatively display the degree of correlation between environmental factors and crop growth abnormality through the Spearman rank association calculation formula, providing a scientific basis for crop management. The extracted environmental change parameter data include multiple environmental factors such as temperature, humidity, light intensity, soil moisture, etc. By sorting and rank difference calculation (S232) of these environmental data, the impact of each environmental factor on the crop growth stage at different time points can be decomposed to obtain detailed correlation information. This multi-dimensional analysis can help farmers or researchers better understand the changing trend of crop growth under complex environmental conditions. Using the environment-growth abnormality correlation detection coefficient, combined with crop growth environment data and crop abnormal growth stage data, correlation detection can be quickly performed to generate environment-growth abnormality correlation detection data. Through this data, farmers can accurately monitor the disease risk caused by the external environment, thereby achieving early intervention and avoiding large-scale damage to crops. Crop environmental disease monitoring data can provide an early warning system for disease prevention and control. For example, if the temperature or humidity rises abnormally during a certain period of time, and at the same time, there are abnormalities in the crop growth stage (such as leaf shrinkage or underdeveloped root system), the monitoring data can predict the possibility of disease occurrence in advance and remind growers to take preventive measures. According to the monitoring data, farmers can take targeted management measures, such as adjusting the amount of irrigation, changing the fertilization strategy, or using preventive drugs to avoid the adverse effects of environmental factors on crop health. This data-based management strategy reduces blindness and improves the efficiency and accuracy of crop management. Through the implementation of step S23, a comprehensive disease monitoring system can be provided for farmland, which not only detects crop abnormalities caused by environmental changes, but also discovers the potential relationship between environment and disease from rank difference association analysis, thereby helping farmers to adopt more effective disease prevention and control strategies and minimize losses.

[0037] Preferably, step S4 comprises the following steps:

[0038] Step S241: collecting images of the abnormal germination stage of crops based on the crop environmental disease monitoring data to obtain images of the abnormal germination stage of crops;

[0039] Step S242: performing a time series analysis of crop leaf and stem regions on the crop abnormal germination stage images to generate a crop leaf region time series image and a crop stem region time series image; performing growth feature extraction on the crop leaf region time series image and the crop stem region time series image to generate leaf growth data and stem growth data;

[0040] Step S243: calculating the growth rate of the leaf growth data and the stem growth data to obtain the growth rate data of the germination stage; monitoring the abnormal internal diseases of the crop during the germination stage according to the growth rate data of the germination stage, and generating the internal disease monitoring data of the germination stage;

[0041] Step S244: Based on the internal disease monitoring data of the budding stage, the abnormal reproductive stage data of the crop is subjected to abnormal fruit stalk development disease monitoring to generate the internal disease monitoring data of the reproductive stage;

[0042] Step S245: Based on the internal disease monitoring data of the reproductive stage, the abnormal crop nutrition supply disease monitoring data is carried out on the abnormal mature stage data of the crop to generate the internal disease monitoring data of the mature stage; the internal disease monitoring data of the germination stage, the internal disease monitoring data of the reproductive stage and the internal disease monitoring data of the mature stage are integrated to generate the internal disease monitoring data of the crop.

[0043] The present invention monitors the diseases of crops from germination to maturity, ensuring targeted disease analysis at each key stage of crop growth. In particular, in the three stages of germination, reproduction, and maturity that are crucial to crop health and yield, systematic internal disease monitoring can help to discover potential health problems early and avoid the spread of diseases. Through time-series image analysis, the growth dynamics of crop leaves and stems are monitored and growth feature data are extracted. This process can not only identify growth abnormalities, but also compare and calculate growth rates at multiple time points, provide accurate disease judgment basis, and ensure earlier intervention opportunities. The calculation of growth rate data in the germination stage (step S243) can intuitively reflect the health status of leaves and stems. For example, abnormal growth rates are often related to diseases or malnutrition, so accurate calculation of growth rates can improve the accuracy of disease detection and provide key data support for subsequent internal disease monitoring. Steps S244 and S245 ensure that disease monitoring at different growth stages can be linked to each other through correlation analysis of disease monitoring data. The discovery of internal diseases in the germination stage indicates abnormal development in the reproductive stage; similarly, diseases in the reproductive stage affect the subsequent maturity stage. Therefore, through multi-stage internal disease monitoring and integrated analysis, a more comprehensive crop health report can be generated to provide more effective guidance for disease prevention and control. Steps S244 and S245 focus on the monitoring of abnormal fruit stalk development and abnormal nutrient supply. Abnormal fruit stalk development usually directly affects fruit quality, while abnormal nutrient supply leads to defects in the crop maturity stage. By monitoring these key growth parts and functions, farmers can take timely intervention measures to prevent diseases from affecting the final crop yield and quality. The final step (S245) integrates the internal disease monitoring data of different growth stages to generate comprehensive crop internal disease monitoring data. This data integration helps farmers or agricultural managers to systematically grasp the health status of crops and support more scientific agricultural management and decision-making. Through the visualization and analysis of these data, it is easier to track the health status of crops and predict future disease risks.

[0044] Preferably, step S244 includes the following steps:

[0045] Step S2441: Based on the internal disease monitoring data of the budding stage, the crop abnormal reproduction stage data is subjected to crop abnormal reproduction stage image acquisition to obtain crop abnormal reproduction stage images; the crop abnormal reproduction stage images are subjected to fruit stalk part segmentation to generate crop fruit stalk part images;

[0046] Step S2442: performing fruit stalk discoloration region recognition on the fruit stalk part image of the crop to generate a fruit stalk discoloration region image; performing aperture detection on the fruit stalk discoloration region image to generate aperture detection data for the fruit stalk discoloration region; performing pixel brightness calculation on the fruit stalk discoloration region image using the aperture detection data for the fruit stalk discoloration region to obtain pixel brightness data for the fruit stalk discoloration region;

[0047] Step S2443: Compare the pixel brightness data of the fruit stalk discoloration area with a preset pixel brightness threshold. When the pixel brightness data of the fruit stalk discoloration area is greater than or equal to the preset pixel brightness threshold, the fruit stalk discoloration area image corresponding to the pixel brightness data of the fruit stalk discoloration area is marked as a fungal disease area image; when the pixel brightness data of the fruit stalk discoloration area is less than the preset pixel brightness threshold, the fruit stalk discoloration area image corresponding to the pixel brightness data of the fruit stalk discoloration area is marked as a rot disease area image;

[0048] Step S2444: Monitor the abnormal fruit stalk development of crops through fungal disease area images and rot disease area images of the abnormal reproductive stage data of crops, and generate internal disease monitoring data of the reproductive stage.

[0049] The present invention can accurately distinguish whether the discoloration of the fruit stem is caused by fungal diseases or rot diseases by aperture detection and pixel brightness calculation of the discoloration area of ​​the fruit stem. This accurate classification can not only improve the accuracy of disease monitoring, but also help farmers take more targeted prevention and control measures. For example, the prevention and control methods of fungal diseases and rot diseases are different, so accurate identification helps to improve the prevention and control efficiency. Through the aperture detection and pixel brightness calculation of the discoloration area of ​​the fruit stem, the optical characteristics are used to identify the disease, which is more efficient and objective than the traditional disease identification. Pixel brightness analysis can quickly detect the degree of lesions in abnormal areas, avoiding subjective errors caused by manual detection. The pixel brightness threshold comparison mechanism in step S2443 ensures the standardization and consistency of detection. By distinguishing between fungal diseases and rot diseases, the system can monitor multiple disease types at the same time, which greatly improves the extensiveness and comprehensiveness of disease monitoring. Especially in the reproductive stage, the fruit stem is an important part of fruit development, and its health status directly affects the fruit yield and quality. Therefore, it is very important to monitor different types of diseases in parallel to ensure the normal development of the fruit stem. By collecting and analyzing the image of the discolored area of ​​the fruit stalk in step S2441, abnormal development of the fruit stalk can be discovered early. This early warning mechanism can help farmers take preventive measures before the disease further spreads, avoid the disease affecting the key stage of fruit development, and improve the yield and quality of crops. In step S2443, the image of the discolored area of ​​the fruit stalk is clearly marked as a fungal disease area or a rot disease area through pixel brightness threshold comparison. This marking method provides a clear basis for subsequent prevention and control measures. Through accurate marking, farmers can intuitively understand the distribution of different diseases and make more reasonable prevention and control decisions. The entire process of step S244 from image collection to disease classification is completed by automated image processing and analysis technology. This automated technology can greatly improve monitoring efficiency, reduce the cost and error risk of manual intervention, and make disease monitoring in large-scale agricultural production more efficient and scalable.

[0050] Preferably, step S245 includes the following steps:

[0051] Step S2451: collecting images of the abnormally mature stage of crops based on the internal disease monitoring data of the reproductive stage, and obtaining images of the abnormally mature stage of crops;

[0052] Step S2452: performing root system image segmentation on the crop abnormal maturity stage image to generate a crop root system segmentation image; performing crop root system density calculation on the crop root system segmentation image to obtain crop root system density data; performing crop rooting penetration analysis on the crop root system segmentation image using the crop root system density data to generate crop rooting penetration data;

[0053] Step S2453: performing fruit part image segmentation on the crop abnormal maturity stage image to generate a crop fruit segmentation image; performing fruit morphological feature extraction on the crop fruit segmentation image to obtain crop fruit morphological feature data; performing fruit fullness analysis on the crop fruit segmentation image based on the crop fruit morphological feature data to generate crop fruit fullness data;

[0054] Step S2454: Based on the crop root penetration data and the crop fruit fullness data, the abnormal mature stage data of the crop is monitored for internal abnormal diseases in the mature stage to generate internal disease monitoring data for the mature stage; the internal disease monitoring data for the germination stage, the internal disease monitoring data for the reproductive stage and the internal disease monitoring data for the mature stage are integrated to generate crop internal disease monitoring data.

