Plant disease and insect pest prediction method and system based on image recognition

Abstract

The invention relates to the technical field of image recognition, and particularly discloses a plant disease and insect pest prediction method and system based on image recognition, and the method comprises the steps: analyzing a to-be-predicted plant growth condition corresponding to a qualified to-be-predicted plant growth condition image, and obtaining a to-be-predicted plant disease and insect pest characteristic factor, obtaining a to-be-predicted plant disease and insect pest microorganism constraint signal, a to-be-predicted plant disease and insect pest heavy metal constraint signal and a to-be-predicted plant disease and insect pest environment constraint factor, determining a to-be-predicted plant disease and insect pest deviation constraint signal, and obtaining a to-be-predicted plant disease and insect pest early warning grade. The problems that a traditional pest and disease monitoring and predicting method usually depends on manual observation and experience judgment, prediction is inaccurate, and dynamic adaptive capacity is insufficient are solved, the growth condition and pest and disease risk of plants are known more comprehensively, the efficiency and sustainability of agricultural production are improved, early symptoms of pests and diseases are helped to be rapidly recognized, and the economic benefit is increased. And early warning and prevention in advance are facilitated.

Description

Technical Field

[0001] The present invention relates to the technical field of image recognition, and in particular to a method and system for predicting plant diseases and insect pests based on image recognition. Background Art

[0002] In recent years, with the rapid development of deep learning and computer vision technology, image recognition technology has been widely used in various fields. Its application in agriculture, especially the automatic detection and identification of plant diseases and pests, has shown great potential and can achieve fast and accurate monitoring. By collecting plant image data, features can be extracted from a large amount of data to achieve early prediction of diseases and pests. Data-driven methods are more scientific and objective than traditional methods. By applying image recognition technology to predict diseases and pests, measures can be taken in advance to reduce the use of fertilizers and pesticides, increase crop yields, protect the ecological environment, and promote the sustainable development of agriculture. With the rise of smart agriculture, combined with technologies such as the Internet of Things and big data analysis, the disease and pest prediction method based on image recognition meets the needs of modern agricultural development, can achieve precision agriculture, and improve resource utilization efficiency.

[0003] Nowadays, there are still some shortcomings in the research on plant disease and pest prediction based on image recognition. Specifically, traditional disease and pest monitoring and prediction methods usually rely on manual observation and empirical judgment, which is not only time-consuming and labor-intensive, but also easily affected by subjective factors, resulting in inaccurate predictions and insufficient dynamic adaptability. Traditional methods are usually based on fixed experience, slow to respond to environmental changes, and unable to update prediction information in a timely manner, thus affecting the prevention and control effect. Traditional methods usually cannot achieve real-time monitoring of diseases and pests, resulting in insufficient early warning and rapid response capabilities, and it is difficult to quickly obtain comprehensive and accurate monitoring results. Summary of the invention

[0004] In view of the deficiencies in the prior art, the present invention provides a plant disease and insect pest prediction method and system based on image recognition, which can effectively solve the problems involved in the above-mentioned background technology.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions: In a first aspect, the present invention provides a method for predicting plant diseases and insect pests based on image recognition, comprising the following steps: obtaining an image of the growth status of the plant to be predicted, and judging whether the quality of the image of the growth status of the plant to be predicted is qualified; if the quality of the image of the growth status of the plant to be predicted is unqualified, re-acquiring the image of the growth status of the plant to be predicted; if the quality of the image of the growth status of the plant to be predicted is qualified, analyzing the growth status of the plant to be predicted, and obtaining characteristic factors of the plant diseases and insect pests to be predicted; analyzing the soil microbial community of the plant to be predicted to determine the microbial constraint signal of the plant diseases and insect pests to be predicted; analyzing the heavy metal characteristics of the soil to be predicted to determine the heavy metal constraint signal of the plant diseases and insect pests to be predicted; analyzing the planting environment of the plant to be predicted to obtain the environmental constraint factor of the plant diseases and insect pests to be predicted; combining the microbial constraint signal of the plant diseases and insect pests to be predicted, the heavy metal constraint signal of the plant diseases and insect pests to be predicted and the environmental constraint factor of the plant diseases and insect pests to be predicted, determining the deviation constraint signal of the plant diseases and insect pests to be predicted; based on the characteristic factors of the plant diseases and insect pests to be predicted and the deviation constraint signal of the plant diseases and insect pests to be predicted, obtaining the warning level of the plant diseases and insect pests to be predicted.

[0006] As a further method, it is determined whether the quality of the image of the growth condition of the plant to be predicted is qualified. The specific analysis process is: obtaining a data set of image quality of the growth condition of the plant to be predicted, the data set of image quality of the growth condition of the plant to be predicted specifically includes the shooting distance deviation rate of the growth condition image, the vertical resolution of the growth condition image, and the contrast deviation rate of the growth condition image; based on the acquired data set of image quality of the growth condition of the plant to be predicted, a comprehensive analysis is performed to obtain a growth condition image quality determination factor, the growth condition image quality determination factor is used as an analysis basis for determining whether the quality of the image of the growth condition of the plant to be predicted is qualified; the growth condition image quality determination factor is compared with the growth condition image quality determination threshold stored in the database; if the growth condition image quality determination factor is lower than the growth condition image quality determination threshold, the quality of the image of the growth condition of the plant to be predicted is unqualified, and the image of the growth condition of the plant to be predicted needs to be re-acquired; if the growth condition image quality determination factor is not lower than the growth condition image quality determination threshold, the quality of the image of the growth condition of the plant to be predicted is qualified.

[0007] As a further method, the growth status image quality determination factor, the specific analysis process is:

[0008] ;

[0009] In the formula, is the growth status image quality determination factor, is the shooting distance deviation rate of the growth status image, is the vertical resolution of the growth status image, is the growth status image contrast deviation rate, For setting The compensation factor, For setting The compensation factor, For setting compensation factor.

[0010] As a further method, the growth conditions of the plants to be predicted are analyzed to obtain characteristic factors of plant diseases and pests to be predicted. The specific analysis process is: obtaining a data set of plant growth conditions to be predicted, the data set of plant growth conditions to be predicted specifically includes the percentage of damaged leaf area of ​​the plants to be predicted, the percentage of damaged stem area of ​​the plants to be predicted, and the average density of leaf distribution of plant diseases and pests to be predicted; based on the acquired data set of plant growth conditions to be predicted, a comprehensive analysis is performed to obtain characteristic factors of plant diseases and pests to be predicted, and the characteristic factors of plant diseases and pests to be predicted are used as the analysis basis for obtaining the warning level of plant diseases and pests to be predicted.

[0011] As a further method, the characteristic factors of plant diseases and insect pests to be predicted, the specific analysis process is as follows:

[0012] ;

[0013] In the formula, are the characteristic factors of plant diseases and insect pests to be predicted, is the percentage of damaged area of ​​plant leaves to be predicted, is the percentage of damaged stem area of ​​the plant to be predicted, is the average density of the leaves of the plant pests and diseases to be predicted, For setting The compensation factor, For setting The compensation factor, For setting compensation factor.

