Information processing method for tea garden

Abstract

The invention provides an information processing method for a tea garden, and the method comprises the steps: obtaining tea garden environment parameters of every day, inputting the tea garden environment parameters of every day into a growth state prediction model to obtain a tea tree growth rate of every day, using a time sequence analysis algorithm to analyze the tea garden environment parameters and the tea tree growth rate every day to obtain the incidence rate of tea garden diseases and insect pests every day, and determining the observation time of the unmanned aerial vehicle according to the incidence rate of tea garden diseases and insect pests every day. The unmanned aerial vehicle is controlled to carry the spectrometer to collect the spectral image of the tea garden at the observation moment of the unmanned aerial vehicle, and the disease and pest recognition model is used to perform disease and pest recognition on the spectral image to obtain tea tree health state data, so that the unmanned aerial vehicle can timely discover diseases and pests in the tea garden, and the unmanned aerial vehicle inspection cost can be reduced.

Description

Technical Field

[0001] The present application relates to the field of agricultural informatization and intelligent algorithm technology, and in particular to an information processing method for a tea garden. Background Art

[0002] With the widespread application of information technology, traditional tea garden management that relies on manual experience and judgment has been replaced by information-based tea garden management.

[0003] With the widespread use of drones, drones are also used in information-based tea garden management. More specifically, drones are used to patrol tea gardens to obtain image data of tea gardens. The image data is then identified through machine learning models to obtain the growth and occurrence of pests and diseases in the tea gardens. The inspection frequency of drones is usually determined through manual experience, and drones are used to patrol tea gardens at regular intervals.

[0004] However, the existing method of determining the inspection frequency of drones through manual experience cannot accurately grasp the laws of pests and diseases in tea gardens. If the inspection frequency is too sparse, it will not be possible to accurately inspect in the early stages of pests and diseases, reducing the accuracy of inspections and failing to detect pests and diseases in time. If the inspection frequency is too high, it will increase the cost of drone inspections. Summary of the invention

[0005] The embodiment of the present application provides an information processing method for a tea garden. By predicting the time when pests and diseases occur, a drone is controlled to patrol the tea garden at the time when the pests and diseases occur, so that the pests and diseases in the tea garden can be discovered in time, and the cost of drone patrol can be reduced.

[0006] The embodiment of the present application provides an information processing method for a tea garden. The information processing method is applied to a background server. The information processing method includes:

[0007] Obtaining daily tea garden environmental parameters, including temperature, humidity, temperature change rate, daily average sunshine duration, and precipitation;

[0008] Input the daily tea garden environmental parameters into the growth state prediction model to obtain the daily tea tree growth rate;

[0009] Use time series analysis algorithms to analyze daily tea garden environmental parameters and daily tea tree growth rates to obtain the daily incidence of tea garden pests and diseases;

[0010] The observation time of the drone is determined according to the daily incidence of tea garden pests and diseases, and the drone is controlled to carry a spectrometer to collect spectral images of the tea garden at the observation time;

[0011] The pest and disease recognition model is used to identify pests and diseases in spectral images to obtain tea tree health status data.

[0012] In the above technical solution, the background server analyzes the daily tea garden environmental parameters to obtain the daily tea tree growth rate, and uses the time series analysis algorithm to process the daily tea tree growth rate and tea garden environmental parameters to obtain the daily incidence of pests and diseases. The observation time is determined based on the daily incidence of pests and diseases, and the drone equipped with a spectrometer is controlled to collect spectral images at the observation time. Since the law of tea tree pests and diseases is obtained through analysis, the observation time is determined based on the law of tea tree pests and diseases. In this way, drones are used for monitoring in the early stage of tea tree pests and diseases. If the recognition result of the monitoring image shows that it is in the early stage of pests and diseases, pests and diseases can be controlled in time.

[0013] In a possible implementation, a time series analysis algorithm is used to analyze the daily tea garden environmental parameters and the daily tea tree growth rate to obtain the daily incidence of tea garden pests and diseases, including:

[0014] Obtain the key factor relationships that cause tea tree diseases and insect pests, and cross-process the daily tea garden environmental parameters and daily tea tree growth rates based on the key factor relationships to obtain daily cross-information;

[0015] The time series analysis algorithm is used to analyze the daily cross-information and obtain the daily incidence of tea garden pests and diseases.

[0016] In the above technical scheme, considering that pests and diseases are caused by multiple factors, when using the time series analysis algorithm to analyze the environmental parameters and the growth rate of tea trees, cross-processing is performed according to the relationship between the key factors causing the pests and diseases of tea trees to obtain daily cross-information, and the time series analysis algorithm is used to analyze the daily cross-information to obtain the daily incidence of pests and diseases in the tea garden, thereby improving the accuracy of pest and disease prediction.

[0017] In a possible implementation, the key factor relationship causing tea tree diseases and insect pests is obtained, and the daily tea garden environmental parameters and tea tree growth rate are processed according to the key factor relationship to obtain daily cross information, including:

[0018] If the key factor relationship includes the temperature and humidity relationship, the product of the temperature and humidity of the tea garden every day is calculated to obtain the daily temperature and humidity cross information;

[0019] If the key factor relationship includes the temperature-growth state relationship, the product of the daily tea garden temperature change rate and the tea tree growth rate is calculated to obtain the daily temperature-growth cross-information.

[0020] In the above technical scheme, since the cross-influence of temperature and humidity can lead to the outbreak of pests and diseases, the cross-influence of the temperature change rate of the tea garden and the growth rate of the tea trees can lead to the outbreak of pests and diseases, the product of the temperature and humidity of the tea garden is calculated as the temperature and humidity cross-information, and the product of the temperature change rate of the tea garden and the growth rate of the tea trees is calculated as the temperature-growth cross-information. In this way, the incidence of pests and diseases can be predicted based on the temperature and humidity cross-information and the temperature-growth cross-information, which can improve the prediction accuracy.

[0021] In a possible implementation, a time series analysis algorithm is used to analyze the daily cross information to obtain the daily incidence of tea garden pests and diseases, including:

[0022] Normalize and vectorize the daily temperature and humidity cross information and the daily temperature growth cross information to obtain the temperature and humidity vector and the temperature growth vector;

[0023] Use the fusion coefficient to fuse the temperature and humidity vector and the temperature growth vector to obtain the input vector;

[0024] The ARIMA model is used to analyze the input vector to obtain the daily incidence of tea garden pests and diseases.

[0025] In the above technical solution, after obtaining the daily temperature and humidity cross information and the daily temperature growth cross information, the input vector is obtained by fusing the temperature and humidity cross information with the temperature growth cross information. In this way, the cross information of the two variables is fused to obtain the input vector of a single variable. In this way, there is no need to use a multi-vector model for analysis and processing, thereby reducing the complexity of data processing.

[0026] In a possible implementation, the temperature and humidity vector and the temperature growth vector are fused using a fusion coefficient to obtain an input vector, specifically including:

[0027] Calculate the weighted sum of the temperature and humidity vector and the temperature growth vector according to the first coefficient and the second coefficient to obtain an input vector;

[0028] The fusion coefficient includes a first coefficient and a second coefficient, and the first coefficient, the second coefficient and the parameters in the ARIMA model are obtained through training.

[0029] In the above technical solution, the first coefficient and the second coefficient are used to calculate the weighted sum of the temperature and humidity vector and the temperature growth vector, so as to convert the time series data of the two variables into the time series data of a single variable. In this way, the time series data processing model of a single variable can be used to analyze the time series data, thereby reducing the number of model parameters and reducing training time.

[0030] In a possible implementation, a pest and disease identification model is used to identify pests and diseases on spectral images to obtain tea tree health status data, specifically including:

[0031] Determine the growth stage of the tea tree every day, and take the growth stage corresponding to the acquisition time of the spectral image as the target growth stage;

[0032] Determine the pest and disease identification model corresponding to the target growth stage according to the target growth stage and the model mapping table;

[0033] The pest and disease identification model corresponding to the target growth stage is used to identify pests and diseases on the spectral images to obtain the health status data of the tea trees.