[0055] The present invention can effectively evaluate the density of crop roots and root penetration ability by segmenting and analyzing the root part image. This analysis helps to understand whether the root system of crops is healthy in the mature stage and whether it can effectively absorb water and nutrients. The health of the root system is crucial to the overall growth of crops, so this analysis helps farmers to find root problems early and prevent growth obstacles of crops in the mature stage. Through the fruit morphological feature extraction and fruit fullness analysis in step S2453, the development of the fruit can be accurately evaluated. The fullness of the fruit is an important indicator for measuring the quality of the fruit. This process not only helps to monitor the health of the fruit, but also provides data basis for the harvest time of the fruit. Farmers can decide whether to harvest in advance or supplement nutrition based on the analysis results of the fruit fullness. By combining the root penetration data and the fruit fullness data, disease monitoring in the mature stage can be effectively carried out. This process ensures a comprehensive evaluation of the root system and the fruit, making the internal disease monitoring more comprehensive. Abnormalities in the root system and the fruit will not only affect the crop yield, but also affect the subsequent storage and sales. Therefore, comprehensive disease monitoring is extremely important. Finally, by integrating the disease monitoring data of the germination stage, the reproductive stage and the maturity stage, a comprehensive crop internal disease monitoring data set is generated. This integration process provides farmers with disease reports across the entire growth cycle, allowing them to understand the health of crops more systematically and provide a reliable basis for making subsequent planting decisions. In particular, this data integration helps to identify whether there is a correlation between diseases at different growth stages, so as to better prevent and control them. Through image acquisition and analysis of key parts such as roots and fruits, step S245 provides real-time internal disease monitoring and early warning functions for crops. Once an abnormality is detected, the system can immediately prompt farmers to take protective measures to prevent the disease from spreading further. Especially when the crop is about to enter the maturity stage, any abnormality will directly affect the yield and quality, so a timely early warning mechanism is particularly important. Through root density and root penetration analysis, fruit fullness assessment, etc., farmers can manage the nutrient supply and field management measures of crops more scientifically. These analysis results provide farmers with accurate data on crops at the maturity stage, helping them adjust water and fertilizer management, thereby improving the final yield and quality of crops.

[0056] Preferably, step S3 comprises the following steps:

[0057] Step S31: screening diseased crops for the standard crop image set using the crop environment disease monitoring data and the crop internal disease monitoring data to obtain a diseased crop screening image; marking high-risk crops based on the diseased crop screening image to generate high-risk crop marking data;

[0058] Step S32: performing crop infection cluster analysis on the diseased crop screening image to generate crop infection cluster data; performing propagation path identification on the crop infection cluster data using high-risk crop marker data to generate crop disease risk propagation path data;

[0059] Step S33: using the crop disease risk propagation path data to mark medium-risk crops, and generating medium-risk crop marking data; using the high-risk crop marking data and the medium-risk crop marking data to identify and mark low-risk crops in the standard crop image set, and generating low-risk crop marking data;

[0060] Step S34: dynamically adjust the monitoring frequency according to the high-risk crop marking data, the medium-risk crop marking data and the low-risk crop marking data, thereby generating crop disease monitoring frequency adjustment data.

[0061] The present invention can screen out diseased crops from a standard crop image set through crop environmental disease monitoring data and crop internal disease monitoring data. This precise screening method ensures the early detection of diseased crops, thereby facilitating timely prevention and control measures to avoid the spread of diseases. In addition, the marking function of high-risk crops can effectively identify the most threatening crops, which is convenient for key monitoring and management. By clustering the screened diseased crops, the propagation pattern of crop diseases can be revealed. Infection cluster analysis not only helps to determine the source of the disease, but also reveals the infection relationship between crops, providing strong data support for subsequent prevention and control work. Through propagation path identification, farmers can understand the direction of disease spread, so as to formulate prevention and control plans in advance to prevent the spread of diseases. Based on the disease propagation path data, medium-risk crops are marked, and combined with the high-risk crop marking data, the risk classification management of crops can be effectively carried out. Through this multi-level risk classification, farmers can clearly distinguish high-risk, medium-risk and low-risk crops, and implement differentiated management strategies. High-risk crops require more frequent monitoring and timely prevention and control measures, while low-risk crops can reduce the monitoring frequency and optimize the use of resources. Step S34 provides a mechanism for dynamically adjusting the monitoring frequency according to the risk level of the crop. By combining the labeling data of high, medium and low risk crops, farmers can reasonably arrange the disease monitoring frequency for different crops. This not only improves the efficiency of disease monitoring, but also reduces the number of unnecessary monitoring times, thereby saving time and cost. Dynamically adjusting the disease monitoring frequency helps to respond to crop disease risks more flexibly and accurately. Step S3 can effectively prevent the spread of diseases by identifying the transmission path of crop diseases and marking risk classification. In particular, the marking and monitoring of high-risk crops enable farmers to take measures in advance to control the spread of diseases. In addition, the identification and marking of low-risk crops helps to concentrate resources on key areas of disease prevention and control and reduce the impact of large-scale diseases. Through disease screening, infection cluster analysis and risk classification management, step S3 provides a large amount of data support to help farmers make more scientific and effective decisions. Farmers can decide whether to increase or decrease the frequency of disease monitoring in a certain area based on the labeling information of various risky crops, thereby optimizing farmland management strategies.

[0062] Preferably, step S32 includes the following steps:

[0063] Step S321: performing crop edge detection on the diseased crop screening image to generate a diseased crop contour image; performing feature normalization on the diseased crop contour image to generate a diseased area feature vector; performing cluster center point calculation on the diseased area feature vector to generate diseased area cluster center point data;

[0064] Step S322: clustering the diseased crop screening image into crop infection clusters using the diseased area cluster center point data to generate crop infection cluster data; performing crop cluster interval distance analysis on the crop infection cluster data using the high-risk crop marker data to generate crop cluster interval distance data;

[0065] Step S323: Analyze the disease infection range of the crop infection cluster data to generate crop disease infection range data; identify the shortest impact range path of the crop cluster interval distance data based on the crop disease infection range data to generate crop disease risk transmission path data.

[0066] The present invention generates a contour image of a diseased crop by detecting crop edges, which can effectively separate the diseased area from the healthy area. This precise contour detection can help identify the specific location of the crop infected, providing a basis for subsequent disease cluster analysis. At the same time, the normalization of the feature vector of the diseased area makes the comparison between different diseased areas more consistent, thereby improving the accuracy of the disease analysis. The calculation of the cluster center point (step S321) provides core data for the subsequent cluster clustering analysis. In this way, farmers can quickly grasp the core position of the diseased area and infer the spread trend of the disease. This not only speeds up the monitoring speed of disease infection, but also provides data support for the formulation of disease control strategies. By clustering the diseased crops using the cluster center point data, crop infection cluster data is generated. This process helps farmers identify the concentrated areas of the disease, so as to better understand the pattern of disease transmission. By analyzing the crop infection clusters, farmers can more effectively divide the key areas for disease prevention and control, avoiding excessive prevention and control or waste of resources. By analyzing the interval distance of the crop infection cluster data (step S322), the distance between different disease clusters and the degree of mutual influence can be evaluated. This interval distance analysis helps to understand the distribution of diseases in the field and helps farmers determine whether the disease will spread between different areas, so as to formulate targeted prevention and control measures. By analyzing the disease infection range of crop infection cluster data, the spread range of the disease in the crop can be accurately estimated. The data on the disease infection range provides necessary information for the subsequent risk transmission path identification, ensuring that farmers can accurately predict the risk area of ​​disease transmission, so as to take early prevention and control measures to reduce the impact of the disease. Based on the disease infection range data and the crop cluster interval distance data, the shortest impact path of disease transmission is identified. This step helps farmers understand the disease transmission route and predict which areas are the next transmission targets of the disease. By identifying the shortest disease impact range path, farmers can take blocking measures in advance to prevent the rapid spread of the disease in crops. Step S32 not only realizes the accurate monitoring of disease infection, but also improves the efficiency of disease monitoring through means such as cluster clustering, infection range analysis and transmission path identification. Through rapid clustering and transmission path analysis, farmers can identify the potential spread trend of the disease at an early stage and reduce the losses caused in the later stage. By accurately analyzing the infection range of the disease and identifying the transmission path, step S32 helps farmers significantly reduce the risk of crop disease transmission. Farmers can quickly understand the spread dynamics of the disease by analyzing the cluster spacing and infection range, and thus formulate scientific prevention and control strategies. In this way, effective prevention and control can be carried out before the disease spreads, reducing the overall risk of crop losses.