[0014] As a further method, the microbial constraint signal of the plant diseases and insect pests to be predicted is determined. The specific analysis process is: obtaining the characteristic data set of the soil microbial community of the plant to be predicted, and the characteristic data set of the soil microbial community of the plant to be predicted specifically includes the deviation rate of the total microbial biomass of the soil and the deviation rate of the respiration rate of the soil; based on the obtained characteristic data set of the soil microbial community of the plant to be predicted, a comprehensive analysis is performed to obtain the microbial constraint signal of the plant diseases and insect pests to be predicted, and the microbial constraint signal of the plant diseases and insect pests to be predicted is used as the analysis basis for obtaining the warning level of the plant diseases and insect pests to be predicted.

[0015] As a further method, the heavy metal constraint signals of the plant diseases and insect pests to be predicted are determined. The specific analysis process is: obtaining the soil heavy metal characteristic data set of the plants to be predicted, the soil heavy metal characteristic data set of the plants to be predicted specifically includes the soil lead concentration, the soil cadmium concentration, and the soil mercury concentration; based on the obtained soil heavy metal characteristic data set of the plants to be predicted, a comprehensive analysis is performed to obtain the heavy metal constraint signals of the plant diseases and insect pests to be predicted, and the heavy metal constraint signals of the plant diseases and insect pests to be predicted are used as the analysis basis for obtaining the warning level of the plant diseases and insect pests to be predicted.

[0016] As a further method, the environmental constraint factors of the plant diseases and insect pests to be predicted are obtained. The specific analysis process is: obtain the planting environment data set of the plants to be predicted, and the planting environment data set of the plants to be predicted specifically includes the planting environment temperature deviation rate, the planting environment humidity deviation rate, and the planting rhizosphere environment oxygen concentration deviation rate; based on the obtained planting environment data set of the plants to be predicted, a comprehensive analysis is performed to obtain the environmental constraint factors of the plant diseases and insect pests to be predicted, and the environmental constraint factors of the plant diseases and insect pests to be predicted are used as the analysis basis for obtaining the warning level of the plant diseases and insect pests to be predicted.

[0017] As a further method, the warning level of the plant diseases and insect pests to be predicted is obtained. The specific analysis process is: the microbial constraint signal of the plant diseases and insect pests to be predicted, the heavy metal constraint signal of the plant diseases and insect pests to be predicted and the environmental constraint factor of the plant diseases and insect pests to be predicted are stored as designated tags, and the designated tags are compared with the deviation constraint signals of the plant diseases and insect pests to be predicted corresponding to each designated tag stored in the database to obtain the deviation constraint signals of the plant diseases and insect pests to be predicted corresponding to the designated tags; the characteristic factors of the plant diseases and insect pests to be predicted and the deviation constraint signals of the plant diseases and insect pests to be predicted are stored as designated codes, and the designated codes are compared with the warning signals of the plant diseases and insect pests to be predicted corresponding to each designated code stored in the database to obtain the warning signals of the plant diseases and insect pests to be predicted corresponding to the designated codes; the warning signals of the plant diseases and insect pests to be predicted are compared with the warning signals of the plant diseases and insect pests to be predicted stored in the database. If the warning signal of plant diseases and insect pests to be predicted falls within the constraint range of the warning signal of plant diseases and insect pests to be predicted, the warning level of plant diseases and insect pests to be predicted corresponding to the warning signal of plant diseases and insect pests to be predicted is level two; if the warning signal of plant diseases and insect pests to be predicted does not fall within the constraint range of the warning signal of plant diseases and insect pests to be predicted, and the warning signal of plant diseases and insect pests to be predicted is less than the minimum value in the constraint range of the warning signal of plant diseases and insect pests to be predicted, the warning level of plant diseases and insect pests to be predicted corresponding to the warning signal of plant diseases and insect pests to be predicted is level one; if the warning signal of plant diseases and insect pests to be predicted does not fall within the constraint range of the warning signal of plant diseases and insect pests to be predicted, and the warning signal of plant diseases and insect pests to be predicted is greater than the maximum value in the constraint range of the warning signal of plant diseases and insect pests to be predicted, the warning level of plant diseases and insect pests to be predicted corresponding to the warning signal of plant diseases and insect pests to be predicted is level three.

[0018] The second aspect of the present invention provides a plant pest and disease prediction system based on image recognition, including a pest and disease characteristic factor analysis module, a microbial constraint signal determination module, a heavy metal constraint signal determination module, an environmental constraint factor acquisition module, a deviation constraint signal determination module and a pest and disease warning level determination module, wherein: the pest and disease characteristic factor analysis module is used to obtain an image of the growth status of the plant to be predicted, and judge whether the quality of the image of the growth status of the plant to be predicted is qualified; if the quality of the image of the growth status of the plant to be predicted is unqualified, the image of the growth status of the plant to be predicted is re-acquired; if the quality of the image of the growth status of the plant to be predicted is qualified, the growth status of the plant to be predicted is analyzed to obtain the characteristic factors of the plant pest and disease to be predicted; the microbial constraint signal determination module is used to determine the soil microbial community of the plant to be predicted. The module is used to analyze the characteristics of the soil heavy metals of the plants to be predicted and determine the heavy metal constraint signals of the plants to be predicted; the module is used to analyze the planting environment of the plants to be predicted and obtain the environmental constraint factors of the plants to be predicted; the module is used to determine the deviation constraint signals of the plants to be predicted by combining the microbial constraint signals of the plants to be predicted, the heavy metal constraint signals of the plants to be predicted and the environmental constraint factors of the plants to be predicted; the module is used to determine the deviation constraint signals of the plants to be predicted by combining the microbial constraint signals of the plants to be predicted, the heavy metal constraint signals of the plants to be predicted and the environmental constraint factors of the plants to be predicted; the module is used to determine the warning level of the plants to be predicted based on the characteristic factors of the plants to be predicted and the deviation constraint signals of the plants to be predicted.

[0019] Compared with the prior art, the embodiments of the present invention have at least the following advantages or beneficial effects:

[0020] (1) The present invention provides a method and system for predicting plant diseases and pests based on image recognition. Through multiple steps, a comprehensive analysis of plant growth conditions, soil microorganisms, heavy metal characteristics and environmental factors is performed, so as to provide a more comprehensive understanding of plant growth conditions and their disease and pest risks. Quality inspection is performed to ensure that the image quality is qualified, which can improve the accuracy and reliability of subsequent analysis and reduce the possibility of misjudgment and missed judgment. Automated image quality judgment and multi-dimensional analysis reduce the subjectivity of manual observation and improve the objectivity and scientificity of prediction. Based on the results of data analysis, the efficiency and sustainability of agricultural production can be improved.

[0021] (2) The present invention obtains characteristic factors of plant diseases and insect pests to be predicted by analyzing the growth status of the plants to be predicted with qualified plant growth status images. Through the analysis of high-quality images, the significant characteristics of plant diseases and insect pests can be accurately captured, directly reflecting the health status of the plants and ensuring the clear presentation of details, so that the analysis algorithm can more accurately extract the characteristics of diseases and insect pests, thereby improving the accuracy of recognition and prediction and reducing misjudgment due to image blur or distortion. The extraction of characteristic factors can help quickly identify the early signs of diseases and insect pests, facilitating early warning and prevention. Through early detection and intervention, the damage to plants caused by diseases and insect pests can be reduced. The acquisition of characteristic factors through image analysis makes the disease and insect pest monitoring process more automated, greatly improving efficiency and reducing subjective bias.