[0034] In the above technical scheme, after obtaining the spectral image collected by the drone, the growth stage corresponding to the spectral image is determined according to the collection time of the spectral image, and the pest and disease identification model of the corresponding stage is obtained. The spectral image is used to identify pests and diseases using the pest and disease identification model corresponding to the target growth stage to obtain the health status data of the tea tree. Since the tree shapes of tea trees at different growth stages are different, each growth stage corresponds to a pest and disease identification model, which can improve the accuracy of the pest and disease identification model.

[0035] In a possible implementation, after using the pest and disease identification model to identify pests and diseases on the spectral image to obtain tea tree health status data, the method further includes:

[0036] If the health status data shows that the tea trees are in the early stages of disease and pest infestation or have disease and pest infestation, the type of disease and pest is obtained, and prevention and control instructions are generated based on the type of disease and pest, so that the drone is equipped with the corresponding pesticide spraying device to patrol the tea garden and spray pesticides.

[0037] In the above technical solution, when the health status data indicates that the tea tree is in the early stage of disease and pest occurrence or when disease and pest occur, prevention and control instructions are generated according to the type of disease and pest, and the drone is controlled to carry out disease and pest control, thereby improving the prevention and control efficiency.

[0038] The embodiment of the present application provides an information processing device for a tea garden, comprising:

[0039] An acquisition module is used to acquire tea garden environmental parameters every day, wherein the tea garden environmental parameters include temperature, humidity, temperature change rate, daily average sunshine duration and precipitation;

[0040] A processing module, used for inputting the daily tea garden environmental parameters into a growth state prediction model to obtain a daily tea tree growth rate;

[0041] The processing module is further used to analyze the daily tea garden environmental parameters and the daily tea tree growth rate using a time series analysis algorithm to obtain the daily incidence of tea garden pests and diseases;

[0042] The processing module is further used to determine the observation time of the drone according to the daily incidence of pests and diseases in the tea garden, and control the drone to carry a spectrometer to collect spectral images of the tea garden at the observation time of the drone;

[0043] The processing module is also used to use a pest and disease identification model to identify pests and diseases on the spectral image to obtain tea tree health status data.

[0044] The embodiment of the present application provides a backend server, including: a memory, a processor;

[0045] Memory stores computer-executable instructions;

[0046] The processor executes the computer-executable instructions stored in the memory, so that the processor executes various possible implementations as described above.

[0047] An embodiment of the present application provides a computer-readable storage medium, in which computer-executable instructions are stored. When the computer-executable instructions are executed by a processor, they are used to implement the various possible implementation methods described above.

[0048] An embodiment of the present application provides a computer program product, including a computer program, which implements the above various possible implementation methods when executed by a processor. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0050] Figure 1 A schematic diagram of a scenario for the information processing method for a tea garden provided in this application;

[0051] Figure 2 A schematic diagram of the process flow of the information processing method for a tea garden provided in this application;

[0052] Figure 3 This is a schematic diagram of the structure of the information processing device for a tea garden provided in this application.

[0053] The above drawings have shown clear embodiments of the present application, which will be described in more detail later. These drawings and text descriptions are not intended to limit the scope of the present application in any way, but to illustrate the concept of the present application to those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION

[0054] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application. Instead, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.

[0055] With the widespread use of drones, drones are also used in information-based tea garden management. More specifically, drones are used to patrol tea gardens to obtain image data of tea gardens. The image data is then identified through machine learning models to obtain the growth and occurrence of pests and diseases in the tea gardens. The inspection frequency of drones is usually determined through manual experience, and drones are used to patrol tea gardens at regular intervals.

[0056] However, the existing method of determining the inspection frequency of drones through manual experience cannot accurately grasp the laws of pest and disease occurrence in tea gardens, resulting in the inability to accurately conduct inspections in the early stages of pest and disease occurrence, reducing inspection accuracy and failing to detect pests and diseases in a timely manner.

[0057] The embodiment of the present application provides an information processing method for a tea garden. By predicting the time when pests and diseases occur, a drone is controlled to patrol the tea garden at the time when the pests and diseases occur, so that the pests and diseases in the tea garden can be discovered in time, and the cost of drone patrol can be reduced.

[0058] The technical solution of the present application and how the technical solution of the present application solves the above-mentioned technical problems are described in detail below with specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of the present application will be described below in conjunction with the accompanying drawings.

[0059] Figure 1 A schematic diagram of a scenario for the information processing method for a tea garden provided in this application. Figure 2 A flow chart of the information processing method for a tea garden provided in this application, such as Figure 1 and Figure 2 As shown, some embodiments of the present application provide an information processing method for a tea garden, and the information processing method specifically includes the following steps:

[0060] S101. The backend server obtains tea garden environmental parameters every day.

[0061] Among them, the tea garden environmental parameters include temperature, humidity, temperature change rate, daily average light duration and precipitation. A temperature and humidity sensor is arranged in the tea garden, and the temperature and humidity sensor is used to collect the temperature and humidity at various times of the day. The daily temperature change rate and the daily average temperature are calculated according to the collected temperatures at various times, and the daily average temperature is used as the daily temperature. The humidity at various times of the day is calculated as the daily average humidity as the daily humidity. A light intensity sensor is arranged in the tea garden, and the light intensity sensor collects the light intensity in the tea garden, and calculates the daily average light duration based on the collected light intensity. A precipitation detector is arranged in the tea garden, and the precipitation detector is used to detect the precipitation in the tea garden.

[0062] S102. The backend server inputs the daily tea garden environmental parameters into the growth state prediction model to obtain the daily tea tree growth rate.

[0063] Among them, the growth status prediction model can be a decision tree, random forest, support vector machine or neural network model, and the tea garden environmental parameters are input into the growth status prediction model to obtain the daily tea tree growth rate.

[0064] S103. The backend server uses a time series analysis algorithm to analyze the daily tea garden environmental parameters and the daily tea tree growth rate to obtain the daily incidence of tea garden pests and diseases.

[0065] The time series analysis algorithm may be a vector autoregression model or a long short-term memory network. By training the vector autoregression model or the long short-term memory network, the vector autoregression model or the long short-term memory network can analyze the daily tea garden environmental parameters and the daily tea tree growth rate to obtain the incidence of tea garden pests and diseases at various future times.

[0066] S104. The backend server determines the observation time according to the daily incidence of diseases and insect pests in the tea garden, and controls the drone equipped with a spectrometer to collect spectral images of the tea garden at the observation time.

[0067] Among them, a threshold of pests and diseases is set. If the incidence of pests and diseases in the tea garden is greater than the threshold, the date corresponding to the incidence of pests and diseases greater than the threshold is used as the observation time of pests and diseases. At the observation time, the drone equipped with a spectrometer is controlled to patrol the tea garden and collect spectral images of the tea garden.

[0068] S105. The backend server uses a pest and disease identification model to identify the spectral image and obtain tea tree health status data.

[0069] If the spectral image collected by the spectrometer is in multiple bands, a spectral image in a single band is extracted from the collected spectral image, and the spectral image in a single band is identified using a pest and disease identification model to obtain tea tree health status data.

[0070] The pest and disease identification model is a convolutional neural network. By training the convolutional neural network, the convolutional neural network is enabled to identify spectral images to obtain tea tree health status data.

[0071] In the above technical solution, the background server analyzes the daily tea garden environmental parameters to obtain the daily tea tree growth rate, and uses the time series analysis algorithm to process the daily tea tree growth rate and tea garden environmental parameters to obtain the daily incidence of pests and diseases. The observation time is determined based on the daily incidence of pests and diseases, and the drone equipped with a spectrometer is controlled to collect spectral images at the observation time. Since the law of tea tree pests and diseases is obtained through analysis, the observation time is determined based on the law of tea tree pests and diseases. In this way, drones are used for monitoring in the early stage of tea tree pests and diseases. If the recognition result of the monitoring image shows that it is in the early stage of pests and diseases, pests and diseases can be controlled in time.