[0067] The beneficial effect of the present invention is that by acquiring a crop image set and screening the growth stage, the relevance and accuracy of the data set are ensured, thereby laying a solid foundation for subsequent analysis. Growth morphological feature extraction and virtual health visualization help to quickly identify the growth status and potential problems of crops. By collecting crop growth environment data, external factors affecting crop health can be more comprehensively understood, and the ability to identify abnormal growth stages can be enhanced. Step-by-step internal disease monitoring combines external environmental monitoring with internal disease analysis, improves the early recognition rate and accuracy of diseases, and reduces losses. By combining environmental disease and internal disease monitoring data, high-risk crops can be quickly marked, which helps to take corresponding management measures in a timely manner. The marking and identification of medium-risk and low-risk crops enable resources to be reasonably allocated and improve monitoring efficiency. In addition, dynamic monitoring frequency adjustment ensures the effective use of resources and can optimize monitoring strategies according to real-time conditions. Through the definition of hierarchical risk rules, the risk of crop diseases can be systematically evaluated, thereby formulating more refined prevention and control strategies. This multi-level prevention and control strategy improves the management efficiency of crops with different risk levels, making prevention and control measures more targeted and effective, thereby reducing the risk of crop losses and improving the overall benefits of agricultural production. Therefore, the present invention improves the accuracy and effectiveness of disease identification through multi-dimensional information acquisition, staged feature extraction, combination of environmental and internal disease monitoring, hierarchical risk assessment and dynamic adjustment of prevention and control strategies. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] Figure 1 A schematic diagram of the steps of a crop disease monitoring and damage assessment method;

[0069] Figure 2 for Figure 1 Detailed implementation steps of step S2 in the flowchart;

[0070] Figure 3 for Figure 1 Detailed implementation steps of step S3 in FIG.

[0071] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0072] The technical method of the present invention is described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by technicians in this field without creative work are within the scope of protection of the present invention.

[0073] In addition, the accompanying drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor methods and / or microcontroller methods.

[0074] It should be understood that, although the terms "first", "second", etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are used only to distinguish one unit from another unit. For example, without departing from the scope of the exemplary embodiments, the first unit may be referred to as the second unit, and similarly the second unit may be referred to as the first unit. The term "and / or" used herein includes any and all combinations of one or more of the listed associated items.

[0075] To achieve this, please refer to Figures 1 to 3 , a crop disease monitoring and damage assessment method, the method comprising the following steps:

[0076] Step S1: Acquire a crop image set; perform crop growth stage image screening on the crop image set to obtain a crop growth stage image set; perform crop growth morphological feature extraction on the crop growth stage image set to obtain crop growth morphological feature data; perform virtual crop health visualization based on the crop growth morphological feature data to obtain a virtual crop growth health image set;

[0077] Step S2: Acquire crop growth environment data; match crop stage growth abnormalities with the crop image set according to the virtual crop growth health image set to generate crop abnormal growth stage data; perform external environmental disease monitoring on the crop abnormal growth stage data to generate crop environmental disease monitoring data; perform step-by-step internal disease monitoring on the crop abnormal growth stage data based on the crop environmental disease monitoring data to generate crop internal disease monitoring data;

[0078] Step S3: Marking the crop image set as high-risk crops through the crop environment disease monitoring data and the crop internal disease monitoring data to generate high-risk crop marking data; marking the diseased crop screening image as propagation path crops to generate medium-risk crop marking data; identifying and marking the crop image set as low-risk crops through the high-risk crop marking data and the medium-risk crop marking data to generate low-risk crop marking data; dynamically adjusting the monitoring frequency according to the high-risk crop marking data, the medium-risk crop marking data and the low-risk crop marking data, thereby generating crop disease monitoring frequency adjustment data;

[0079] Step S4: Define hierarchical risk rules for diseased crop screening images according to crop disease monitoring frequency adjustment data to generate hierarchical risk rule data; generate prevention and control strategies for high-risk crop marking data, medium-risk crop marking data and low-risk crop marking data through the hierarchical risk rule data, thereby obtaining a multi-level crop disease prevention and control strategy.

[0080] The present invention establishes a rich data foundation by acquiring a crop image set, which provides reliable data support for subsequent analysis. Screening the crop image set for growth stages can help farmers focus on specific growth stages and ensure effective monitoring and management during critical periods. By extracting the morphological characteristics of crop growth, the health status of crops can be quantified, providing an objective basis for health assessment. Through virtual crop health visualization, farmers can intuitively understand the health status of crops and take timely intervention measures to promote the healthy growth of crops. Collecting crop growth environment data provides an environmental background for subsequent analysis and helps to understand the impact of the environment on crop growth. According to the virtual crop growth health image set, stage growth abnormality matching can be performed to identify growth abnormalities early and provide a time window for disease prevention and control. External environmental disease monitoring can timely identify external factors that affect crop health and provide a basis for disease prevention and control. Based on the stepped internal disease monitoring data of environmental disease monitoring, diseases can be managed in layers, thereby improving the accuracy and timeliness of disease detection. By analyzing the environmental disease monitoring data and the internal disease monitoring data, accurate marking of high-risk crops is achieved, which helps farmers give priority to crops with serious diseases. Through the crop marking of the propagation path, medium-risk crops are identified to provide early warning for the further spread of the disease. Through the systematic classification of high, medium and low risk markers, farmers are helped to optimize resource allocation, concentrate on dealing with high-risk areas, and reduce overall disease losses. Dynamic adjustment of monitoring frequency makes monitoring more flexible and efficient, and the monitoring strategy can be adjusted in real time according to the risk level to improve monitoring efficiency. Define hierarchical risk rules according to the monitoring frequency adjustment data to ensure that risk management measures are targeted and reduce unnecessary waste of resources. The multi-level crop disease prevention and control strategy generated by the hierarchical risk rule data enables farmers to implement differentiated management and take corresponding prevention and control measures for crops with different risk levels. The multi-level disease prevention and control strategy makes the prevention and control work more systematic and targeted, thereby improving the overall prevention and control efficiency and reducing the risk of crop losses. The construction of the entire monitoring and prevention and control system can promote the development of sustainable agriculture, help farmers achieve higher crop yields and quality, and reduce dependence on chemical control methods. Therefore, the present invention improves the accuracy and effectiveness of disease identification through multi-dimensional information acquisition, staged feature extraction, combination of environmental and internal disease monitoring, hierarchical risk assessment and dynamic adjustment of prevention and control strategies.

[0081] In the embodiment of the present invention, reference Figure 1 The above is a schematic diagram of the steps of a crop disease monitoring and damage assessment method of the present invention. In this example, the crop disease monitoring and damage assessment method includes the following steps:

[0082] Step S1: Acquire a crop image set; perform crop growth stage image screening on the crop image set to obtain a crop growth stage image set; perform crop growth morphological feature extraction on the crop growth stage image set to obtain crop growth morphological feature data; perform virtual crop health visualization based on the crop growth morphological feature data to obtain a virtual crop growth health image set;

[0083] In an embodiment of the present invention, a crop image set is obtained by using a high-definition camera, a drone or a satellite image to ensure that the obtained image has good resolution and clarity. Crop images are regularly collected at different growth stages (such as germination, seedlings, heading, maturity, etc.) to establish a comprehensive image library. Image acquisition is ensured under good lighting and weather conditions to reduce the impact of uneven lighting and shadows. Gaussian filtering or median filtering is used to denoise the image to eliminate noise in the image and improve image quality. Color correction is performed on the image to ensure that the images collected under different environments are consistent in color. According to the growth stage of the crop, the crop images are classified using an image processing algorithm (such as image segmentation), and images that meet specific growth stages (such as germination, vegetative growth, reproductive growth, etc.) are screened out to generate a crop growth stage image set. Edge detection, texture analysis, shape analysis and other methods are used to extract crop growth morphological characteristics, such as leaf width, stem height, number of leaves, etc. The extracted features are classified and analyzed using a machine learning algorithm (such as a support vector machine, a random forest, etc.) to obtain crop growth morphological feature data. Standardize the extracted morphological feature data for subsequent analysis and visualization. Based on the crop growth morphological feature data, use 3D modeling software (such as Blender, SketchUp) to create a virtual crop model to reflect the morphology of the crop at different growth stages. Generate health indicators (such as growth rate, leaf photosynthetic capacity) based on the growth morphological feature data, and map these indicators to different parts of the virtual model, which are manifested as color changes, transparency adjustments, etc. Use virtual reality (VR) or augmented reality (AR) technology to display a set of virtual crop growth health images, so that users can intuitively observe the growth status and health of the crops. Provide user interaction functions, allowing users to select crops at different growth stages for observation and analysis, and generate dynamic health visualization effects.

[0084] Step S2: Acquire crop growth environment data; match crop stage growth abnormalities with the crop image set according to the virtual crop growth health image set to generate crop abnormal growth stage data; perform external environmental disease monitoring on the crop abnormal growth stage data to generate crop environmental disease monitoring data; perform step-by-step internal disease monitoring on the crop abnormal growth stage data based on the crop environmental disease monitoring data to generate crop internal disease monitoring data;

[0085] In an embodiment of the present invention, by arranging environmental sensors (such as temperature, humidity, light intensity, soil moisture and other sensors) in the crop growth area, crop growth environment data is monitored in real time. Collect various types of environmental data, including meteorological data (temperature, humidity, precipitation), soil data (pH value, EC value, nutrient content), etc. Store the collected environmental data in a database to ensure data integrity and traceability. Update environmental data regularly to ensure the latest environmental change information. Select crop growth indicators (such as leaf number, plant height, fruit development, etc.) based on a virtual crop growth health image set. Establish a crop growth model and use historical data and virtual image sets for training to generate growth standards based on growth stages. Use image processing algorithms (such as morphological operations, feature matching, etc.) to analyze the crop image set, match the virtual crop growth health image set, identify abnormal growth, and generate crop abnormal growth stage data. Determine environmental factors for monitoring diseases, such as temperature fluctuations, humidity changes, light conditions, etc., to evaluate their impact on crop growth. Use data analysis algorithms (such as time series analysis and machine learning) to process environmental data, identify potential disease risks, and generate crop environmental disease monitoring data. Establish a disease early warning model, compare environmental disease monitoring data with historical disease occurrence data, set early warning thresholds, and achieve real-time early warning. According to the crop growth stage and environmental disease monitoring data, set a step-by-step monitoring strategy, including preliminary monitoring, in-depth analysis, and final diagnosis. Integrate abnormal growth stage data with environmental disease monitoring data to generate a comprehensive monitoring data set. Use image-based disease detection algorithms (such as deep learning models, convolutional neural networks, etc.) to analyze the integrated data, identify internal diseases, generate crop internal disease monitoring data, mark affected crop areas, and assess the extent of the disease.