[0022] (3) The present invention determines the deviation constraint signal of the plant pests and diseases to be predicted by combining the microbial constraint signal of the plant pests and diseases to be predicted, the heavy metal constraint signal of the plant pests and diseases to be predicted, and the environmental constraint factor of the plant pests and diseases to be predicted. It can more comprehensively identify and analyze the potential factors of pests and diseases. The occurrence of pests and diseases is usually the result of the superposition of multiple factors. Through multi-dimensional constraint signals, the source of pest and disease risk can be more accurately characterized. The generation of deviation constraint signals links the characteristic factors of pests and diseases with factors such as the environment, soil and heavy metals, which helps to more accurately identify the conditions and characteristics of the occurrence of pests and diseases, thereby improving the accuracy of prediction, avoiding misjudgment caused by a single factor, and enhancing early warning capabilities. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] The present invention is further described using the accompanying drawings, but the embodiments in the accompanying drawings do not constitute any limitation to the present invention. A person skilled in the art can obtain other drawings based on the following drawings without creative work.

[0024] Figure 1 The figure is a schematic flow chart of the method steps of the present invention.

[0025] Figure 2 It is a schematic diagram of system module connection of the present invention.

[0026] Figure 3 It is the analytical basis for judging whether the image quality of the plant growth condition to be predicted is qualified. DETAILED DESCRIPTION

[0027] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0028] Reference Figure 1 As shown, the first aspect of the present invention provides a plant disease and insect pest prediction method based on image recognition, including: obtaining an image of the growth status of the plant to be predicted, and judging whether the quality of the image of the growth status of the plant to be predicted is qualified; if the quality of the image of the growth status of the plant to be predicted is unqualified, re-obtaining the image of the growth status of the plant to be predicted; if the quality of the image of the growth status of the plant to be predicted is qualified, analyzing the growth status of the plant to be predicted, and obtaining characteristic factors of the plant diseases and insect pests to be predicted.

[0029] like Figure 3 As shown, the specific analysis process is: obtaining a data set of image quality of the growth status of the plant to be predicted, the data set of image quality of the growth status of the plant to be predicted specifically includes the shooting distance deviation rate of the growth status image, the vertical resolution of the growth status image, and the contrast deviation rate of the growth status image; based on the acquired data set of image quality of the growth status of the plant to be predicted, a comprehensive analysis is performed to obtain a growth status image quality determination factor, and the growth status image quality determination factor is used as an analysis basis for judging whether the quality of the image of the growth status of the plant to be predicted is qualified; comparing the growth status image quality determination factor with the growth status image quality determination threshold stored in the database; if the growth status image quality determination factor is lower than the growth status image quality determination threshold, the quality of the image of the growth status of the plant to be predicted is unqualified, and the image of the growth status of the plant to be predicted needs to be re-acquired; if the growth status image quality determination factor is not lower than the growth status image quality determination threshold, the quality of the image of the growth status of the plant to be predicted is qualified.

[0030] The shooting distance deviation rate refers to the deviation ratio between the actual distance when the image is taken and the preset ideal distance. The shooting distance can be measured by a camera equipped with a distance sensor (such as ultrasonic, infrared, laser rangefinder, etc.). The vertical resolution refers to the pixel density of the image in the vertical direction, that is, the number of pixels per unit height. The resolution value is usually directly determined by the parameters of the shooting equipment. The contrast deviation rate refers to the deviation ratio between the actual contrast of the image (that is, the difference in brightness and darkness of the image) and the ideal contrast. It can be obtained through a professional camera equipped with contrast measurement and adjustment functions. The deviation rate is the ratio of the absolute value of the actual value minus the reference value to the reference value.

[0031] The shooting distance deviation rate reflects the accuracy of the shooting distance. If the shooting distance deviates too much from the preset ideal distance, it will affect the image resolution, especially the vertical resolution. When the shooting distance is too close or too far, the contrast will also be affected. A distance that is too close may cause the image to be overexposed, resulting in a bright area and reduced contrast, while a distance that is too far may cause the image to be insufficiently bright, forming a dark area, which also affects the contrast. The image resolution and the contrast deviation rate jointly determine the clarity of the image details. High resolution can capture more details, but if the contrast deviation is too large (such as too low contrast resulting in a dark image or too high contrast resulting in overexposure of some areas), these details will be difficult to distinguish.

[0032] By using multi-dimensional data analysis such as shooting distance deviation rate, vertical resolution, contrast deviation rate, etc., image quality can be evaluated more accurately to ensure that all input images meet the standards and ensure the consistency and reliability of subsequent analysis. High-quality images can more clearly show the growth status and pest characteristics of plants, making the feature factors extracted by subsequent image analysis algorithms more accurate, which helps to improve the accuracy of pest prediction and classification and reduce misidentification caused by image quality differences. Unqualified image quality may lead to misjudgment or missed judgment in analysis. Through this set of image quality assessment processes, unqualified images can be filtered out, which can significantly reduce the possibility of misjudgment and missed judgment and improve the credibility of the entire pest monitoring system. If the image quality is unqualified but still enters the analysis stage, it will waste the system's computing resources and may lead to erroneous results. By screening unqualified images in advance, resource waste can be reduced and system operation efficiency can be improved. The overall efficiency can be improved through automated comparison of image quality judgment factors and judgment thresholds.

[0033] Furthermore, the growth status image quality determination factor, the specific analysis process is as follows:

[0034] ;

[0035] In the formula, is the growth status image quality determination factor, is the growth status image shooting distance deviation rate, is the vertical resolution of the growth status image, is the growth status image contrast deviation rate, For setting The compensation factor, For setting The compensation factor, For setting compensation factor.

[0036] It should be explained that the above growth status image quality determination factor is calculated by the growth status image shooting distance deviation rate, growth status image vertical resolution, and growth status image contrast deviation rate. , , Normalization processing is performed. The shooting distance deviation rate, vertical resolution and contrast deviation rate can ensure that the image meets consistent standards when shooting, avoid inconsistent image quality due to differences in distance, resolution and contrast, and ensure the reliability and comparability of subsequent data analysis. The distance deviation rate ensures the consistency of the shooting position and avoids crop size deviation caused by distance changes, thereby improving the measurement accuracy. The vertical resolution ensures image clarity, and the contrast deviation rate ensures image clarity and distinct layers, which facilitates detail recognition. High-quality images can be screened out and input into the image recognition algorithm, thereby improving the accuracy and stability of the model. The image quality has a direct impact on the recognition effect of the algorithm, especially in tasks such as detail judgment and lesion recognition. Deviations in image shooting distance, resolution and contrast will cause errors and noise in the image, affecting the accuracy of recognition. Calculating the image quality judgment factor can eliminate poor quality image data and reduce the misjudgment rate.