[0072] Specifically, the sensor network accurately obtains the daily tea garden environmental parameters, including temperature, humidity, temperature change rate, daily average light duration, and precipitation. The comprehensive acquisition of these parameters provides a solid data foundation for subsequent analysis and prediction. Through the analysis of these environmental parameters, the background server can grasp the environmental conditions of the tea garden in real time, providing strong support for subsequent growth status prediction and pest and disease prediction. By inputting the tea garden environmental parameters into the trained growth status prediction model, the background server can scientifically predict the daily growth rate of tea trees. This step makes full use of the complex relationship between tea garden environmental parameters and tea tree growth, and realizes high-precision prediction of tea tree growth status through algorithm models (such as long and short neural networks). This not only helps tea garden managers understand the growth status of tea trees, but also provides an important reference for subsequent pest and disease prediction.

[0073] By using time series analysis algorithms (such as the ARIMA model) to analyze tea garden environmental parameters and tea tree growth rates, the backend server can accurately predict the daily incidence of tea garden pests and diseases. This step fully considers the relationship between key factors that cause tea tree pests and diseases, such as the relationship between temperature and humidity, and the relationship between temperature and growth status. Through cross-processing and analysis, the accuracy and timeliness of pest and disease predictions are significantly improved. Based on the prediction results of pest and disease incidence, tea garden managers can formulate prevention and control measures more accurately to reduce the damage of pests and diseases to tea gardens.

[0074] According to the daily incidence of pests and diseases in the tea garden, the backend server can intelligently determine the observation time of the drone. By setting a reasonable threshold for the incidence of pests and diseases and the corresponding observation frequency, the refined management of drone inspections is achieved. During the high incidence period of pests and diseases, the frequency of drone inspections is increased to ensure timely detection and treatment of pests and diseases; during the low incidence period of pests and diseases, the inspection frequency is appropriately reduced to reduce the inspection cost. This data-driven drone inspection strategy not only improves the efficiency of tea garden management, but also significantly reduces labor costs.

[0075] During the observation time of the drone, the drone equipped with a spectrometer can collect spectral images of the tea garden and perform real-time pest and disease identification through pest and disease identification models (such as convolutional neural networks). Not only is the identification speed fast and accurate, but it can also detect and take corresponding prevention and control measures in the early stage of pest and disease. Once the recognition results of the monitoring image show that the tea tree is in the early stage of pest and disease or has already suffered from pest and disease, the background server can immediately generate prevention and control instructions, control the drone equipped with pesticide spraying devices for precise prevention and control, effectively prevent the spread and spread of pests and diseases, and ensure the quality and yield of tea.

[0076] In a possible implementation, S103, the backend server uses a time series analysis algorithm to analyze the daily tea garden environmental parameters and the daily tea tree growth rate to obtain the daily incidence of tea garden pests and diseases, specifically including:

[0077] S201. The backend server obtains the key factor relationships that cause tea tree diseases and insect pests, and cross-processes the daily tea garden environmental parameters and the daily tea tree growth rates according to the key factor relationships to obtain daily cross-information.

[0078] Among them, the influence relationship between the key factors causing tea tree diseases and insect pests was obtained through experimental analysis of tea gardens. For example, there is an influence relationship between temperature and humidity. High temperature and high humidity conditions will induce the occurrence of diseases and insect pests.

[0079] After obtaining the key factor relationships that cause tea tree diseases and insect pests, the daily tea garden environmental parameters and tea tree growth rates are cross-processed according to the key factor relationships to obtain daily cross-information.

[0080] S202. The backend server uses a time series analysis algorithm to analyze the daily cross information to obtain the daily incidence of tea garden pests and diseases.

[0081] The time series analysis algorithm may be a vector autoregression model or a long short-term memory network. By training the vector autoregression model or the long short-term memory network, the vector autoregression model or the long short-term memory network can analyze cross information at multiple moments and obtain the incidence of tea garden pests and diseases at each future time.

[0082] In the above technical scheme, considering that pests and diseases are caused by multiple factors, when using the time series analysis algorithm to analyze the environmental parameters and the growth rate of tea trees, cross-processing is performed according to the relationship between the key factors causing the pests and diseases of tea trees to obtain daily cross-information, and the time series analysis algorithm is used to analyze the daily cross-information to obtain the daily incidence of pests and diseases in the tea garden, thereby improving the accuracy of pest and disease prediction.

[0083] Specifically, the occurrence of tea tree diseases and pests is often affected by a variety of environmental factors, such as temperature, humidity, light, precipitation, and the growth status of the tea tree itself. Through long-term experimental observation and data analysis of tea gardens, the complex relationship between these key factors can be identified. For example, high temperature and high humidity environmental conditions are prone to induce the occurrence of diseases and pests, and there is also a certain correlation between the growth rate of tea trees and the degree of infection of diseases and pests. Accurately obtaining the relationship between these key factors provides a scientific basis for subsequent cross-processing.

[0084] On the basis of obtaining the relationship between key factors, the above technical solution uses a cross-processing method to process the daily tea garden environmental parameters and tea tree growth rate. Cross-processing can refer to combining data of different dimensions according to certain rules to reveal the intrinsic relationship between them. In this example, by multiplying temperature and humidity to obtain temperature and humidity cross-information, and multiplying the temperature change rate with the tea tree growth rate to obtain temperature growth cross-information, the information of tea garden environmental parameters and tea tree growth status is effectively integrated. This cross-processing method can more comprehensively reflect the potential risk of tea tree diseases and pests, and provide a richer data basis for subsequent time series analysis.

[0085] After obtaining the daily cross information, the above technical solution uses a time series analysis algorithm (such as the ARIMA model) to analyze the cross information to predict the daily incidence of tea garden pests and diseases. The time series analysis algorithm is a statistical method specifically used to process and analyze time series data, which can reveal the laws and trends of data changes over time. By applying the time series analysis algorithm, this implementation method can accurately capture the dynamic change characteristics of the incidence of tea garden pests and diseases, and improve the accuracy and timeliness of the prediction.

[0086] Combining the above steps, by obtaining the relationship between key factors, performing cross processing, and applying time series analysis algorithms, the accuracy of tea garden pest and disease prediction has been significantly improved. Compared with traditional prediction methods, the above technical solution not only considers more influencing factors, but also conducts in-depth mining and analysis of complex data relationships through cross processing and time series analysis algorithms. Therefore, it shows higher accuracy and reliability in tea garden pest and disease prediction, providing tea garden managers with a more scientific decision-making basis.

[0087] In a possible implementation, S201, the backend server obtains the key factor relationship causing tea tree diseases and insect pests, processes the daily tea garden environmental parameters and tea tree growth rate according to the key factor relationship, and obtains daily cross information, specifically including:

[0088] S301. If the key factor relationship includes the temperature and humidity relationship, the product of the temperature and humidity of the tea garden is calculated to obtain the daily temperature and humidity cross information.

[0089] The daily tea garden environmental parameters include the temperature and humidity on the first day, the temperature and humidity on the second day, ..., the temperature and humidity on the nth day. The temperature and humidity on the first day are obtained by multiplying the temperature and humidity on the first day, the temperature and humidity on the second day are obtained by multiplying the temperature and humidity on the second day, ..., the temperature and humidity on the nth day are obtained by multiplying the temperature and humidity on the nth day.

[0090] S302: If the key factor relationship includes the temperature change rate growth state relationship, the product of the daily tea garden temperature change rate and the tea tree growth rate is calculated to obtain the daily temperature growth cross information.

[0091] The daily tea garden environmental parameters include the temperature change rate on the first day, the temperature change rate on the second day, ..., the temperature change rate on the nth day. The product of the temperature change rate on the first day and the tea tree growth rate on the first day is calculated to obtain the temperature growth cross information on the first day. The product of the temperature change rate on the second day and the tea tree growth rate on the second day is calculated to obtain the temperature growth cross information on the second day. ... The product of the temperature change rate on the nth day and the tea tree growth rate on the nth day is calculated to obtain the temperature growth cross information on the nth day.

[0092] In the above technical scheme, since the cross-influence of temperature and humidity can lead to the outbreak of pests and diseases, the cross-influence of the temperature change rate of the tea garden and the growth rate of the tea trees can lead to the outbreak of pests and diseases, the product of the temperature and humidity of the tea garden is calculated as the temperature and humidity cross-information, and the product of the temperature change rate of the tea garden and the growth rate of the tea trees is calculated as the temperature-growth cross-information. In this way, the incidence of pests and diseases can be predicted based on the temperature and humidity cross-information and the temperature-growth cross-information, which can improve the prediction accuracy.