[0086] Step S3: Marking the crop image set as high-risk crops through the crop environment disease monitoring data and the crop internal disease monitoring data to generate high-risk crop marking data; marking the diseased crop screening image as propagation path crops to generate medium-risk crop marking data; identifying and marking the crop image set as low-risk crops through the high-risk crop marking data and the medium-risk crop marking data to generate low-risk crop marking data; dynamically adjusting the monitoring frequency according to the high-risk crop marking data, the medium-risk crop marking data and the low-risk crop marking data, thereby generating crop disease monitoring frequency adjustment data;

[0087] In an embodiment of the present invention, crop environmental disease monitoring data and crop internal disease monitoring data are collected. The collected data is cleaned, including removing outliers and filling missing values ​​to ensure data quality. Based on the environmental and internal disease monitoring data, a crop risk assessment model is constructed to determine the risk threshold (such as disease incidence, environmental suitability, etc.). The crop image set is analyzed, and high-risk crops are marked according to the model output, and high-risk crop marking data is generated. The propagation path analysis is performed on the diseased crop screening image to identify the potential path of disease transmission between crops. Cluster analysis, graph theory algorithms and other methods are applied to identify the propagation path of crop infection, thereby generating medium-risk crop marking data. The medium-risk crops obtained by propagation path identification are marked and integrated into the crop image set to form medium-risk crop marking data. The high-risk crop marking data and the medium-risk crop marking data are compared to identify crops that are not marked as high-risk and medium-risk. By eliminating high-risk and medium-risk crops, low-risk crops are identified from the crop image set, and low-risk crop marking data is generated. Update low-risk crop labeling data to the crop image set together with high-risk and medium-risk data to form a complete risk labeling system. Set dynamic adjustment standards for monitoring frequency based on high-risk, medium-risk, and low-risk crop labeling data. High-risk crops should be monitored frequently, while low-risk crops can be monitored less frequently. Use machine learning algorithms (such as reinforcement learning) to dynamically adjust monitoring frequency to adapt to environmental changes and crop disease development trends. Generate crop disease monitoring frequency adjustment data based on the adjusted monitoring frequency and output it to the monitoring system to achieve intelligent management of crop disease monitoring.

[0088] Step S4: Define hierarchical risk rules for diseased crop screening images according to crop disease monitoring frequency adjustment data to generate hierarchical risk rule data; generate prevention and control strategies for high-risk crop marking data, medium-risk crop marking data and low-risk crop marking data through the hierarchical risk rule data, thereby obtaining a multi-level crop disease prevention and control strategy.

[0089] In the embodiment of the present invention, by collecting crop disease monitoring frequency adjustment data, diseased crop screening images and corresponding risk marking data (high risk, medium risk, low risk). The collected data is cleaned and standardized as necessary to ensure data consistency and accuracy. Crop risks are hierarchically divided according to the degree of disease, and the standards of high risk, medium risk and low risk are defined. Based on statistical analysis and expert knowledge, hierarchical risk rules are generated. For example: high-risk crops: high disease transmission rate, poor environmental suitability, and disease incidence exceeding a certain threshold. Medium-risk crops: medium disease transmission rate, general environmental suitability, and disease incidence close to the threshold. Low-risk crops: low disease transmission rate, good environmental suitability, and disease incidence below the threshold. The generated hierarchical risk rules are recorded in the form of data to form hierarchical risk rule data, which provides a basis for the subsequent generation of prevention and control strategies. According to the defined hierarchical risk rules, corresponding prevention and control strategies are formulated for crops of different risk levels. For example: high-risk crop prevention and control strategies: strengthen monitoring frequency, apply targeted pesticides, set up physical barriers, etc. Medium-risk crop prevention and control strategies: regular monitoring, application of preventive pesticides, education and training, etc. Low-risk crop control strategies: maintain routine monitoring, optimize the growing environment, etc. Use decision trees, rule engines and other algorithms to automatically generate control strategies based on hierarchical risk rules. Digitalize the generated control strategies and integrate them into the crop disease management system. Integrate the control strategies for high-risk, medium-risk and low-risk crops to form a complete multi-level crop disease control strategy. Record the multi-level control strategies in the database and apply them in the crop management system for the convenience of farmers and managers.

[0090] Preferably, step S1 comprises the following steps:

[0091] Step S11: using a camera to acquire a crop image set;

[0092] Step S12: performing image preprocessing on the crop image set to generate a standard crop image set, wherein the image preprocessing includes image cleaning, image filtering and image brightness enhancement;

[0093] Step S13: screening the crop growth stage images of the standard crop image set to obtain a crop growth stage image set;

[0094] Step S14: extracting crop growth morphological features from the crop growth stage image set to obtain crop growth morphological feature data; and performing virtual crop health visualization based on the crop growth morphological feature data to obtain a virtual crop growth health image set.

[0095] In the embodiment of the present invention, a high-resolution digital camera or a camera mounted on a drone is selected to ensure image quality. The camera used should have strong light sensing capabilities to adapt to different lighting conditions. The position and angle of the camera are configured to ensure coverage of the entire crop planting area. It is recommended to use a combination of a fixed camera and a mobile drone to maximize the comprehensiveness of data collection. A detailed data collection plan is formulated to determine the shooting time, frequency and cycle, especially high-frequency shooting at key growth stages (such as sowing, germination, flowering, and maturity). The shooting plan is adjusted according to climate change to capture changes in growth status caused by climate change. Image acquisition is performed to ensure that crop images from multiple angles are captured at each predetermined time point. Timestamps and weather conditions should be recorded for each shot for subsequent analysis. Scripts are written using image processing software (such as OpenCV) to remove background noise and irrelevant objects from the image. Threshold segmentation and region growing methods can be used to separate crops from the background. Random noise in the image is reduced by technical means such as denoising algorithms (such as wavelet transform). Smoothing filtering techniques, such as Gaussian filtering and median filtering, are applied to smooth the image to reduce the risk of detail loss. In specific operations, you can set an appropriate filter kernel size to balance clarity and noise suppression. Combine edge detection algorithms (such as the Canny algorithm) to enhance the visibility of important features in the image, ensuring that the image can better identify the structure of the crop in subsequent processing. Use histogram equalization technology to improve the brightness and contrast of the image to make the image details more obvious. CLAHE (contrast-limited adaptive histogram equalization) can be used to avoid over-enhancement. Perform brightness adjustment and gamma correction to ensure the readability of the image under different lighting conditions. Identify the different stages of crop growth, such as germination, seedling, jointing, flowering, and maturity, and establish a standardized classification of crop growth stages. Use timestamps to classify images and assign them to different growth stages. Machine learning algorithms (such as SVM or decision trees) can be combined to train models to automatically identify growth stages in images. Combine image content analysis techniques, such as feature point detection and matching, to ensure that images at different stages can be accurately identified and classified. Organize and annotate images at each growth stage to form a crop growth stage image set and save it as a structured database for subsequent analysis and visualization.

[0096] Preferably, step S14 comprises the following steps:

[0097] Step S141: extracting three-dimensional morphological features from the crop growth stage image set to obtain three-dimensional morphological feature data of the crop, wherein the three-dimensional morphological feature extraction includes leaf spatial distribution extraction, stem three-dimensional structure extraction, and root underground development morphology extraction;

[0098] Step S142: Projecting the three-dimensional crop morphological feature data into a virtual three-dimensional space to generate a three-dimensional crop morphological feature matrix; performing morphological correction on the three-dimensional crop morphological feature matrix and a preset standard crop morphological feature matrix to generate a virtual crop growth morphological feature matrix;

[0099] Step S143: performing point cloud conversion on the virtual crop morphology feature matrix to generate virtual crop growth morphology point cloud data; constructing a crop growth three-dimensional model based on the virtual crop growth morphology point cloud data to generate a virtual crop growth three-dimensional model;

[0100] Step S144: Visualizing the growth of the virtual crop three-dimensional growth model at the healthy stage, thereby generating a set of virtual crop growth healthy images.