[0037] It should be explained that the above setting , , The compensation factor is obtained from the database, and the historical measured growth image shooting distance deviation rate, growth image vertical resolution, growth image contrast deviation rate and , , The mapping set of compensation factors is obtained , , Corresponding , , compensation factor.

[0038] It should be noted that the following , , , , , , , , They are also obtained through a mapping set of historical data and compensation factors established in a database, that is, the corresponding compensation factors are obtained according to the current data.

[0039] A data set of growth conditions of plants to be predicted is obtained, which specifically includes the percentage of damaged leaf area of ​​plants to be predicted, the percentage of damaged stem area of ​​plants to be predicted, and the average density of leaf distribution of plants with diseases and insect pests to be predicted; based on the obtained data set of growth conditions of plants to be predicted, characteristic factors of plant diseases and insect pests to be predicted are obtained through comprehensive analysis, and the characteristic factors of plant diseases and insect pests to be predicted are used as the analysis basis for obtaining the warning level of plant diseases and insect pests to be predicted.

[0040] The percentage of damaged leaf area refers to the proportion of the damaged area of ​​the leaf attacked by pests and diseases to the total area of ​​the leaf, usually expressed as a percentage. Use image processing algorithms (such as algorithms based on edge detection or color segmentation) to segment the leaf from the background, identify the damaged area by the color, texture features or brightness changes of the lesions, and compare the number of pixels in the damaged area with the total number of pixels on the leaf to get the percentage of damaged area. The percentage of damaged stem area refers to the proportion of the damaged area of ​​the plant stem attacked by pests and diseases to the total area of ​​the stem, expressed as a percentage. Use image segmentation technology to separate the stem from the image. Algorithms based on color and shape features or deep learning segmentation models can be used. Analyze the color changes, lesion shapes or texture features of the stems to identify the damaged areas. The damaged areas usually show obvious features such as discoloration, spots or rot. Calculate the proportion of the damaged area in the total stem area to obtain the proportion of the damaged area of ​​the stem. The average density of leaf distribution of pests and diseases refers to the distribution density of pests and diseases on the leaves, which is usually expressed by the number of lesions or damaged areas of pests and diseases per unit area. The leaf area is separated from the background through a segmentation algorithm, and the lesions are identified using features such as color, texture or shape. The distribution position and number of each lesion on the leaf are counted, and the total area of ​​the lesions or damaged areas is divided by the total leaf area to obtain the distribution density of pests and diseases per unit area.

[0041] The percentage of damaged leaf area and the percentage of damaged stem area can respectively reflect the spread and degree of damage of pests and diseases in different parts of the plant (leaves and stems). If these two indicators are high at the same time, it means that the pests and diseases have spread widely in different parts of the plant, and the overall health of the plant is under great threat. Damage to both leaves and stems often indicates that the pests and diseases have entered a serious stage, and immediate prevention and control measures may be needed. The average density of leaf distribution of pests and diseases describes the density of distribution of pests and diseases on leaves. An increase in leaf distribution density may be a precursor to the spread of pests and diseases to other parts (such as stems). The percentage of damaged leaf area and the average density of leaf distribution of pests and diseases can be used in combination to assess the severity of pests and diseases on leaves. The combination of high distribution density and high percentage of damaged area usually indicates that pests and diseases are extremely serious on leaves and may have affected the normal photosynthesis and metabolic functions of leaves.

[0042] Through specific indicators such as the percentage of damaged leaf area, the percentage of damaged stem area, and the average density of pest and disease distribution, the impact and risk of pests and diseases can be assessed more accurately, and the damage of pests and diseases to different parts of the plant can be intuitively reflected, thereby providing a more accurate basis for the formulation of early warning levels. Pest and disease characteristic factors are based on objective data from many aspects and can comprehensively assess the severity of pests and diseases, thereby improving the accuracy and reliability of the early warning system and reducing false alarms or missed reports. By tracking data such as changes in damaged leaves and stems of plants, the distribution density of pests and diseases, the system can dynamically monitor the development trend of pests and diseases, and provide a basis for continuous monitoring and adjustment of prevention and control strategies. Indicators such as the damaged area of ​​leaves and stems can help the system identify abnormalities in the early stages of pests and diseases, issue early warnings, and take prevention and control measures before pests and diseases spread to avoid greater damage to crops.

[0043] Furthermore, the specific analysis process of the characteristic factors of plant diseases and insect pests to be predicted is as follows:

[0044] ;

[0045] In the formula, are the characteristic factors of plant diseases and insect pests to be predicted, is the percentage of damaged area of ​​plant leaves to be predicted, is the percentage of damaged stem area of ​​the plant to be predicted, is the average density of the leaves of the plant pests and diseases to be predicted, For setting The compensation factor, For setting The compensation factor, For setting compensation factor.

[0046] It should be explained that the above-mentioned characteristic factors of plant diseases and insect pests to be predicted are calculated by the percentage of damaged area of ​​leaves of plants to be predicted, the percentage of damaged area of ​​stems of plants to be predicted, and the average density of leaf distribution of plant diseases and insect pests to be predicted. , , After normalization, the percentage of damaged area on leaves and stems can quantify the degree of damage suffered by plants by pests and diseases, and provide an objective assessment of the overall health status of plants. The higher the percentage of damaged area, the more serious the damage caused by pests and diseases. By calculating the average density of pest and disease distribution on leaves, the distribution of pests and diseases on leaves can be identified. The higher the average density of pest and disease distribution, the more concentrated the pests and diseases are, and there is a higher risk of transmission. The percentage of damaged area on leaves and stems can help managers determine the concentrated areas and main affected parts of pests and diseases, and can be used for precise prevention and control, such as taking stronger prevention and control measures for severely damaged areas, and conducting preventive management for healthy areas. Pest and disease characteristic factors can provide rich data input, help optimize pest and disease prediction models, and improve the accuracy of model predictions. By monitoring the damage to leaves and stems, more scientific predictions can be made on the cycle, rate and peak period of pest and disease occurrence.

[0047] Analyze the soil microbial community of the predicted plant to determine the microbial constraint signals of the predicted plant pests and diseases.

[0048] The specific analysis process is as follows: obtaining the soil microbial community characteristic data set of the plant to be predicted, which specifically includes the total microbial biomass deviation rate and the respiration rate deviation rate of the soil; based on the obtained soil microbial community characteristic data set of the plant to be predicted, a comprehensive analysis is performed to obtain the microbial constraint signal of the plant disease and insect pest to be predicted, and the microbial constraint signal of the plant disease and insect pest to be predicted is used as the analysis basis for obtaining the warning level of the plant disease and insect pest to be predicted.

[0049] The deviation rate of total soil microbial biomass usually refers to the relative deviation between the actually measured soil microbial biomass (the mass of microorganisms per unit volume or weight) and the reference soil microbial biomass, which is obtained through soil microbial biomass measuring instruments, carbon source and phosphorus source measuring and analyzing instruments, etc. The deviation rate of soil respiration rate refers to the relative deviation between the soil respiration rate (the rate at which carbon dioxide is produced during the decomposition of organic matter by microorganisms) and the reference value. The soil respiration rate is generally measured by soil respirometers, gas analyzers, infrared carbon dioxide analyzers and other equipment.