[0093] Specifically, the occurrence of tea plant diseases and pests is often affected by a variety of environmental factors, among which temperature and humidity are key factors. Through long-term experimental observation and data analysis, it is found that the cross-effect of temperature and humidity has a significant effect on the outbreak of diseases and pests. In addition, the cross-effect of the temperature change rate of the tea garden and the growth rate of the tea tree cannot be ignored. Accurately identifying the relationship between these key factors provides a scientific basis for subsequent cross-information processing.

[0094] After identifying the key factor relationship, the above technical solution processes the daily tea garden environmental parameters and tea tree growth rate to obtain cross-information. Specifically, if the key factor relationship includes the temperature and humidity relationship, the product of the daily tea garden temperature and humidity is calculated to obtain the daily temperature and humidity cross-information. This calculation method effectively integrates the information of temperature and humidity, reflecting their joint effect on the occurrence of pests and diseases. Similarly, if the key factor relationship includes the temperature-growth state relationship, the product of the daily tea garden temperature change rate and the tea tree growth rate is calculated to obtain the daily temperature-growth cross-information. This calculation method further takes into account the impact of the growth state of the tea tree on the occurrence of pests and diseases, making the cross-information more comprehensive and accurate.

[0095] Based on the calculation of the above cross information, the above technical solution can more accurately predict the incidence of pests and diseases in tea gardens. Since the cross-influence of temperature and humidity and the cross-influence of the temperature change rate of the tea garden and the growth rate of tea trees are important factors leading to the outbreak of pests and diseases, the potential risk of pests and diseases can be more comprehensively reflected by calculating the temperature and humidity cross information and the temperature and growth cross information. Using this cross information as the input of the time series analysis algorithm can more accurately capture the dynamic change characteristics of the incidence of pests and diseases, thereby improving the accuracy of the prediction.

[0096] Specifically, in the process of time series analysis, the algorithm can use the trend, periodicity and seasonality of the temperature-humidity cross-information and the temperature-growth cross-information to make a more accurate prediction of the incidence of pests and diseases. For example, when the temperature-humidity cross-information shows an upward trend, it may indicate that pests and diseases are about to break out; and when the temperature-growth cross-information shows abnormal fluctuations, it may also indicate that the risk of pests and diseases is increasing. By comprehensively analyzing this information, the algorithm can more accurately judge the trend and timing of pests and diseases, and provide more scientific decision-making support for tea garden managers.

[0097] The above technical solutions not only improve the accuracy of pest and disease prediction, but also help promote the intelligent and refined management of tea gardens. By accurately predicting the occurrence of pests and diseases, tea garden managers can formulate prevention and control measures more scientifically, reduce the use of pesticides, reduce prevention and control costs, and improve prevention and control effects. In addition, the formulation of drone intelligent patrol strategies based on prediction results can also achieve refined management of tea garden patrols, improve patrol efficiency and accuracy, and provide more comprehensive technical support for tea garden management.

[0098] In a possible implementation, S202, the backend server uses a time series analysis algorithm to analyze the daily cross information to obtain the daily incidence of tea garden pests and diseases, specifically including:

[0099] S401. The backend server normalizes and vectorizes the daily temperature and humidity cross information and the daily temperature growth cross information to obtain a temperature and humidity vector and a temperature growth vector.

[0100] The temperature and humidity cross information of each day includes the temperature and humidity cross information x1 of the first day, the temperature and humidity cross information x2 of the second day, ..., the temperature and humidity cross information xn of the nth day.

[0101] The temperature and humidity cross information x1 of the first day, the temperature and humidity cross information x2 of the second day, ..., the temperature and humidity cross information xn of the nth day are normalized to obtain the normalized temperature and humidity cross information x1' of the first day, the normalized temperature and humidity cross information x2' of the second day, ..., the normalized temperature and humidity cross information xn' of the nth day.

[0102] The normalized temperature and humidity cross information is vectorized to obtain a temperature and humidity vector X = [x1', x2', ..., xn'].

[0103] The temperature growth cross information of each day includes the temperature growth cross information y1 of the first day, the temperature growth cross information y2 of the second day, ..., the temperature growth cross information yn of the nth day.

[0104] The temperature growth cross information y1 of the first day, the temperature growth cross information y2 of the second day, ..., the temperature growth cross information yn of the nth day are normalized to obtain the normalized temperature growth cross information y1' of the first day, the normalized temperature growth cross information y2' of the second day, ..., the normalized temperature growth cross information yn' of the nth day.

[0105] The normalized temperature growth cross information is vectorized to obtain a temperature growth vector Y = [y1', y2', ..., yn'].

[0106] S402. The backend server uses a fusion coefficient to fuse the temperature and humidity vector and the temperature growth vector to obtain an input vector.

[0107] The backend server uses the fusion coefficient to perform weighted averaging on the temperature and humidity vector and the temperature growth vector to obtain the input vector. More specifically, the fusion coefficient includes a first coefficient and a second coefficient, and the product of the first coefficient and the temperature and humidity vector is calculated, the product of the second coefficient and the temperature growth vector is calculated, and then the sum of the two products is calculated to obtain the input vector.

[0108] S403: The backend server uses the ARIMA model to analyze and process the input vector to obtain the daily incidence of tea garden pests and diseases.

[0109] Among them, since the ARIMA model usually processes univariate time series, after using the fusion coefficient to perform weighted averaging on the temperature and humidity vector and the temperature growth vector, the input vector obtained is a single variable. The ARIMA model is then used to analyze and process the input vector to obtain the daily incidence of tea garden pests and diseases.

[0110] In some examples, the training data includes a training temperature and humidity vector and a training temperature growth vector. The sum of the training temperature and humidity vector and the training temperature growth vector is calculated according to the fusion coefficient to obtain a training input vector. The time series data in the training input vector is tested for stationarity. If it is stationary data, the autocorrelation coefficient and partial autocorrelation function coefficient of the training data are calculated. The autocorrelation function and partial autocorrelation function graph are used to preliminarily determine the order of the AR(p) and MA(q) parts of the model. The parameter values ​​of the model are determined by methods such as maximum likelihood estimation, so as to obtain a trained ARIMA model. The input vector is analyzed and processed using the trained ARIMA model to obtain the incidence of pests and diseases in the tea garden every day.

[0111] In the above technical solution, after obtaining the daily temperature and humidity cross information and the daily temperature growth cross information, the input vector is obtained by fusing the temperature and humidity cross information with the temperature growth cross information. In this way, the cross information of the two variables is fused to obtain the input vector of a single variable. In this way, there is no need to use a multi-vector model for analysis and processing, thereby reducing the complexity of data processing.

[0112] Specifically, the daily temperature and humidity cross information and temperature growth cross information were normalized and vectorized. Normalization is to convert data of different dimensions into the same dimension, so that different data are comparable. Vectorization is to convert scalar data into vector form, which is convenient for subsequent mathematical operations and model processing. Through this step, the temperature and humidity cross information and the temperature growth cross information are converted into temperature and humidity vectors and temperature growth vectors, laying the foundation for subsequent data fusion and analysis. Normalization and vectorization effectively avoid calculation errors caused by inconsistent data dimensions and improve the accuracy and stability of data processing. At the same time, vectorization enables data to be more efficiently calculated and processed in subsequent time series analysis algorithms.

[0113] After obtaining the temperature and humidity vector and the temperature growth vector, the above scheme uses the fusion coefficient to fuse the two vectors to obtain a single variable input vector. The fusion coefficient is obtained through training and can reflect the different degrees of influence of temperature and humidity and temperature growth status on the incidence of pests and diseases. Through weighted summation, the information of the two vectors is effectively fused together to form a more comprehensive and accurate input vector. Data fusion significantly reduces the complexity of data processing. Traditional methods may need to process multiple variables separately and use multi-vector models for analysis, which not only increases the amount of calculation, but also may introduce additional errors. The above scheme integrates the information of multiple variables into one input vector through data fusion, simplifies the subsequent analysis process, and improves processing efficiency.