[0101] In an embodiment of the present invention, the crop image is processed by using an image segmentation algorithm (such as U-Net based on deep learning) to separate the leaf area. The geometric information of the leaves in the two-dimensional image is converted into three-dimensional data by three-dimensional reconstruction technology, the position and angle of the leaves in three-dimensional space are recorded, and a spatial distribution model of the leaves is generated. The outline of the stem is identified by edge detection and contour extraction technology. The shape and position of the stem can be determined by algorithms such as Hough transform. A three-dimensional model of the stem is generated by combining the known height and perimeter parameters with a geometric modeling method, and its features such as diameter, length and angle are recorded. The three-dimensional structural data of the root system is obtained by a non-destructive detection method (such as X-ray imaging or ground penetrating radar). The acquired data is analyzed using image processing technology to extract the distribution, depth and structural information of the root system to form a three-dimensional model of the root system. According to the extracted three-dimensional morphological feature data, a three-dimensional morphological feature matrix is ​​constructed, in which each feature point contains its coordinates (X, Y, Z) in three-dimensional space and related morphological parameters (such as color, size, etc.). These feature data are projected in three dimensions using data visualization software (such as MATLAB or Blender) to form a preliminary three-dimensional crop morphological feature matrix. The generated three-dimensional crop morphological feature matrix is ​​compared with the preset standard crop morphological feature matrix, and the least squares method and other algorithms are used for morphological correction to identify and adjust the differences between morphological features. Through morphological correction, a virtual crop growth morphological feature matrix is ​​generated to ensure that it conforms to the actual crop growth characteristics. The virtual crop growth morphological feature matrix is ​​converted into point cloud data. The spatial coordinates and related attributes (such as color, density, etc.) of each feature point are converted into a point in the point cloud. Point cloud processing libraries (such as PCL) can be used for data processing and conversion to ensure the high quality and accuracy of the point cloud. Based on the generated virtual crop growth morphological point cloud data, a surface reconstruction algorithm (such as Poisson surface reconstruction or triangular mesh reconstruction) is used to construct a three-dimensional model of the crop. Ensure that the three-dimensional model contains complete morphological information and adjust the level of detail of the model according to actual needs. Combined with key parameters of crop growth (such as growth rate, nutritional status, pests and diseases, etc.), the health status of the virtual crop growth three-dimensional model is evaluated. Using the existing health assessment model, different health states are classified by algorithm (such as health, water shortage, pests and diseases, etc.). Use visualization tools (such as Unity or Blender) to dynamically render the 3D model and generate visualization effects in different health states. Use animation, color change or morphological transformation to show the growth changes of crops at different health stages and generate a set of virtual crop growth health images.

[0102] As an example of the present invention, refer to Figure 2 As shown, in this example, step S2 includes:

[0103] Step S21: using environmental sensors to obtain crop growth environment data;

[0104] Step S22: matching the standard crop image set with the virtual crop healthy growth image set to find out the abnormal growth of the crop stage, and generating the abnormal growth stage data of the crop, wherein the abnormal growth stage data of the crop includes the abnormal germination stage data of the crop, the abnormal reproductive stage data of the crop and the abnormal maturity stage data of the crop;

[0105] Step S23: Performing external environmental disease monitoring on the crop abnormal growth stage data through the crop growth environment data to generate crop environmental disease monitoring data;

[0106] Step S24: Based on the crop environmental disease monitoring data, the crop abnormal germination stage data, the crop abnormal reproduction stage data and the crop abnormal maturity stage data are subjected to step-by-step internal disease monitoring to generate crop internal disease monitoring data.

[0107] In an embodiment of the present invention, suitable environmental sensors, such as temperature and humidity sensors, soil moisture sensors, light sensors, and CO2 sensors, are selected and deployed in the crop growth area. Ensure that the location of the sensors can fully cover different areas where crops grow for continuous monitoring. Set the sampling frequency of the sensor to ensure real-time acquisition of environmental data (such as every minute, every hour, etc.). Use the Internet of Things (IoT) technology to transmit the collected environmental data to the central processing unit through a wireless network (such as LoRa, Wi-Fi, etc.). Establish a database in the central processing unit to store the collected environmental data, including timestamps, temperature, humidity, soil moisture, light intensity, and CO2 concentration. Perform data cleaning to remove outliers and missing values ​​to ensure the accuracy and integrity of the data. Use the virtual crop growth healthy image set generated in step S14 to establish a standard health model for each growth stage of the crop, including morphological features under normal growth conditions. Apply image processing algorithms (such as template matching, feature point matching, etc.) to analyze the standard crop image set and identify differences from the virtual crop growth healthy image set. Perform anomaly detection through a deep learning model (such as a convolutional neural network) to mark abnormal stages in the crop growth process. According to the matching results, the data of abnormal crop growth stages are classified and generated, including: data of abnormal crop germination stage: images of crops with poor germination or delayed germination are identified. Data of abnormal crop reproductive stage: images of poor fruit development or abnormal flower buds in the reproductive period are detected. Data of abnormal crop maturity stage: images of rotten or dried fruits in the maturity period are identified. According to the environmental parameters of crop growth and historical disease data, a machine learning model (such as random forest or support vector machine) is established to predict environmental disease risks. Indicators for environmental disease monitoring are defined, such as the effects of temperature, humidity, soil moisture, light intensity, etc. on the occurrence of diseases. Data obtained by environmental sensors are input into the established model for analysis to evaluate the disease risks caused by the crop growth environment, generate crop environmental disease monitoring data, and record the disease risk level, disease type and corresponding environmental parameters. Combine the data of abnormal crop growth stage and environmental disease monitoring data to establish a multi-level disease monitoring model. Set monitoring indicators, including physiological characteristics of crops (such as leaf yellowing, wilting, etc.) and environmental influencing factors. According to the crop growth stage, disease monitoring is carried out for abnormal germination, abnormal reproduction and abnormal maturity of crops respectively, forming three independent monitoring systems. Comprehensive analysis is carried out on the data of each stage, combined with environmental disease monitoring data, to determine the internal disease risk. The monitoring results of each stage are summarized to generate crop internal disease monitoring data, recording the disease type, disease severity and impact range of each stage. Targeted management and prevention measures are formulated based on the monitoring data to improve the healthy level of crop growth.

[0108] Preferably, step S23 includes the following steps:

[0109] Step S231: extracting environmental change parameters from the crop abnormal growth stage data through the crop growth environment data to obtain environmental change parameter data;

[0110] Step S232: sorting the environmental change parameter data and the crop abnormal growth stage data to generate sorted environmental change parameter data and crop abnormal growth stage data; performing rank calculation on the sorted environmental change parameter data and crop abnormal growth stage data to obtain an environmental change rank value table and a crop abnormal growth rank value table;

[0111] Step S233: performing rank difference calculation on the environmental change rank value table and the crop abnormal growth rank value table to obtain environment-growth stage rank difference data; performing Spearman rank correlation calculation on the environment-growth stage rank difference data to generate environment-growth abnormality correlation detection coefficient;

[0112] Step S234: performing correlation detection on the crop growth environment data and the crop abnormal growth stage data using the environment-growth abnormality correlation detection coefficient to generate environment-growth abnormality correlation detection data; performing external environmental disease monitoring on the crop growth environment data according to the environment-growth abnormality correlation detection data to generate crop environmental disease monitoring data;

[0113] The calculation formula for Spearman rank correlation is as follows:

[0114]

[0115] In the formula, Expressed as the environment-growth anomaly association detection coefficient, Expressed as the sum of squares of the environment-growth stage rank differences, Expressed as the total number of data points.

[0116] In an embodiment of the present invention, crop growth environment data collected by sensors, including temperature, humidity, soil moisture, light intensity, etc., are recorded in a database. Data on abnormal crop growth stages are collected, including data on abnormal germination, abnormal reproduction, and abnormal maturity stages. Statistical analysis is performed on the crop growth environment data to extract key environmental change parameters. For example, the change amplitude and volatility of each environmental factor are calculated to generate environmental change parameter data, and the change value of each parameter and its timestamp are recorded for subsequent analysis. The extracted environmental change parameter data and the crop abnormal growth stage data are sorted by variables. Each group of data is sorted from small to large using a sorting algorithm (such as quick sort, merge sort, etc.) to generate a sorted environmental change parameter data table and a crop abnormal growth stage data table. A rank value is assigned to each sorted data. The rank value starts from 1 and increases in sequence. For example, the smallest value after sorting corresponds to rank 1, the second smallest value corresponds to rank 2, and so on. An environmental change rank value table and a crop abnormal growth rank value table are generated, and the rank value of each data point is recorded. The environmental change rank value table and the crop abnormal growth rank value table are compared one by one, and the rank difference of each pair of data is calculated, which is recorded as . Calculate the square of the rank difference And summing, we get . Use the formula for the Spearman rank correlation coefficient to calculate: in: Expressed as the environment-growth anomaly association detection coefficient, Expressed as the sum of squares of the environment-growth stage rank differences, Expressed as the total number of data points. The calculated rank difference data and Spearman rank correlation coefficient are saved as data records for subsequent analysis. The correlation between crop growth environment data and crop abnormal growth stage data is tested. The value of determines the strength of the correlation: Usually, A value close to 1 indicates a high positive correlation, close to -1 indicates a high negative correlation, and close to 0 indicates no correlation. Based on the results of the correlation test, the environment-growth anomaly correlation detection data is generated, and the correlation coefficient and its corresponding environmental change parameters and growth stage information are recorded. Based on the generated environment-growth anomaly correlation detection data, a comprehensive analysis of the crop growth environment data is performed to monitor potential disease risks, generate crop environmental disease monitoring data, and record the types of diseases monitored, the risk of occurrence, and possible preventive measures.