[0050] Soil microbial biomass refers to the number of microorganisms in the soil per unit volume or weight, representing the total amount of active microorganisms in the soil. The higher the microbial biomass, the stronger the microbial activity in the soil. The soil respiration rate reflects the rate at which microorganisms release carbon dioxide in the process of decomposing organic matter. The higher the respiration rate, the higher the activity of soil microorganisms and the faster the rate of decomposition of organic matter. The size of soil microbial biomass will affect the soil respiration rate. When the soil microbial biomass increases, the number and activity of microorganisms will increase, thereby accelerating the decomposition of organic matter and the release of carbon, which increases the soil respiration rate.

[0051] Data such as total microbial biomass and soil respiration rate reflect the microbial activity and health of the soil, helping to judge the risk of pests and diseases more scientifically. Microbial constraint signals, as multi-dimensional environmental data, can improve the accuracy of the pest and disease early warning system. Poor microbial activity or abnormal microbial numbers are usually associated with decreased plant resistance. Therefore, microbial constraint signals can serve as an important predictive basis to help improve the accuracy of pest and disease early warning. Changes in soil microbial activity can reflect the health of the soil. Through continuous monitoring of microbial characteristics, the early warning system can identify the deterioration of soil conditions in advance, making it easier to adjust soil management measures before pests and diseases occur. Microbial constraint signals can help identify soil microbial abnormalities that may cause pests and diseases, and can manage soil microecology more specifically, improve soil health, thereby indirectly reducing the probability of pests and diseases and supporting scientific prevention and control programs.

[0052] The specific analysis process of the plant pest and disease microbial constraint signal to be predicted is as follows:

[0053] ;

[0054] In the formula, To predict the microbial constraint signals of plant pests and diseases, is the deviation rate of total soil microbial biomass, is the soil respiration rate deviation rate, For setting The compensation factor, For setting compensation factor.

[0055] It should be explained that the above-mentioned plant disease and insect pest microbial constraint signal to be predicted is calculated by the deviation rate of the total microbial biomass of the soil and the deviation rate of the respiration rate of the soil. , After normalization, the health status of soil microbial communities directly affects the incidence of diseases and insect pests. The total microbial biomass deviation rate reflects the number and activity of beneficial microorganisms in the soil. The respiration rate deviation rate reflects the microbial activity and the decomposition rate of organic matter, which can help identify the degree of support of soil microorganisms for crop health. Healthy microbial communities usually have an inhibitory effect on pathogens. Microbial constraint signals can reflect the deviation of the microecological balance in the soil and provide guidance for improving soil microecology. If the microbial biomass is low, or the respiration rate is too high or too low, it means that the soil microbial ecosystem may be unstable.

[0056] Analyze the heavy metal characteristics of the soil to which the predicted plants belong, and determine the heavy metal constraint signals of the predicted plant diseases and insect pests.

[0057] The specific analysis process is: obtaining the soil heavy metal characteristic data set of the plants to be predicted, which specifically includes the soil lead concentration, soil cadmium concentration, and soil mercury concentration; based on the obtained soil heavy metal characteristic data set of the plants to be predicted, a comprehensive analysis is performed to obtain the heavy metal constraint signals of the plants to be predicted, and the heavy metal constraint signals of the plants to be predicted are used as the analysis basis for obtaining the warning level of the plants to be predicted.

[0058] Soil lead concentration refers to the content of lead in soil per unit volume or weight. Lead is a common heavy metal pollutant, mainly derived from industrial waste, fertilizers and pesticides. High lead concentration in soil will have a negative impact on plant growth, microbial activity and the ecological environment. Commonly used detection equipment includes atomic absorption spectrometer (AAS), inductively coupled plasma mass spectrometer (ICP-MS) and inductively coupled plasma emission spectrometer (ICP-OES). Soil cadmium concentration refers to the content of cadmium in soil per unit volume or weight. Commonly used detection methods include graphite furnace atomic absorption spectrometer. Soil mercury concentration refers to the content of mercury in soil per unit volume or weight, which is obtained using a cold atomic absorption spectrometer.

[0059] The pollution sources of lead, cadmium and mercury are similar, usually coming from industrial emissions (such as mineral smelting, battery manufacturing, petrochemicals), agricultural activities (such as pesticides, fertilizers), urban landfills and atmospheric deposition. When the concentration of a certain heavy metal in the soil is high, the concentration of other heavy metals may also be high. The mobility and enrichment of lead, cadmium and mercury in the soil are greatly affected by factors such as soil pH, water content, and organic matter content. In acidic soils, these heavy metals are more easily released and migrate to plant roots, increasing the risk of heavy metal absorption by plants. The combined presence of lead, cadmium and mercury will increase the pressure on the soil ecosystem. Lead will interfere with plant photosynthesis, cadmium will affect plant nutrient absorption, and mercury has a strong toxic effect. When the three exist together, their combined effect will reduce the growth ability of plants and affect the structure of soil microbial communities.

[0060] The accumulation of heavy metals such as lead, cadmium, and mercury in the soil will affect the growth and disease resistance of plants and increase the probability of pests and diseases. Heavy metal constraint signals can help identify potential pollution risks in the soil and reveal the impact of the plant growth environment on pests and diseases. Soil with excessive heavy metals is often associated with decreased plant immunity. The pest and disease early warning system can more accurately identify the actual pest and disease risks faced by plants through heavy metal constraint signals, thereby reducing misjudgments caused by ignoring soil pollution. Heavy metal data can help detect soil pollution conditions at an early stage, allowing managers to take timely measures such as improving the soil and planting pollution-resistant plants before pollution affects plant health and pests and diseases occur, so as to avoid the further spread of pests and diseases. Understanding the concentration of heavy metals in the soil can help formulate accurate prevention and control plans. If the heavy metal concentration is high, soil improvement measures may be given priority to reduce the risk of infection from pests and diseases, thereby reducing dependence on pesticides.

[0061] The specific analysis process of heavy metal constraint signals of plant pests and diseases to be predicted is as follows:

[0062] ;

[0063] In the formula, To predict the heavy metal constraint signals of plant pests and diseases, is the soil lead concentration, is the soil cadmium concentration, is the mercury concentration in the soil, For setting The compensation factor, For setting The compensation factor, For setting The compensation factor of , e is a natural constant.

[0064] It should be explained that the heavy metal constraint signals of plant diseases and insect pests to be predicted are calculated by the soil lead concentration, soil cadmium concentration, soil mercury concentration, soil lead concentration, soil cadmium concentration, and soil mercury concentration. , , After normalization, excessive heavy metals such as lead, cadmium and mercury will inhibit the normal physiological activities of crops and weaken their immune systems, thereby increasing the risk of diseases and insect pests. The calculated heavy metal constraint signal can help identify the potential stress of soil heavy metals on crops and determine whether there are conditions that induce diseases and insect pests. Heavy metal pollution will change the structure of microbial communities in the soil, inhibit the growth of beneficial microorganisms, and promote the reproduction of certain pathogens. Heavy metal constraint signals can more accurately predict the probability of disease and insect pests by comprehensively analyzing the concentrations of lead, cadmium and mercury. The heavy metal constraint signal can clarify the relationship between the risk of disease and insect pests and specific heavy metal pollution, helping to select appropriate soil improvement measures.