[0114] After obtaining the input vector, the above scheme uses the ARIMA model to analyze and process the input vector to obtain the daily incidence of tea garden pests and diseases. The ARIMA model is a commonly used time series analysis model that can capture characteristics such as trends, seasonality, and random fluctuations in the data, and accurately predict the incidence of pests and diseases. The application of the ARIMA model improves the accuracy and timeliness of the prediction. Since the input vector has integrated the information of temperature, humidity, and temperature growth status, the ARIMA model can more comprehensively consider the various factors that affect the occurrence of pests and diseases, thereby making more accurate predictions. At the same time, the ARIMA model has strong adaptability and flexibility, and can adjust the prediction results in time according to changes in tea garden environmental parameters, providing more timely decision support for tea garden managers.

[0115] In summary, the above scheme significantly reduces the complexity of data processing through normalization, vectorization, data fusion and the application of ARIMA model. Traditional methods may need to process multiple variables separately and use complex multi-vector models for analysis, which not only increases the amount of calculation but also may introduce additional errors. This implementation method integrates the information of multiple variables into one input vector through a series of optimization steps and uses a simple ARIMA model for analysis and processing, thereby significantly improving the efficiency and accuracy of data processing.

[0116] In a possible implementation, S402, the backend server uses a fusion coefficient to fuse the temperature and humidity vector and the temperature growth vector to obtain an input vector, specifically including:

[0117] S501. The backend server calculates the weighted sum of the temperature and humidity vector and the temperature growth vector according to the first coefficient and the second coefficient to obtain an input vector.

[0118] The fusion coefficient includes the first coefficient λ1 and the second coefficient λ2. The weighted sum of the temperature and humidity vector X = [x1', x2', ..., xn'] and the temperature growth vector Y = [y1', y2', ..., yn'] is calculated to obtain the input vector Z = [λ1×x1'+λ2×y1', λ1×x2'+λ2×y2', ..., λ1×xn'+λ2×yn'].

[0119] The first coefficient, the second coefficient and the parameters in the ARIMA model are obtained through training, and the optimal first coefficient, the optimal second coefficient and the parameters in the ARIMA model are obtained through training.

[0120] In the above technical solution, the first coefficient and the second coefficient are used to calculate the weighted sum of the temperature and humidity vector and the temperature growth vector, so as to convert the time series data of the two variables into the time series data of a single variable. In this way, the time series data processing model of a single variable can be used to analyze the time series data, thereby reducing the number of model parameters and reducing training time.

[0121] Specifically, fusion coefficients (including the first coefficient and the second coefficient) are introduced to fuse the temperature and humidity vector and the temperature growth vector. These fusion coefficients are obtained through training and can accurately reflect the different degrees of influence of temperature and humidity and temperature growth states on the incidence of pests and diseases. By calculating the weighted sum of the temperature and humidity vector and the temperature growth vector, we obtain a single variable input vector that integrates the information of multiple variables. The introduction of the fusion coefficient realizes the effective integration of multivariate data. In traditional methods, it may be necessary to process multiple variables separately, which not only increases the computational complexity, but also may reduce the prediction accuracy due to mutual interference between variables. The above scheme organically integrates the information of multiple variables through the precise calculation of the fusion coefficient to form a more accurate and comprehensive input vector.

[0122] After obtaining the input vector of a single variable, we can use a single variable time series data processing model (such as an ARIMA model) for analysis and processing. Compared with multivariate models, single variable models have the advantages of a small number of parameters and low computational complexity. This means that while maintaining prediction accuracy, we can significantly improve computational efficiency and reduce training time. The use of a single variable input vector significantly reduces the complexity and computational cost of the model. Due to the reduction in the number of model parameters, the model training and optimization process can be completed faster, allowing it to be put into practical application faster. In addition, single variable models are easier to interpret and analyze, which helps tea garden managers better understand and apply prediction results.

[0123] In addition, due to the introduction of fusion coefficients and single variable input vectors, this implementation significantly improves the training efficiency of the model. During the training process, it can converge to the optimal solution faster, thus saving a lot of computing resources and time. This is especially important for application scenarios such as tea garden pest and disease prediction that require rapid response and real-time decision-making. The improvement in model training efficiency allows for faster deployment and updating of prediction models to adapt to changes in tea garden environmental parameters and the evolution of pest and disease occurrence patterns. This helps tea garden managers to formulate and adjust prevention and control measures in a timely manner and improve tea quality and yield.

[0124] It is worth noting that although the above scheme improves computational efficiency by simplifying the model structure and reducing the number of parameters, it does not sacrifice prediction accuracy. Through precise fusion coefficient calculation and effective data fusion strategy, it is still possible to capture the key effects of temperature, humidity and temperature growth status on the incidence of pests and diseases, thereby making accurate predictions.

[0125] In a possible implementation, S102, inputting daily tea garden environmental parameters into a growth state prediction model to obtain daily tea tree growth rate, specifically includes:

[0126] S601. Calculate the product of daily temperature and humidity to obtain daily temperature and humidity, and generate an input matrix according to daily temperature and humidity, temperature change rate, daily average sunshine duration, and precipitation.

[0127] Among them, each column in the input matrix represents the environmental parameters of the tea garden for one day, and each row in the input matrix represents an environmental parameter.

[0128] Among them, the input matrix A is expressed as:

[0129]

[0130] a 11 represents the temperature and humidity on the first day, a 12 represents the temperature and humidity on the second day, ..., a 1n Indicates the temperature and humidity on the nth day. 21 represents the temperature change rate on the first day, b 22 represents the temperature change rate on the second day, ..., b 2n Indicates the temperature change rate on the nth day. c 31 represents the average daily light duration on the first day, c 32 represents the average daily light duration on the second day, ..., c 3n Indicates the average daily sunshine duration on day n. 41 represents the precipitation on the first day, d 42 represents the precipitation on the second day, ..., d 4n Represents the precipitation on day n.

[0131] S602, inputting the input matrix into the trained long-short neural network, so that the trained long-short neural network outputs the daily growth rate of the tea trees.

[0132] The long-short neural network is trained so that the long-short neural network can output the growth rate of the tea tree. The input matrix is ​​input into the trained long-short neural network so that the trained long-short neural network outputs the growth rate of the tea tree every day.

[0133] In the above technical solution, after obtaining the tea garden environmental parameters, considering the combined effects of temperature and humidity on tea tree growth, the product of daily temperature and humidity is calculated to obtain daily temperature and humidity, and then an input matrix is ​​generated based on daily temperature and humidity, temperature change rate, daily average light duration, and precipitation. Each column in the input matrix represents a tea garden environmental parameter for a day, and each row in the input matrix represents an environmental parameter. In this way, the trained long-short neural network can be used to predict the growth rate of the tea trees, thereby improving the prediction accuracy.

[0134] Specifically, we first considered the combined effects of temperature and humidity on tea tree growth, and obtained the daily temperature and humidity by calculating the product of daily temperature and humidity. This step fully considered the interaction of temperature and humidity as important environmental factors for tea tree growth, providing a more accurate data basis for subsequent predictions.

[0135] Next, an input matrix was generated based on the daily temperature and humidity, temperature change rate, daily average sunshine duration, and precipitation. Each column in the input matrix represents a tea garden environmental parameter for one day, and each row represents an environmental parameter. This matrix format not only facilitates data storage and processing, but also preserves the correlation and time dependence between environmental parameters, providing strong support for subsequent long-short neural network analysis. The construction of the input matrix achieves a comprehensive integration and orderly organization of tea garden environmental parameters, laying a solid foundation for subsequent data analysis and model prediction.

[0136] Inputting the input matrix into the trained long-short neural network is the core step of the above scheme. As a special recurrent neural network (RNN), the long-short neural network (LSTM) has the ability to process long time series data and can capture the time dependency and complex patterns in the data. In this application scenario, LSTM can make full use of the time series characteristics of tea garden environmental parameters to learn the complex relationship between tea tree growth rate and environmental parameters. Through the output of LSTM, the daily tea tree growth rate can be obtained. Compared with traditional methods, the LSTM model can better handle the complexity and dynamics of tea garden environmental parameters, and improve the accuracy and robustness of prediction. The application of long-short neural networks significantly improves the prediction accuracy of tea tree growth rate. Through its powerful time series data processing capabilities, the LSTM model captures the complex relationship between tea garden environmental parameters and tea tree growth rate, providing tea garden managers with a more reliable decision-making basis.