[0117] Preferably, step S4 comprises the following steps:

[0118] Step S241: collecting images of the abnormal germination stage of crops based on the crop environmental disease monitoring data to obtain images of the abnormal germination stage of crops;

[0119] Step S242: performing a time series analysis of crop leaf and stem regions on the crop abnormal germination stage images to generate a crop leaf region time series image and a crop stem region time series image; performing growth feature extraction on the crop leaf region time series image and the crop stem region time series image to generate leaf growth data and stem growth data;

[0120] Step S243: calculating the growth rate of the leaf growth data and the stem growth data to obtain the growth rate data of the germination stage; monitoring the abnormal internal diseases of the crop during the germination stage according to the growth rate data of the germination stage, and generating the internal disease monitoring data of the germination stage;

[0121] Step S244: Based on the internal disease monitoring data of the budding stage, the abnormal reproductive stage data of the crop is subjected to abnormal fruit stalk development disease monitoring to generate the internal disease monitoring data of the reproductive stage;

[0122] Step S245: Based on the internal disease monitoring data of the reproductive stage, the abnormal crop nutrition supply disease monitoring data is carried out on the abnormal mature stage data of the crop to generate the internal disease monitoring data of the mature stage; the internal disease monitoring data of the germination stage, the internal disease monitoring data of the reproductive stage and the internal disease monitoring data of the mature stage are integrated to generate the internal disease monitoring data of the crop.

[0123] In an embodiment of the present invention, the characteristics and symptoms of the abnormal germination stage of the crop are confirmed based on the crop environmental disease monitoring data to determine suitable image acquisition conditions. Use a high-resolution camera to collect images of the crop in the abnormal germination stage. Multiple acquisitions should be performed at different time points to ensure the timeliness and comprehensiveness of the data. The collected images should include close-ups of leaves, stems and other growth parts for subsequent analysis. The collected images of the abnormal germination stage of the crop are processed and arranged in a time series. The image changes in the leaf area and the stem area are analyzed to generate crop leaf area time series images and crop stem area time series images. Growth features are extracted from the time series images, including leaf shape, color changes, stem height, diameter, etc., to generate leaf growth data and stem growth data, and record the corresponding growth feature parameters and their changing trends. Based on the extracted leaf growth data and stem growth data, the growth rate in the germination stage is calculated. The following formula can be used: in, is the growth rate, For the moment The growth length, is the initial length, is the time interval. According to the growth rate data of the budding stage, analyze the abnormal fluctuation of the growth rate, judge the diseases existing inside the crop, generate the internal disease monitoring data of the budding stage, and record the disease type and its impact on crop growth. Based on the internal disease monitoring data of the budding stage, conduct data analysis of the reproductive stage. Collect crop images of the reproductive stage, analyze the development of the fruit stalk, and judge whether there is any abnormality. Record the abnormal development of the fruit stalk and the disease type, and generate the internal disease monitoring data of the reproductive stage. Use the internal disease monitoring data of the reproductive stage to analyze the crops in the mature stage and monitor the nutritional supply of the crops. Observe the maturity of the crops and their response to environmental changes to judge whether there are problems such as insufficient nutritional supply. Record the disease types and influencing factors in the mature stage, and generate the internal disease monitoring data of the mature stage. Integrate the internal disease monitoring data of the budding stage, the internal disease monitoring data of the reproductive stage, and the internal disease monitoring data of the mature stage to form a comprehensive crop internal disease monitoring data. Record the disease type, severity, impact range and recommended treatment measures of the integrated data for subsequent agricultural management decisions.

[0124] Preferably, step S244 includes the following steps:

[0125] Step S2441: Based on the internal disease monitoring data of the budding stage, the crop abnormal reproduction stage data is subjected to crop abnormal reproduction stage image acquisition to obtain crop abnormal reproduction stage images; the crop abnormal reproduction stage images are subjected to fruit stalk part segmentation to generate crop fruit stalk part images;

[0126] Step S2442: performing fruit stalk discoloration region recognition on the fruit stalk part image of the crop to generate a fruit stalk discoloration region image; performing aperture detection on the fruit stalk discoloration region image to generate aperture detection data for the fruit stalk discoloration region; performing pixel brightness calculation on the fruit stalk discoloration region image using the aperture detection data for the fruit stalk discoloration region to obtain pixel brightness data for the fruit stalk discoloration region;

[0127] Step S2443: Compare the pixel brightness data of the fruit stalk discoloration area with a preset pixel brightness threshold. When the pixel brightness data of the fruit stalk discoloration area is greater than or equal to the preset pixel brightness threshold, the fruit stalk discoloration area image corresponding to the pixel brightness data of the fruit stalk discoloration area is marked as a fungal disease area image; when the pixel brightness data of the fruit stalk discoloration area is less than the preset pixel brightness threshold, the fruit stalk discoloration area image corresponding to the pixel brightness data of the fruit stalk discoloration area is marked as a rot disease area image;

[0128] Step S2444: Monitor the abnormal fruit stalk development of crops through fungal disease area images and rot disease area images of the abnormal reproductive stage data of crops, and generate internal disease monitoring data of the reproductive stage.

[0129] In an embodiment of the present invention, the characteristics of the abnormal reproductive stage of the crop are determined based on the internal disease monitoring data of the bud stage, so as to carry out targeted image acquisition. Use high-definition camera equipment to collect images of the abnormal reproductive stage of the crop, and ensure that the lighting conditions are good to improve the image quality. The collected images should include close-ups of the fruit stalk part for subsequent analysis and processing. The collected images of the abnormal reproductive stage of the crop are processed, and the fruit stalk part is extracted using an image segmentation algorithm (such as edge detection, region growing, etc.) to generate an image of the fruit stalk part of the crop. The image of the fruit stalk part is subjected to color analysis, and the contrast of the discolored area is enhanced using color space conversion (such as HSV color space). The discolored area of ​​the fruit stalk is identified by threshold segmentation technology, and an image of the discolored area of ​​the fruit stalk is generated. In the image of the discolored area of ​​the fruit stalk, an aperture detection algorithm (such as Hough transform) is applied to identify and extract the aperture features of the discolored area, and the aperture detection data of the discolored area of ​​the fruit stalk is generated. According to the detected aperture features, the pixel brightness of the discolored area of ​​the fruit stalk is calculated to obtain the pixel brightness data of the discolored area of ​​the fruit stalk. The brightness can be calculated using the following formula: in, is the average brightness of the area, is the number of pixels, is the pixel point with the horizontal coordinate , the vertical axis is The brightness value at the location. A preset pixel brightness threshold is set to distinguish the brightness characteristics of fungal diseases and rot diseases. The threshold can be adjusted according to historical data or expert experience. The pixel brightness data of the fruit stalk discoloration area is compared with the preset pixel brightness threshold: if the pixel brightness data of the fruit stalk discoloration area is greater than or equal to the preset pixel brightness threshold, the area is marked as a fungal disease area image. If the pixel brightness data of the fruit stalk discoloration area is less than the preset pixel brightness threshold, the area is marked as a rot disease area image. By analyzing the images marked as fungal disease areas and rot disease areas, the abnormal disease situation of the fruit stalk development of the abnormal reproductive stage data of the crop is evaluated. Combined with the fruit stalk development situation and disease markings, the internal disease monitoring data of the reproductive stage is generated, including information such as disease type, occurrence site, and disease severity.

[0130] Preferably, step S245 includes the following steps:

[0131] Step S2451: collecting images of the abnormally mature stage of crops based on the internal disease monitoring data of the reproductive stage, and obtaining images of the abnormally mature stage of crops;

[0132] Step S2452: performing root system image segmentation on the crop abnormal maturity stage image to generate a crop root system segmentation image; performing crop root system density calculation on the crop root system segmentation image to obtain crop root system density data; performing crop rooting penetration analysis on the crop root system segmentation image using the crop root system density data to generate crop rooting penetration data;

[0133] Step S2453: performing fruit part image segmentation on the crop abnormal maturity stage image to generate a crop fruit segmentation image; performing fruit morphological feature extraction on the crop fruit segmentation image to obtain crop fruit morphological feature data; performing fruit fullness analysis on the crop fruit segmentation image based on the crop fruit morphological feature data to generate crop fruit fullness data;

[0134] Step S2454: Based on the crop root penetration data and the crop fruit fullness data, the abnormal mature stage data of the crop is monitored for internal abnormal diseases in the mature stage to generate internal disease monitoring data for the mature stage; the internal disease monitoring data for the germination stage, the internal disease monitoring data for the reproductive stage and the internal disease monitoring data for the mature stage are integrated to generate crop internal disease monitoring data.

[0135] In an embodiment of the present invention, the characteristics of the abnormal maturity stage of the crop are determined based on the internal disease monitoring data of the reproductive stage, so as to carry out targeted image acquisition. Use high-definition camera equipment to collect images of the abnormal maturity stage of the crop to ensure the clarity and detail integrity of the image. The collected images should include close-ups of the roots and fruits for subsequent analysis and processing. The collected images of the abnormal maturity stage of the crop are processed, and the root parts are extracted using image segmentation algorithms (such as threshold segmentation, regional growth, etc.) to generate crop root segmentation images. In the root segmentation image, the number and distribution of roots are calculated, and the crop root density data is obtained by the following formula: in is the root density, is the number of root pixels, is the area of ​​the image. Based on the crop root density data, the crop root segmentation image is analyzed for root penetration, and crop root penetration data is generated to evaluate the root system's ability to absorb soil nutrients and water. The image of the abnormal maturity stage of the crop is processed to extract the fruit part and generate the crop fruit segmentation image to ensure the integrity of the fruit features. In the fruit segmentation image, the morphological feature data of the fruit is extracted, including the diameter, circumference, color and other indicators of the fruit. These features can be used to analyze the maturity state of the fruit. Based on the crop fruit morphological feature data, the fullness of the fruit is evaluated and the crop fruit fullness data is generated. The fruit fullness can be calculated using the following formula: Fruit fullness = actual fruit volume / ideal fruit volume (can be determined based on variety data) × 100%; Based on the crop root penetration data and the crop fruit fullness data, combined with the previous reproductive stage internal disease monitoring data, the crop abnormal maturity stage data is monitored for internal abnormal diseases, and the mature stage internal disease monitoring data is generated. The disease situation is judged by analyzing the rooting and fruit status. The internal disease monitoring data of the budding stage, the reproductive stage and the mature stage are integrated to form a complete crop internal disease monitoring data. The integrated results will provide a comprehensive assessment basis for the health status of crops and facilitate the formulation of corresponding management and prevention strategies.