[0065] Analyze the planting environment of the plants to be predicted and obtain the environmental constraints of the plant diseases and insect pests to be predicted.

[0066] The specific analysis process is as follows: obtaining the planting environment data set of the plants to be predicted, which specifically includes the planting environment temperature deviation rate, the planting environment humidity deviation rate, and the planting rhizosphere environment oxygen concentration deviation rate; based on the obtained planting environment data set of the plants to be predicted, a comprehensive analysis is performed to obtain the environmental constraint factors of the plants to be predicted, and the environmental constraint factors of the plants to be predicted are used as the analysis basis for obtaining the warning level of the plants to be predicted.

[0067] The temperature deviation rate of the planting environment refers to the percentage deviation between the current ambient temperature and the set or ideal temperature. Commonly used equipment is a temperature sensor or a temperature and humidity recorder. The humidity deviation rate of the planting environment indicates the percentage deviation between the current ambient humidity and the ideal humidity level. Air humidity sensors or temperature and humidity recorders are used to measure the actual humidity data, and then the difference with the set humidity is calculated. The oxygen concentration deviation rate of the planting rhizosphere environment indicates the percentage deviation between the actual oxygen concentration and the ideal oxygen concentration. The oxygen concentration in the rhizosphere plays a key role in root respiration, microbial activity and nutrient conversion. Excessive deviation will lead to root hypoxia or excessive exposure to the oxygen environment. The rhizosphere oxygen concentration is generally measured by a dissolved oxygen sensor, a gas analyzer or a dedicated soil oxygen probe, and the deviation rate is calculated using the sensor data.

[0068] Temperature, humidity and oxygen concentration all directly affect the growth and metabolic process of crop roots. Appropriate temperature and humidity can promote the absorption of water and nutrients by the roots, while the oxygen concentration in the rhizosphere is the key condition for root respiration. There is a direct interaction between temperature and humidity. High temperature usually increases the saturated vapor pressure of the air and reduces the relative humidity, while low temperature increases humidity. Ambient temperature deviation will affect humidity deviation, which in turn affects the transpiration of plants. Soil moisture directly affects the oxygen concentration in the rhizosphere, because water will occupy soil pores, reduce air flow, and make it difficult for oxygen to reach the rhizosphere environment. When the humidity is high, the oxygen concentration in the rhizosphere is often low, resulting in root hypoxia and affecting root respiration.

[0069] Environmental conditions such as temperature, humidity and oxygen concentration directly affect the growth of plants and the reproduction conditions of pests and diseases. Pest and disease environmental constraints can reveal the impact of the environment on the occurrence of pests and diseases, and provide key background data for accurately judging the risk of pests and diseases. Environmental factors play a key role in the occurrence of pests and diseases. Comprehensive environmental constraints can more accurately assess the possibility of pest and disease occurrence, thereby improving the accuracy of the early warning system and reducing false alarms or omissions. Environmental conditions such as temperature and humidity fluctuate over time. Through dynamic monitoring of environmental constraints, the early warning system can update the pest and disease risk level in a timely manner according to environmental changes, and achieve more flexible and adaptive pest and disease prediction. Within a specific range of temperature, humidity or oxygen concentration, the breeding and spread of pests and diseases are often accelerated. Different environmental conditions may require different pest and disease control methods. Understanding environmental constraints can provide farmers with scientific prevention and control suggestions. When the humidity is high, methods can be adopted to reduce humidity to reduce disease risks and avoid unnecessary use of pesticides.

[0070] The specific analysis process of the environmental constraint factors of plant diseases and insect pests to be predicted is as follows:

[0071] ;

[0072] In the formula, The environmental constraints of plant pests and diseases to be predicted are: is the planting environment temperature deviation rate, is the humidity deviation rate of the planting environment, is the oxygen concentration deviation rate of the rhizosphere environment. For setting The compensation factor, For setting The compensation factor, For setting The compensation factor of , e is a natural constant.

[0073] It should be explained that the above-mentioned environmental constraint factors of plant diseases and insect pests to be predicted are calculated by the deviation rate of planting environment temperature, planting environment humidity, and planting rhizosphere environment oxygen concentration. , , After normalization, temperature, humidity and oxygen concentration are key environmental factors for the occurrence of pests and diseases. The deviation rate can quantify the gap between these factors and the ideal values, and help determine whether the environment is in a state suitable for the occurrence of pests and diseases. By comprehensively calculating the environmental constraint factors through the deviation rate, we can more sensitively identify environmental changes that are not conducive to crop growth but conducive to the spread of pests and diseases, and improve the accuracy of pest and disease prediction. The environmental deviation rate can reflect the fluctuation of environmental conditions in advance. When the environmental constraint factors show deviations from the ideal values, early warnings can be issued before pests and diseases break out, and prevention and control measures can be taken in advance to reduce the incidence and impact of pests and diseases.

[0074] The deviation constraint signals of the plant diseases and insect pests to be predicted are determined by combining the microbial constraint signals of the plant diseases and insect pests to be predicted, the heavy metal constraint signals of the plant diseases and insect pests to be predicted, and the environmental constraint factors of the plant diseases and insect pests to be predicted.

[0075] Based on the characteristic factors of the plant diseases and insect pests to be predicted and the deviation constraint signals of the plant diseases and insect pests to be predicted, the warning level of the plant diseases and insect pests to be predicted is obtained.

[0076] The specific analysis process is as follows: storing the microbial constraint signal of the plant pests and diseases to be predicted, the heavy metal constraint signal of the plant pests and diseases to be predicted and the environmental constraint factor of the plant pests and diseases to be predicted as a designated label, comparing the designated label with the deviation constraint signal of the plant pests and diseases to be predicted corresponding to each designated label stored in the database, and obtaining the deviation constraint signal of the plant pests and diseases to be predicted corresponding to the designated label; storing the characteristic factor of the plant pests and diseases to be predicted and the deviation constraint signal of the plant pests and diseases to be predicted as a designated code, comparing the designated code with the warning signal of the plant pests and diseases to be predicted corresponding to each designated code stored in the database, and obtaining the warning signal of the plant pests and diseases to be predicted corresponding to the designated code; comparing the warning signal of the plant pests and diseases to be predicted with the constraint range of the warning signal of the plant pests and diseases to be predicted stored in the database; if If the warning signal of the plant diseases and insect pests to be predicted falls within the constraint range of the warning signal of the plant diseases and insect pests to be predicted, the warning level of the plant diseases and insect pests to be predicted corresponding to the warning signal of the plant diseases and insect pests to be predicted is level two; if the warning signal of the plant diseases and insect pests to be predicted does not fall within the constraint range of the warning signal of the plant diseases and insect pests to be predicted, and the warning signal of the plant diseases and insect pests to be predicted is less than the minimum value in the constraint range of the warning signal of the plant diseases and insect pests to be predicted, the warning level of the plant diseases and insect pests to be predicted corresponding to the warning signal of the plant diseases and insect pests to be predicted is level one; if the warning signal of the plant diseases and insect pests to be predicted does not fall within the constraint range of the warning signal of the plant diseases and insect pests to be predicted, and the warning signal of the plant diseases and insect pests to be predicted is greater than the maximum value in the constraint range of the warning signal of the plant diseases and insect pests to be predicted, the warning level of the plant diseases and insect pests to be predicted corresponding to the warning signal of the plant diseases and insect pests to be predicted is level three.