[0137] In addition to the above steps, the above scheme further improves the accuracy of prediction by considering the joint effects of temperature and humidity on tea tree growth, generating input matrices, and applying long-short neural networks. Specifically, the product calculation of temperature and humidity fully considers the interaction between temperature and humidity, the construction of the input matrix retains the correlation and time dependence between environmental parameters, and the application of long-short neural networks makes full use of these characteristics to improve the accuracy and robustness of prediction.

[0138] In a possible implementation, S104, determining the observation time of the drone according to the daily incidence of tea garden pests and diseases, specifically includes:

[0139] S701, let i traverse from 1 to n, determine whether the incidence of tea garden pests and diseases on the i-th day is greater than the first threshold, if so, determine the observation frequency of the drone to be the first frequency, and determine the observation time of the drone on the i-th day according to the first frequency.

[0140] S702: If not, determine whether the incidence of tea garden pests and diseases on the i-th day is greater than the second threshold; if so, determine the observation frequency of the drone to be the second frequency, and determine the observation time of the drone on the i-th day according to the second frequency.

[0141] Wherein, i and n are both positive integers, the first threshold is greater than the second threshold, and the first frequency is greater than the second frequency.

[0142] In the first loop, it is determined whether the incidence of tea garden pests and diseases on the first day is greater than or equal to the first threshold. If so, the observation frequency of the drone is determined to be the first frequency, and the observation time of the drone on the first day is determined according to the first frequency.

[0143] In the second loop, it is determined whether the incidence of tea garden pests and diseases on the second day is greater than or equal to the first threshold. If not, it is determined whether the incidence of tea garden pests and diseases on the second day is greater than or equal to the second threshold. The observation frequency of the drone is determined to be the second frequency, and the observation time of the drone on the second day is determined according to the second frequency.

[0144] In the third loop, it is determined whether the incidence of tea garden pests and diseases on the third day is greater than or equal to the first threshold. If not, it is determined whether the incidence of tea garden pests and diseases on the third day is greater than or equal to the second threshold. If not, the observation frequency of the drone is determined to be zero.

[0145] And so on, until the incidence of tea garden pests and diseases on all days is traversed and the observation time of the drone is determined.

[0146] In the above technical solution, by setting the first threshold and the second threshold, when the incidence of pests and diseases is greater than the first threshold, the observation time of the drone on the day is determined by the first frequency, and when the incidence of pests and diseases is greater than the second threshold, the observation time of the drone on the day is determined by the second frequency. Different collection frequencies are used under different incidences of pests and diseases, so that pests and diseases can be discovered in time and the patrol cost of the drone can be reduced.

[0147] Specifically, two key thresholds are set first: the first threshold and the second threshold, to distinguish different levels of pest and disease incidence. When the pest and disease incidence in the tea garden on the i-th day exceeds the higher first threshold, the system responds immediately and sets the drone's observation frequency to the higher first frequency. This mechanism ensures that during periods of high pest and disease risk, drones can conduct more frequent aerial observations, thereby timely discovering and tracking the dynamics of pests and diseases, and providing key information support for subsequent prevention and control work.

[0148] If the incidence of pests and diseases on the i-th day does not reach the first threshold, but exceeds the lower second threshold, the solution adopts a milder response strategy and adjusts the observation frequency of the drone to the lower second frequency. This hierarchical treatment method not only avoids unnecessary high-frequency flights during the period when pests and diseases are relatively controllable, but also ensures that necessary monitoring activities can be carried out continuously, effectively improving the utilization efficiency of monitoring resources.

[0149] By setting different thresholds and corresponding observation frequencies, the above scheme achieves refined management and response to different incidence rates of pests and diseases. High-frequency monitoring is used during the high-incidence period to ensure the timeliness and accuracy of information acquisition; during the low-incidence period, the frequency is appropriately reduced to reduce the number of drones used, thereby effectively controlling operating costs. This dynamic adjustment strategy not only meets the urgent needs of pest and disease monitoring, but also takes into account the economic considerations of long-term operation.

[0150] From the perspective of technical implementation, the above solution can determine the observation time and frequency of the drone through simple conditional judgment logic (i.e. judging the relationship between the incidence of pests and diseases and the threshold), without the need for complex algorithms or model support. This design is not only easy to understand and implement, but also can maintain high operating efficiency and stability in practical applications, which is convenient for wide promotion and application in agricultural scenarios such as tea gardens.

[0151] In a possible implementation, S105, the backend server uses a pest identification model to identify pests and diseases on the spectral image to obtain tea tree health status data, specifically including:

[0152] S801. The backend server obtains the growth stage of each day, and obtains the growth stage corresponding to the acquisition time of the spectral image as the target growth stage.

[0153] Among them, the growth process of tea trees in one year is statistically analyzed to obtain the growth stage of each day in a year. For example, the growth stages of tea trees include spring germination, the first dormant period, summer growth period, the second dormant period, autumn growth period and winter dormant period. Obtain the dates covered by the spring germination period, the first dormant period, the summer growth period, the second dormant period, the autumn growth period and the winter dormant period.

[0154] The growth stage corresponding to the acquisition time of the spectral image is obtained as the target growth stage. For example, if the spectral image is acquired on March 12, it is determined that the target growth stage corresponding to the spectral image is the spring germination period.

[0155] S802: The backend server determines a pest and disease identification model corresponding to the target growth stage according to the target growth stage and the model mapping table.

[0156] Among them, the model mapping table represents the mapping relationship between the identification of the pest and disease identification model at different growth stages and the growth stage identification. After obtaining the target growth stage corresponding to the spectral image, the identification of the target growth stage is used to query the mapping relationship table to obtain the identification of the pest and disease identification model corresponding to the target growth stage.

[0157] S803. The backend server uses the pest and disease identification model corresponding to the target growth stage to identify pests and diseases on the spectral image to obtain tea tree health status data.

[0158] Among them, the pest and disease identification model corresponding to the target growth stage is run in the background server, and the pest and disease identification model is used to identify pests and diseases on the spectral image to obtain the health status data of the tea tree.

[0159] In the above technical scheme, after obtaining the spectral image collected by the drone, the growth stage corresponding to the spectral image is determined according to the collection time of the spectral image, and the pest and disease identification model of the corresponding stage is obtained. The spectral image is used to identify pests and diseases using the pest and disease identification model corresponding to the target growth stage to obtain the health status data of the tea tree. Since the tree shapes of tea trees at different growth stages are different, each growth stage corresponds to a pest and disease identification model, which can improve the accuracy of the pest and disease identification model.

[0160] Specifically, by determining the growth stage of the tea tree every day, the life cycle of the tea tree can be managed in a refined manner. This step provides a basic framework for subsequent pest and disease identification, because tea trees at different growth stages often have different physiological characteristics and the types of pests and diseases they are susceptible to. By collecting the specific time of the spectral image, the system can automatically match and determine the target growth stage, laying a precise time dimension foundation for the subsequent analysis.

[0161] Next, the above scheme introduces a model mapping table, which maps each growth stage of the tea tree to the corresponding pest and disease identification model. This design ensures that the most suitable model for the current growth stage can be selected when identifying pests and diseases. Through this precise matching, the inaccuracy that may be caused by the traditional method of using a single model to process all growth stages is avoided, thereby significantly improving the pertinence and efficiency of identification.

[0162] The spectral image is analyzed using the pest and disease identification model corresponding to the target growth stage. The pest and disease identification model specific to each growth stage is trained and optimized based on the specific characteristics of the tea tree at that stage and the common pest and disease types, so it can more accurately identify the signs of pests and diseases in the spectral image. This customized processing method fully considers the variability of the tea tree growth cycle and significantly improves the accuracy of pest and disease identification.