[0136] As an example of the present invention, refer to Figure 3 As shown, in this example, step S3 includes:

[0137] Step S31: screening diseased crops for the standard crop image set using the crop environment disease monitoring data and the crop internal disease monitoring data to obtain a diseased crop screening image; marking high-risk crops based on the diseased crop screening image to generate high-risk crop marking data;

[0138] Step S32: performing crop infection cluster analysis on the diseased crop screening image to generate crop infection cluster data; performing propagation path identification on the crop infection cluster data using high-risk crop marker data to generate crop disease risk propagation path data;

[0139] Step S33: using the crop disease risk propagation path data to mark medium-risk crops, and generating medium-risk crop marking data; using the high-risk crop marking data and the medium-risk crop marking data to identify and mark low-risk crops in the standard crop image set, and generating low-risk crop marking data;

[0140] Step S34: dynamically adjust the monitoring frequency according to the high-risk crop marking data, the medium-risk crop marking data and the low-risk crop marking data, thereby generating crop disease monitoring frequency adjustment data.

[0141] In an embodiment of the present invention, a standard crop image set is collected and sorted, including healthy crop images and diseased crop images. Crop environmental disease monitoring data and crop internal disease monitoring data are integrated for screening. A machine learning algorithm (such as a convolutional neural network (CNN)) is used to train the standard crop image set to identify the characteristics of diseased crops. The trained model is used to predict the standard crop image set, screen out diseased crops, and generate diseased crop screening images. According to the diseased crop screening images, a high-risk threshold (such as disease severity, transmission probability, etc.) is set, and the screened diseased crops are marked to generate high-risk crop marking data. Spatial clustering analysis is performed on the diseased crop screening images, and crop infection cluster data is generated using a clustering algorithm (such as DBSCAN or K-means). The cluster center and cluster range are identified to determine the concentrated area of ​​infection. Based on the high-risk crop marking data, the relationship between the infection clusters is analyzed, and the disease transmission path is identified using a network analysis method (such as the shortest path algorithm), and the crop disease risk transmission path data is generated, and the transmission path and impact range are displayed in a visual form. Using the crop disease risk transmission path data, determine the crops adjacent to or close to high-risk crops, set the medium-risk threshold, and generate medium-risk crop labeling data. By analyzing the standard crop image set and combining the high-risk and medium-risk crop labeling data, identify the unaffected crops and generate low-risk crop labeling data. This step uses machine learning classification algorithms (such as random forests or support vector machines SVM) for risk assessment and labeling. According to the distribution of high-risk crop labeling data, medium-risk crop labeling data, and low-risk crop labeling data, evaluate whether the current monitoring frequency is reasonable. Set dynamic adjustment rules, such as high-risk crops need to be monitored more frequently, while low-risk crops can reduce the monitoring frequency. Based on the evaluation results, generate crop disease monitoring frequency adjustment data, including monitoring frequency recommendations for various crops. Implement a dynamic monitoring strategy and adjust the monitoring frequency in real time to ensure the effectiveness of disease monitoring and the rational allocation of resources.

[0142] Preferably, step S32 includes the following steps:

[0143] Step S321: performing crop edge detection on the diseased crop screening image to generate a diseased crop contour image; performing feature normalization on the diseased crop contour image to generate a diseased area feature vector; performing cluster center point calculation on the diseased area feature vector to generate diseased area cluster center point data;

[0144] Step S322: clustering the diseased crop screening image into crop infection clusters using the diseased area cluster center point data to generate crop infection cluster data; performing crop cluster interval distance analysis on the crop infection cluster data using the high-risk crop marker data to generate crop cluster interval distance data;

[0145] Step S323: Analyze the disease infection range of the crop infection cluster data to generate crop disease infection range data; identify the shortest impact range path of the crop cluster interval distance data based on the crop disease infection range data to generate crop disease risk transmission path data.

[0146] In an embodiment of the present invention, the diseased crop screening image is processed by using the Canny edge detection algorithm or the Sobel operator to extract the edge information of the diseased crop and generate the contour image of the diseased crop. The contour image is binarized to enhance the edge features for subsequent analysis. The contour image features of the diseased area are normalized to generate a diseased area feature vector. This step ensures that the feature data is within the same scale range for effective comparison and analysis. The normalization method can use minimum-maximum normalization or Z-score standardization. The K-means clustering algorithm is used to process the diseased area feature vector to calculate the center point data of the cluster. An appropriate K value (number of clusters) is selected to improve the clustering effect, and the center point data of the diseased area cluster is generated for subsequent infection cluster cluster analysis. The diseased crop screening image is clustered into crop infection clusters using the calculated center point data of the diseased area cluster to generate crop infection cluster data. The clustering method can use hierarchical clustering or DBSCAN to identify the cluster distribution of crop infection. Based on the high-risk crop marking data, the distance between crop infection clusters is analyzed to generate crop cluster distance data. The distance between clusters is calculated using Euclidean distance or Manhattan distance to evaluate the relative position between infected areas. The crop infection cluster data is analyzed to identify the infection range of the disease and generate crop disease infection range data. The bounding box or contour area calculation method can be used to determine the impact area of ​​each infection cluster. Based on the crop disease infection range data, the crop cluster distance data is analyzed to identify the shortest impact range path and generate crop disease risk transmission path data. Graph theory algorithms (such as Dijkstra algorithm) are used to find the shortest transmission path between clusters to ensure rapid identification of potential infection transmission channels.

[0147] The beneficial effect of the present invention is that by acquiring a crop image set and screening the growth stage, the relevance and accuracy of the data set are ensured, thereby laying a solid foundation for subsequent analysis. Growth morphological feature extraction and virtual health visualization help to quickly identify the growth status and potential problems of crops. By collecting crop growth environment data, external factors affecting crop health can be more comprehensively understood, and the ability to identify abnormal growth stages can be enhanced. Step-by-step internal disease monitoring combines external environmental monitoring with internal disease analysis, improves the early recognition rate and accuracy of diseases, and reduces losses. By combining environmental disease and internal disease monitoring data, high-risk crops can be quickly marked, which helps to take corresponding management measures in a timely manner. The marking and identification of medium-risk and low-risk crops enable resources to be reasonably allocated and improve monitoring efficiency. In addition, dynamic monitoring frequency adjustment ensures the effective use of resources and can optimize monitoring strategies according to real-time conditions. Through the definition of hierarchical risk rules, the risk of crop diseases can be systematically evaluated, thereby formulating more refined prevention and control strategies. This multi-level prevention and control strategy improves the management efficiency of crops with different risk levels, making prevention and control measures more targeted and effective, thereby reducing the risk of crop losses and improving the overall benefits of agricultural production. Therefore, the present invention improves the accuracy and effectiveness of disease identification through multi-dimensional information acquisition, staged feature extraction, combination of environmental and internal disease monitoring, hierarchical risk assessment and dynamic adjustment of prevention and control strategies.

[0148] Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is therefore intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.

[0149] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A method for monitoring crop diseases and assessing the degree of damage, characterized in that: The following steps are involved: Step S1: Acquire a crop image set; perform crop growth stage image screening on the crop image set to obtain a crop growth stage image set; perform crop growth morphological feature extraction on the crop growth stage image set to obtain crop growth morphological feature data; perform virtual crop health visualization based on the crop growth morphological feature data to obtain a virtual crop growth health image set; Step S2: Acquire crop growth environment data; match crop stage growth abnormalities with the crop image set according to the virtual crop growth health image set to generate crop abnormal growth stage data; perform external environmental disease monitoring on the crop abnormal growth stage data to generate crop environmental disease monitoring data; perform step-by-step internal disease monitoring on the crop abnormal growth stage data based on the crop environmental disease monitoring data to generate crop internal disease monitoring data; Step S3: marking high-risk crops on the crop image set by using the crop environmental disease monitoring data and the crop internal disease monitoring data to generate high-risk crop marking data; Mark the crops along the diseased crop transmission path in the diseased crop screening images to generate medium-risk crop marking data; Identify and label low-risk crops in the crop image set using high-risk crop labeling data and medium-risk crop labeling data to generate low-risk crop labeling data; Dynamically adjust the monitoring frequency according to the high-risk crop marking data, the medium-risk crop marking data and the low-risk crop marking data, thereby generating crop disease monitoring frequency adjustment data; step S3 includes the following steps: Step S31: screening diseased crops for the standard crop image set using the crop environment disease monitoring data and the crop internal disease monitoring data to obtain a diseased crop screening image; marking high-risk crops based on the diseased crop screening image to generate high-risk crop marking data; Step S32: performing crop infection cluster analysis on the diseased crop screening image to generate crop infection cluster data; performing propagation path identification on the crop infection cluster data using high-risk crop marker data to generate crop disease risk propagation path data; Step S33: using the crop disease risk propagation path data to mark medium-risk crops, and generating medium-risk crop marking data; using the high-risk crop marking data and the medium-risk crop marking data to identify and mark low-risk crops in the standard crop image set, and generating low-risk crop marking data; Step S34: dynamically adjusting the monitoring frequency according to the high-risk crop marking data, the medium-risk crop marking data, and the low-risk crop marking data, thereby generating crop disease monitoring frequency adjustment data; Step S4: Define hierarchical risk rules for diseased crop screening images according to crop disease monitoring frequency adjustment data to generate hierarchical risk rule data; generate prevention and control strategies for high-risk crop marking data, medium-risk crop marking data and low-risk crop marking data through the hierarchical risk rule data, thereby obtaining a multi-level crop disease prevention and control strategy.