[0077] It should be noted that the constraint range of the plant pest and disease warning signal to be predicted stored in the above database is a left-closed and right-closed interval. Deviation constraint signals are generated through constraint signals such as microorganisms, heavy metals and the environment, and warning signals are generated together with characteristic factors, so that the assessment of pests and diseases is more refined, thereby making an accurate judgment on the risk level. This multi-level warning mechanism can accurately divide different warning levels according to the position of the pest and disease warning signal within the constraint range, so that the system can sensitively respond to the size and potential trend of pest and disease risks, thereby improving the accuracy of the prediction. The warning levels are divided into level one, level two and level three, which is convenient for farmers or managers to take targeted prevention and control measures. Level two warning indicates moderate risk and is suitable for close monitoring, while level three warning indicates high risk and requires immediate intervention. Different warning levels make prevention and control more scientific. By accurately comparing deviation signals and warning signals, excessive or insufficient prevention and control measures can be avoided, the waste of resources such as pesticides and labor can be reduced, and the negative impact on the environment can be reduced. At the same time, it also avoids the loss of pests and diseases caused by insufficient prevention and control, and improves the automation and stability of the system.

[0078] Reference Figure 2As shown, the second aspect of the present invention provides a plant disease and pest prediction system based on image recognition, including a disease and pest characteristic factor analysis module, a microbial constraint signal determination module, a heavy metal constraint signal determination module, an environmental constraint factor acquisition module, a deviation constraint signal determination module and a disease and pest warning level determination module.

[0079] The pest and disease characteristic factor analysis module is used to obtain the image of the plant growth condition to be predicted and determine whether the quality of the image of the plant growth condition to be predicted is qualified; if the quality of the image of the plant growth condition to be predicted is unqualified, the image of the plant growth condition to be predicted is re-acquired; if the quality of the image of the plant growth condition to be predicted is qualified, the plant growth condition to be predicted is analyzed to obtain the characteristic factors of the plant pest and disease to be predicted.

[0080] The microbial constraint signal determination module is used to analyze the soil microbial community of the predicted plant and determine the microbial constraint signal of the predicted plant disease and insect pest.

[0081] The heavy metal constraint signal determination module is used to analyze the heavy metal characteristics of the soil of the predicted plant and determine the heavy metal constraint signals of the predicted plant pests and diseases.

[0082] The environmental constraint factor acquisition module is used to analyze the planting environment of the plant to be predicted and obtain the environmental constraint factors of the plant diseases and insect pests to be predicted.

[0083] The deviation constraint signal determination module is used to determine the deviation constraint signal of the plant pests and diseases to be predicted by combining the microbial constraint signal of the plant pests and diseases to be predicted, the heavy metal constraint signal of the plant pests and diseases to be predicted and the environmental constraint factor of the plant pests and diseases to be predicted.

[0084] The pest warning level determination module is used to obtain the warning level of the plant pests to be predicted based on the characteristic factors of the plant pests to be predicted and the deviation constraint signal of the plant pests to be predicted.

[0085] The above contents are merely examples and explanations of the structure of the present invention. The technicians in this technical field may make various modifications or additions to the specific embodiments described or replace them in a similar manner. As long as they do not deviate from the structure of the invention or exceed the scope defined by the claims, they should all fall within the protection scope of the present invention.

Claims

1. A method for predicting plant diseases and insect pests based on image recognition, characterized in that: The following steps are involved: Acquire an image of the growth condition of the plant to be predicted, and determine whether the quality of the image of the growth condition of the plant to be predicted is qualified; If the quality of the image of the growth condition of the plant to be predicted is unqualified, reacquire the image of the growth condition of the plant to be predicted; If the image quality of the plant growth condition to be predicted is qualified, the plant growth condition to be predicted is analyzed to obtain characteristic factors of the plant diseases and insect pests to be predicted; Analyze the soil microbial community of the predicted plant to determine the constraint signal of the plant pests and diseases; Analyze the heavy metal characteristics of the soil of the predicted plants and determine the heavy metal constraint signals of the predicted plant diseases and insect pests; Analyze the planting environment of the plants to be predicted and obtain the environmental constraint factors of the plant diseases and insect pests to be predicted; Determine the deviation constraint signal of the plant pests and diseases to be predicted by combining the constraint signal of the microorganisms of the plant pests and diseases to be predicted, the constraint signal of the heavy metals of the plant pests and diseases to be predicted and the environmental constraint factor of the plant pests and diseases to be predicted; Based on the characteristic factors of the plant diseases and insect pests to be predicted and the deviation constraint signals of the plant diseases and insect pests to be predicted, the warning level of the plant diseases and insect pests to be predicted is obtained.

2. The method for predicting plant diseases and insect pests based on image recognition according to claim 1, characterized in that: The specific analysis process of judging whether the quality of the image of the plant growth condition to be predicted is qualified is as follows: Acquire a data set of image quality of plant growth conditions to be predicted, wherein the data set of image quality of plant growth conditions to be predicted specifically includes a growth condition image shooting distance deviation rate, a growth condition image vertical resolution, and a growth condition image contrast deviation rate; Based on the acquired plant growth condition image quality data set, a growth condition image quality determination factor is obtained through comprehensive analysis, and the growth condition image quality determination factor is used as an analysis basis for judging whether the quality of the plant growth condition image to be predicted is qualified; comparing the growth condition image quality determination factor with the growth condition image quality determination threshold stored in the database; If the growth condition image quality determination factor is lower than the growth condition image quality determination threshold, the quality of the plant growth condition image to be predicted is unqualified, and the plant growth condition image to be predicted needs to be reacquired; If the growth condition image quality determination factor is not lower than the growth condition image quality determination threshold, the quality of the growth condition image of the plant to be predicted is qualified.

3. The method for predicting plant diseases and insect pests based on image recognition according to claim 2, characterized in that: The growth status image quality determination factor, the specific analysis process is: ； In the formula, is the growth status image quality determination factor, is the growth status image shooting distance deviation rate, is the vertical resolution of the growth status image, is the growth status image contrast deviation rate, For setting The compensation factor, For setting The compensation factor, For setting compensation factor.

4. The method for predicting plant diseases and insect pests based on image recognition according to claim 1, characterized in that: The growth status of the plant to be predicted is analyzed to obtain the characteristic factors of the plant diseases and insect pests to be predicted. The specific analysis process is as follows: Obtaining a data set of plant growth conditions to be predicted, wherein the data set of plant growth conditions to be predicted specifically includes a percentage of damaged leaf areas of plants to be predicted, a percentage of damaged stem areas of plants to be predicted, and an average density of leaf distribution of plant diseases and insect pests to be predicted; Based on the acquired data set of the growth status of the plants to be predicted, a comprehensive analysis is performed to obtain the characteristic factors of the plant diseases and insect pests to be predicted, and the characteristic factors of the plant diseases and insect pests to be predicted are used as the analysis basis for obtaining the warning level of the plant diseases and insect pests to be predicted.