[0163] Finally, through the above series of refined operations, the system can output data reflecting the current health status of tea trees. These data not only help to timely discover potential pests and diseases, but also provide a scientific basis for subsequent precise prevention and control, thereby optimizing tea production management and ensuring tea quality and yield.

[0164] Some embodiments of the present application provide an information processing method for a tea garden, the information processing method specifically comprising the following steps:

[0165] S901. The backend server obtains the environmental parameters of the tea garden every day.

[0166] This step has been described in detail in the above embodiment and will not be repeated here.

[0167] S902. The backend server inputs the daily tea garden environmental parameters into the growth state prediction model to obtain the daily tea tree growth rate.

[0168] This step has been described in detail in the above embodiment and will not be repeated here.

[0169] S903. The backend server uses a time series analysis algorithm to analyze the daily tea garden environmental parameters and the daily tea tree growth rate to obtain the daily incidence of tea garden pests and diseases.

[0170] This step has been described in detail in the above embodiment and will not be repeated here.

[0171] S904, obtaining training images of the tea garden at different growth stages, and using the training images at the same growth stage to train the convolutional neural network to obtain pest and disease recognition models at different stages.

[0172] Among them, training images of the spring germination period, training images of the first dormant period, training images of the summer growth period, training images of the second dormant period, training images of the autumn growth period and training images of the winter dormant period are obtained.

[0173] The convolutional neural network is trained using the training images of the spring germination period to obtain the pest and disease recognition model for the spring germination period. The convolutional neural network is trained using the training images of the first dormant period to obtain the pest and disease recognition model for the first dormant period. The convolutional neural network is trained using the training images of the summer growth period to obtain the pest and disease recognition model for the summer growth period. The convolutional neural network is trained using the training images of the second dormant period to obtain the pest and disease recognition model for the second dormant period. The convolutional neural network is trained using the training images of the autumn growth period to obtain the pest and disease recognition model for the autumn growth period. The convolutional neural network is trained using the training images of the winter dormant period to obtain the pest and disease recognition model for the winter dormant period.

[0174] S905: Generate a model mapping table according to the model identifiers of the pest and disease identification models at different growth stages and the stage identifiers of different growth stages.

[0175] Among them, a mapping relationship between the identifier of the spring germination period and the identifier of the pest and disease identification model for the spring germination period is established, a mapping relationship between the identifier of the first dormant period and the identifier of the pest and disease identification model for the first dormant period is established, a mapping relationship between the identifier of the summer growing period and the identifier of the pest and disease identification model for the summer growing period is established, a mapping relationship between the identifier of the second dormant period and the identifier of the pest and disease identification model for the second dormant period is established, a mapping relationship between the identifier of the autumn growing period and the identifier of the pest and disease identification model for the autumn growing period is established, and a mapping relationship between the identifier of the winter dormancy period and the identifier of the pest and disease identification model for the winter dormancy period is established.

[0176] S906. The backend server determines the observation time according to the daily incidence of pests and diseases in the tea garden, and controls the drone equipped with a spectrometer to collect spectral images of the tea garden at the observation time.

[0177] This step has been described in detail in the above embodiment and will not be repeated here.

[0178] S907. The backend server determines the growth stage of the tea tree every day according to the growth rate of the tea tree every day, and obtains the growth stage corresponding to the acquisition time of the spectral image as the target growth stage.

[0179] For example, the spectral image is collected on March 12, and the target growth stage corresponding to the spectral image is determined to be the spring germination stage.

[0180] S908. The backend server determines a pest and disease identification model corresponding to the target growth stage according to the target growth stage and the model mapping table.

[0181] Among them, the identification of the spring germination period is used to search the model mapping table to obtain the identification of the pest and disease identification model of the spring germination period, the storage area of ​​the pest and disease identification model is determined according to the identification of the pest and disease identification model of the spring germination period, the pest and disease identification model of the spring germination period is read from the storage area, and the pest and disease identification model of the spring germination period is run.

[0182] S909. The backend server uses the pest and disease identification model corresponding to the target growth stage to identify pests and diseases on the spectral image to obtain tea tree health status data.

[0183] Among them, the background server uses the pest and disease identification model of the spring germination period to identify pests and diseases on the spectral images and obtain the health status data of the tea trees.

[0184] S910. If the health status data indicates that the tea plant is in the early stage of pest and disease occurrence or when pest and disease occurrence occurs, the pest and disease type is obtained, and a prevention and control instruction is generated according to the pest and disease type, so that the drone is equipped with a corresponding pesticide spraying device to patrol the tea garden and spray pesticides.

[0185] In S910, the system can accurately determine whether the tea trees are in the early stages of pests and diseases or have already been attacked by pests and diseases. This capability is based on the model's high-precision recognition of subtle features in spectral images, ensuring early detection of pest and disease signs and providing the possibility for timely intervention.

[0186] Once the tea tree's health status is confirmed to be abnormal, the system will further identify the specific type of pests and diseases. This step is crucial because different types of pests and diseases require different prevention and control measures. Accurate identification can ensure the pertinence and effectiveness of subsequent prevention and control measures.

[0187] Based on the identified pest and disease types, the system automatically generates targeted prevention and control instructions. This process is highly automated, avoiding the subjectivity and delay of human judgment, and ensuring the scientificity and timeliness of the prevention and control strategy.

[0188] After the command is generated, the system controls the drone to carry the corresponding pesticide spraying device, quickly responds and executes the prevention and control task. The drone's patrol and precise spraying in the tea garden not only greatly improves the operation efficiency and reduces labor costs, but also reduces pesticide waste and environmental pollution through precise application, which is in line with the concept of sustainable development of modern agriculture.

[0189] Combining the above steps, this technical solution significantly improves the efficiency of tea tree pest control through the seamless connection of intelligent identification, command generation and drone automated operation. From early warning to rapid response, and then to precise treatment, a closed-loop, efficient pest control system has been formed, providing a strong guarantee for the healthy growth of tea trees.

[0190] After obtaining the health status data, it is determined whether the health status data is in the early stage of the occurrence of pests and diseases or the occurrence of pests and diseases, and the pest and disease type is extracted from the health status data, and prevention and control instructions are generated by encoding the pest and disease type.

[0191] The backend server sends the prevention and control instructions to the drone. After receiving the prevention and control instructions, the drone parses the type of pests and diseases from the instructions, and carries a pesticide spraying device according to the type of pests and diseases, injects the corresponding pesticide into the pesticide spraying device, and the drone patrols the tea garden and sprays pesticides.

[0192] In the above technical solution, since the tree shapes are different at different growth stages, in order to reduce the influence of the tree shape on the recognition accuracy of the pest and disease recognition model, training images of each growth stage are obtained, and the convolutional neural network is trained using the training images of one growth stage to obtain pest and disease recognition models at different growth stages. In this way, when the recognition model is used to recognize spectral images, the recognition accuracy of the model is improved.

[0193] Specifically, we first obtain training images of tea gardens at different growth stages. This step is based on the scientific understanding of the changes in morphological and physiological characteristics of tea trees during their growth cycle. Tea trees at different growth stages (such as seedlings, growth stages, and mature stages) have significant differences in crown morphology, leaf size, and color. These differences directly affect the capture content and quality of spectral images. By using training images from the same growth stage to conduct targeted training on the convolutional neural network, a pest and disease recognition model specifically for that stage can be generated. This approach fully considers the image features unique to the growth stage, which helps the model learn the manifestations of pests and diseases that are closely related to the current stage, thereby improving the accuracy of recognition.

[0194] The training images of all growth stages are traversed, and the pest and disease recognition models of each stage are trained in turn to ensure that the model library can fully cover the entire process of tea tree growth. This means that a matching recognition model can be found at any growth stage, greatly improving the practicality and flexibility of the system.

[0195] In order to efficiently manage and apply these pest and disease identification models for different growth stages, an innovative model mapping table is proposed. This table associates the unique identifier of the model (such as model ID) with the stage identifier of the corresponding growth stage (such as growth period, maturity period, etc.), so that in practical applications, the corresponding identification model can be quickly found according to the current growth stage of the tea tree. This design not only simplifies the process of model selection and application, but also greatly improves the processing speed and efficiency, allowing pest and disease monitoring to respond more promptly, which is crucial for timely detection and control of pests and diseases.