2. The crop disease monitoring and damage assessment method according to claim 1, characterized in that: Step S1 includes the following steps: Step S11: using a camera to acquire a crop image set; Step S12: performing image preprocessing on the crop image set to generate a standard crop image set, wherein the image preprocessing includes image cleaning, image filtering and image brightness enhancement; Step S13: screening the crop growth stage images of the standard crop image set to obtain a crop growth stage image set; Step S14: extracting crop growth morphological features from the crop growth stage image set to obtain crop growth morphological feature data; and performing virtual crop health visualization based on the crop growth morphological feature data to obtain a virtual crop growth health image set.

3. The crop disease monitoring and damage assessment method according to claim 2, characterized in that: Step S14 includes the following steps: Step S141: extracting three-dimensional morphological features from the crop growth stage image set to obtain three-dimensional morphological feature data of the crop, wherein the three-dimensional morphological feature extraction includes leaf spatial distribution extraction, stem three-dimensional structure extraction, and root underground development morphology extraction; Step S142: Projecting the three-dimensional crop morphological feature data into a virtual three-dimensional space to generate a three-dimensional crop morphological feature matrix; performing morphological correction on the three-dimensional crop morphological feature matrix and a preset standard crop morphological feature matrix to generate a virtual crop growth morphological feature matrix; Step S143: performing point cloud conversion on the virtual crop morphology feature matrix to generate virtual crop growth morphology point cloud data; constructing a crop growth three-dimensional model based on the virtual crop growth morphology point cloud data to generate a virtual crop growth three-dimensional model; Step S144: Visualizing the growth of the virtual crop three-dimensional growth model at the healthy stage, thereby generating a set of virtual crop growth healthy images.

4. The crop disease monitoring and damage assessment method according to claim 3, characterized in that: Step S2 includes the following steps: Step S21: using environmental sensors to obtain crop growth environment data; Step S22: matching the standard crop image set with the virtual crop healthy growth image set to find out the abnormal growth of the crop stage, and generating the abnormal growth stage data of the crop, wherein the abnormal growth stage data of the crop includes the abnormal germination stage data of the crop, the abnormal reproductive stage data of the crop and the abnormal maturity stage data of the crop; Step S23: Performing external environmental disease monitoring on the crop abnormal growth stage data through the crop growth environment data to generate crop environmental disease monitoring data; Step S24: Based on the crop environmental disease monitoring data, the crop abnormal germination stage data, the crop abnormal reproduction stage data and the crop abnormal maturity stage data are subjected to step-by-step internal disease monitoring to generate crop internal disease monitoring data.

5. The method for crop disease monitoring and damage assessment according to claim 4, characterized in that: Step S23 includes the following steps: Step S231: extracting environmental change parameters from the crop abnormal growth stage data through the crop growth environment data to obtain environmental change parameter data; Step S232: sorting the environmental change parameter data and the crop abnormal growth stage data to generate sorted environmental change parameter data and crop abnormal growth stage data; performing rank calculation on the sorted environmental change parameter data and crop abnormal growth stage data to obtain an environmental change rank value table and a crop abnormal growth rank value table; Step S233: performing rank difference calculation on the environmental change rank value table and the crop abnormal growth rank value table to obtain environment-growth stage rank difference data; performing Spearman rank correlation calculation on the environment-growth stage rank difference data to generate environment-growth abnormality correlation detection coefficient; Step S234: performing correlation detection on the crop growth environment data and the crop abnormal growth stage data using the environment-growth abnormality correlation detection coefficient to generate environment-growth abnormality correlation detection data; performing external environmental disease monitoring on the crop growth environment data according to the environment-growth abnormality correlation detection data to generate crop environmental disease monitoring data; The calculation formula for Spearman rank correlation is as follows: In the formula, Expressed as the environment-growth anomaly association detection coefficient, Expressed as the sum of squares of the environment-growth stage rank differences, Expressed as the total number of data points.

6. The method for crop disease monitoring and damage assessment according to claim 4, characterized in that: Step S4 includes the following steps: Step S241: collecting images of the abnormal germination stage of crops based on the crop environmental disease monitoring data to obtain images of the abnormal germination stage of crops; Step S242: performing a time series analysis of crop leaf and stem regions on the crop abnormal germination stage images to generate a crop leaf region time series image and a crop stem region time series image; performing growth feature extraction on the crop leaf region time series image and the crop stem region time series image to generate leaf growth data and stem growth data; Step S243: calculating the growth rate of the leaf growth data and the stem growth data to obtain the growth rate data of the germination stage; monitoring the abnormal internal diseases of the crop during the germination stage according to the growth rate data of the germination stage, and generating the internal disease monitoring data of the germination stage; Step S244: Based on the internal disease monitoring data of the budding stage, the abnormal reproductive stage data of the crop is subjected to abnormal fruit stalk development disease monitoring to generate the internal disease monitoring data of the reproductive stage; Step S245: Based on the internal disease monitoring data of the reproductive stage, the abnormal crop nutrition supply disease monitoring data is carried out on the abnormal mature stage data of the crop to generate the internal disease monitoring data of the mature stage; the internal disease monitoring data of the germination stage, the internal disease monitoring data of the reproductive stage and the internal disease monitoring data of the mature stage are integrated to generate the internal disease monitoring data of the crop.

7. The method for crop disease monitoring and damage assessment according to claim 6, characterized in that: Step S244 includes the following steps: Step S2441: Based on the internal disease monitoring data of the budding stage, the crop abnormal reproduction stage data is subjected to crop abnormal reproduction stage image acquisition to obtain crop abnormal reproduction stage images; the crop abnormal reproduction stage images are subjected to fruit stalk part segmentation to generate crop fruit stalk part images; Step S2442: performing fruit stalk discoloration region recognition on the fruit stalk part image of the crop to generate a fruit stalk discoloration region image; performing aperture detection on the fruit stalk discoloration region image to generate aperture detection data for the fruit stalk discoloration region; performing pixel brightness calculation on the fruit stalk discoloration region image using the aperture detection data for the fruit stalk discoloration region to obtain pixel brightness data for the fruit stalk discoloration region; Step S2443: Compare the pixel brightness data of the fruit stalk discoloration area with a preset pixel brightness threshold. When the pixel brightness data of the fruit stalk discoloration area is greater than or equal to the preset pixel brightness threshold, the fruit stalk discoloration area image corresponding to the pixel brightness data of the fruit stalk discoloration area is marked as a fungal disease area image; when the pixel brightness data of the fruit stalk discoloration area is less than the preset pixel brightness threshold, the fruit stalk discoloration area image corresponding to the pixel brightness data of the fruit stalk discoloration area is marked as a rot disease area image; Step S2444: Monitor the abnormal fruit stalk development of crops through fungal disease area images and rot disease area images of the abnormal reproductive stage data of crops, and generate internal disease monitoring data of the reproductive stage.

8. The method for crop disease monitoring and damage assessment according to claim 6, characterized in that: Step S245 includes the following steps: Step S2451: collecting images of the abnormally mature stage of crops based on the internal disease monitoring data of the reproductive stage, and obtaining images of the abnormally mature stage of crops; Step S2452: performing root system image segmentation on the crop abnormal maturity stage image to generate a crop root system segmentation image; performing crop root system density calculation on the crop root system segmentation image to obtain crop root system density data; performing crop rooting penetration analysis on the crop root system segmentation image using the crop root system density data to generate crop rooting penetration data; Step S2453: performing fruit part image segmentation on the crop abnormal maturity stage image to generate a crop fruit segmentation image; performing fruit morphological feature extraction on the crop fruit segmentation image to obtain crop fruit morphological feature data; performing fruit fullness analysis on the crop fruit segmentation image based on the crop fruit morphological feature data to generate crop fruit fullness data; Step S2454: Based on the crop root penetration data and the crop fruit fullness data, the abnormal mature stage data of the crop is monitored for internal abnormal diseases in the mature stage to generate internal disease monitoring data for the mature stage; the internal disease monitoring data for the germination stage, the internal disease monitoring data for the reproductive stage and the internal disease monitoring data for the mature stage are integrated to generate crop internal disease monitoring data.

9. The crop disease monitoring and damage assessment method according to claim 1, characterized in that: Step S32 includes the following steps: Step S321: performing crop edge detection on the diseased crop screening image to generate a diseased crop contour image; performing feature normalization on the diseased crop contour image to generate a diseased area feature vector; performing cluster center point calculation on the diseased area feature vector to generate diseased area cluster center point data; Step S322: clustering the diseased crop screening image into crop infection clusters using the diseased area cluster center point data to generate crop infection cluster data; performing crop cluster interval distance analysis on the crop infection cluster data using the high-risk crop marker data to generate crop cluster interval distance data; Step S323: Analyze the disease infection range of the crop infection cluster data to generate crop disease infection range data; identify the shortest impact range path of the crop cluster interval distance data based on the crop disease infection range data to generate crop disease risk transmission path data.

Images