5. The method for predicting plant diseases and insect pests based on image recognition according to claim 4, characterized in that: The specific analysis process of the characteristic factors of plant diseases and insect pests to be predicted is as follows: ； In the formula, are the characteristic factors of plant diseases and insect pests to be predicted, is the percentage of damaged area of ​​plant leaves to be predicted, is the percentage of damaged plant stem area to be predicted, is the average density of the leaves of the plant pests and diseases to be predicted, For setting The compensation factor, For setting The compensation factor, For setting compensation factor.

6. The method for predicting plant diseases and insect pests based on image recognition according to claim 1, characterized in that: The specific analysis process of determining the constraint signal of the plant pest and disease microorganism to be predicted is as follows: Acquire a soil microbial community characteristic dataset of the plant to be predicted, wherein the soil microbial community characteristic dataset of the plant to be predicted specifically includes a deviation rate of the total microbial biomass of the soil and a deviation rate of the respiration rate of the soil; Based on the acquired soil microbial community characteristic data set of the plants to be predicted, a comprehensive analysis is performed to obtain the microbial constraint signals of the plant diseases and insect pests to be predicted, which are used as the analysis basis for obtaining the early warning level of the plant diseases and insect pests to be predicted.

7. The method for predicting plant diseases and insect pests based on image recognition according to claim 1, characterized in that: The specific analysis process of determining the heavy metal constraint signal of the plant pest to be predicted is as follows: Obtaining a heavy metal characteristic dataset of the soil of the plant to be predicted, wherein the heavy metal characteristic dataset of the soil of the plant to be predicted specifically includes the lead concentration, the cadmium concentration, and the mercury concentration of the soil; Based on the acquired soil heavy metal characteristic data set of the plants to be predicted, a comprehensive analysis is performed to obtain the heavy metal constraint signals of the plants' diseases and insect pests to be predicted. The heavy metal constraint signals of the plants' diseases and insect pests to be predicted are used as the analysis basis for obtaining the warning level of the plants' diseases and insect pests to be predicted.

8. The method for predicting plant diseases and insect pests based on image recognition according to claim 1, characterized in that: The specific analysis process of obtaining the environmental constraint factors of the plant pests and diseases to be predicted is as follows: Obtaining a plant planting environment data set to be predicted, the plant planting environment data set to be predicted specifically includes a planting environment temperature deviation rate, a planting environment humidity deviation rate, and a planting rhizosphere environment oxygen concentration deviation rate; Based on the acquired planting environment data set of the plants to be predicted, a comprehensive analysis is performed to obtain the environmental constraint factors of the plants' diseases and insect pests to be predicted, and the environmental constraint factors of the plants' diseases and insect pests to be predicted are used as the analysis basis for obtaining the warning level of the plants' diseases and insect pests to be predicted.

9. The method for predicting plant diseases and insect pests based on image recognition according to claim 1, characterized in that: The specific analysis process of obtaining the early warning level of the plant pests and diseases to be predicted is as follows: The microbial constraint signal of the plant pests and diseases to be predicted, the heavy metal constraint signal of the plant pests and diseases to be predicted, and the environmental constraint factor of the plant pests and diseases to be predicted are stored as designated tags, and the designated tags are compared with the deviation constraint signals of the plant pests and diseases to be predicted corresponding to the designated tags stored in the database to obtain the deviation constraint signals of the plant pests and diseases to be predicted corresponding to the designated tags; The characteristic factors of the plant diseases and insect pests to be predicted and the deviation constraint signals of the plant diseases and insect pests to be predicted are stored as designated codes, and the designated codes are compared with the early warning signals of the plant diseases and insect pests to be predicted corresponding to the designated codes stored in the database to obtain the early warning signals of the plant diseases and insect pests to be predicted corresponding to the designated codes; comparing the early warning signal of the plant disease and insect pest to be predicted with the constraint range of the early warning signal of the plant disease and insect pest to be predicted stored in the database; If the warning signal of the plant disease and insect pest to be predicted falls within the constraint range of the warning signal of the plant disease and insect pest to be predicted, the warning level of the plant disease and insect pest to be predicted corresponding to the warning signal of the plant disease and insect pest to be predicted is level 2; If the warning signal of the plant disease and insect pest to be predicted does not fall within the constraint range of the warning signal of the plant disease and insect pest to be predicted, and the warning signal of the plant disease and insect pest to be predicted is less than the minimum value in the constraint range of the warning signal of the plant disease and insect pest to be predicted, the warning level of the plant disease and insect pest to be predicted corresponding to the warning signal of the plant disease and insect pest to be predicted is level one; If the warning signal of the plant disease and insect pest to be predicted does not fall within the constraint range of the warning signal of the plant disease and insect pest to be predicted, and the warning signal of the plant disease and insect pest to be predicted is greater than the maximum value in the constraint range of the warning signal of the plant disease and insect pest to be predicted, then the warning level of the plant disease and insect pest to be predicted corresponding to the warning signal of the plant disease and insect pest to be predicted is level three.

10. A plant disease and insect pest prediction system based on image recognition, applied to a plant disease and insect pest prediction method based on image recognition according to any one of claims 1 to 9, characterized in that: It includes a pest and disease characteristic factor analysis module, a microbial constraint signal determination module, a heavy metal constraint signal determination module, an environmental constraint factor acquisition module, a deviation constraint signal determination module and a pest and disease warning level determination module, among which: The pest and disease characteristic factor analysis module is used to obtain the image of the plant growth condition to be predicted and determine whether the quality of the image of the plant growth condition to be predicted is qualified; If the quality of the image of the growth condition of the plant to be predicted is unqualified, reacquire the image of the growth condition of the plant to be predicted; If the image quality of the plant growth condition to be predicted is qualified, the plant growth condition to be predicted is analyzed to obtain characteristic factors of the plant diseases and insect pests to be predicted; The microbial constraint signal determination module is used to analyze the soil microbial community of the predicted plant to determine the microbial constraint signal of the plant disease and insect pest to be predicted; The heavy metal constraint signal determination module is used to analyze the heavy metal characteristics of the soil of the predicted plant and determine the heavy metal constraint signal of the plant disease and insect pest to be predicted; The environmental constraint factor acquisition module is used to analyze the planting environment of the plant to be predicted and obtain the environmental constraint factors of the plant diseases and insect pests to be predicted; The deviation constraint signal determination module is used to determine the deviation constraint signal of the plant pests and diseases to be predicted by combining the microbial constraint signal of the plant pests and diseases to be predicted, the heavy metal constraint signal of the plant pests and diseases to be predicted and the environmental constraint factor of the plant pests and diseases to be predicted; The pest warning level determination module is used to obtain the warning level of the plant pest to be predicted based on the characteristic factors of the plant pest to be predicted and the deviation constraint signal of the plant pest to be predicted.

Images