[0196] Through phased model training, the interference caused by changes in the shape of tea trees at different growth stages on the identification of pests and diseases is effectively reduced. The model of each stage is trained based on the image features unique to that stage, thus reducing misjudgments caused by morphological differences while improving the robustness and reliability of the recognition system.

[0197] Figure 3 This is a schematic diagram of the structure of the information processing device for tea gardens provided in this application. Figure 3 As shown, the information processing device 200 for a tea garden provided in this embodiment includes:

[0198] An acquisition module 210 is used to acquire tea garden environmental parameters every day, wherein the tea garden environmental parameters include temperature, humidity, temperature change rate, daily average sunshine duration and precipitation;

[0199] The processing module 220 is used to input the daily tea garden environmental parameters into the growth state prediction model to obtain the daily tea tree growth rate;

[0200] The processing module 220 is further used to analyze the daily tea garden environmental parameters and the daily tea tree growth rate using a time series analysis algorithm to obtain the daily incidence of tea garden pests and diseases;

[0201] The processing module 220 is further used to determine the observation time of the drone according to the daily incidence rate of pests and diseases in the tea garden, and control the drone to carry a spectrometer to collect spectral images of the tea garden at the observation time of the drone;

[0202] The processing module 220 is further used to use a pest and disease identification model to identify pests and diseases on the spectral image to obtain tea tree health status data.

[0203] The background server provided in this embodiment includes: at least one processor and a memory. Optionally, the device also includes a communication component. The processor, the memory and the communication component are connected via a bus.

[0204] In a specific implementation process, at least one processor executes computer-executable instructions stored in a memory, so that at least one processor executes the above method.

[0205] The specific implementation process of the processor can be found in the above method embodiment, and its implementation principle and technical effect are similar, so this embodiment will not be repeated here.

[0206] In the above embodiments, it should be understood that the processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), etc. A general-purpose processor may be a microprocessor or any conventional processor. The steps of the method disclosed in the invention may be directly implemented as being executed by a hardware processor, or may be executed by a combination of hardware and software modules in the processor.

[0207] The memory may include a high-speed memory (Random Access Memory, RAM), and may also include a non-volatile memory (Non-volatile Memory, NVM), such as at least one disk memory.

[0208] The bus may be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, the bus in the drawings of the present application is not limited to only one bus or one type of bus.

[0209] The present application also provides a computer program product, including a computer program, which implements the above method when executed by a processor.

[0210] The present application also provides a computer-readable storage medium, in which computer-executable instructions are stored. When a processor executes the computer-executable instructions, the above method is implemented.

[0211] The above-mentioned readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk. The readable storage medium can be any available medium that can be accessed by a general or special-purpose computer.

[0212] An exemplary readable storage medium is coupled to a processor so that the processor can read information from the readable storage medium and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can be located in an application specific integrated circuit (Application Specific Integrated Circuits, referred to as: ASIC). Of course, the processor and the readable storage medium can also exist in the device as discrete components.

[0213] The division of units is only a logical function division, and there may be other divisions in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interface, device or unit, which can be electrical, mechanical or other forms.

[0214] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0215] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0216] If the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods of each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0217] Those skilled in the art can understand that all or part of the steps of implementing the above-mentioned method embodiments can be completed by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, the steps of the above-mentioned method embodiments are executed; and the aforementioned storage medium includes: ROM, RAM, disk or optical disk and other media that can store program codes.

[0218] Finally, it should be noted that those skilled in the art will readily conceive of other embodiments of the present invention after considering the specification and practicing the invention disclosed herein. The present invention is intended to cover any variations, uses or adaptations of the present invention, which follow the general principles of the present invention and include common knowledge or customary technical means in the art not disclosed by the present invention, are not limited to the precise structure described above and shown in the drawings, and may be modified and changed in various ways without departing from the scope thereof. The scope of the present invention is limited only by the appended claims.

Claims

1. An information processing method for a tea garden, characterized in that: The information processing method is applied to a backend server, and the information processing method includes: Obtaining daily tea garden environmental parameters, wherein the tea garden environmental parameters include temperature, humidity, temperature change rate, daily average sunshine duration, and precipitation; Inputting the daily tea garden environmental parameters into a growth state prediction model to obtain the daily tea tree growth rate; Using a time series analysis algorithm to analyze the daily tea garden environmental parameters and the daily tea tree growth rate, to obtain the daily incidence of tea garden pests and diseases; Determine the observation time of the drone according to the daily incidence of pests and diseases in the tea garden, and control the drone to carry a spectrometer to collect spectral images of the tea garden at the observation time; The pest and disease identification model is used to identify pests and diseases on the spectral image to obtain tea tree health status data.

2. The information processing method for a tea garden according to claim 1, characterized in that: The daily tea garden environmental parameters and the daily tea tree growth rate are analyzed using a time series analysis algorithm to obtain the daily incidence of tea garden pests and diseases, specifically including: Obtaining the key factor relationship causing tea tree diseases and insect pests, and cross-processing the daily tea garden environmental parameters and the daily tea tree growth rate according to the key factor relationship to obtain daily cross-information; The daily cross information is analyzed using a time series analysis algorithm to obtain the daily incidence of tea garden pests and diseases.

3. The information processing method for tea garden according to claim 2, characterized in that: Obtain the key factor relationship that causes tea tree diseases and insect pests, process the daily tea garden environmental parameters and tea tree growth rate according to the key factor relationship, and obtain daily cross information, specifically including: If the key factor relationship includes the temperature and humidity relationship, the product of the temperature and humidity of the tea garden is calculated to obtain the daily temperature and humidity cross information; If the key factor relationship includes the temperature-growth state relationship, the product of the daily tea garden temperature change rate and the tea tree growth rate is calculated to obtain the daily temperature-growth cross information.

4. The information processing method for a tea garden according to claim 3, characterized in that: The daily cross information is analyzed using a time series analysis algorithm to obtain the daily incidence of tea garden pests and diseases, including: Normalize and vectorize the daily temperature and humidity cross information and the daily temperature growth cross information to obtain the temperature and humidity vector and the temperature growth vector; Using a fusion coefficient to fuse the temperature and humidity vector and the temperature growth vector to obtain an input vector; The input vector is analyzed and processed using the ARIMA model to obtain the daily incidence of tea garden pests and diseases.

5. The information processing method for tea garden according to claim 4, characterized in that: The temperature and humidity vector and the temperature growth vector are fused using a fusion coefficient to obtain the input vector, specifically including: Calculate the weighted sum of the temperature and humidity vector and the temperature growth vector according to the first coefficient and the second coefficient to obtain the input vector; The fusion coefficient includes a first coefficient and a second coefficient, and the first coefficient, the second coefficient and the parameters in the ARIMA model are obtained through training.

6. The information processing method for a tea garden according to any one of claims 1 to 5, characterized in that: Using a pest and disease identification model to identify pests and diseases on the spectral image to obtain the health status data of the tea tree specifically includes: Determine the growth stage of the tea tree every day, and obtain the growth stage corresponding to the acquisition time of the spectral image as the target growth stage; Determine a pest and disease identification model corresponding to the target growth stage according to the target growth stage and the model mapping table; The pest and disease identification model corresponding to the target growth stage is used to identify pests and diseases on the spectral image to obtain the health status data of the tea tree.

7. The information processing method for a tea garden according to any one of claims 1 to 5, characterized in that: After using the pest and disease identification model to identify pests and diseases on the spectral image to obtain the tea tree health status data, the method further includes: If the health status data indicates that the tea plantation is in the early stage of disease and pest occurrence or that disease and pest occurrence has occurred, the type of disease and pest is obtained, and a prevention and control instruction is generated according to the type of disease and pest, so that the drone is equipped with a corresponding pesticide spraying device to patrol the tea garden and spray pesticides.

8. An information processing device for a tea garden, characterized in that: include: An acquisition module is used to acquire tea garden environmental parameters every day, wherein the tea garden environmental parameters include temperature, humidity, temperature change rate, daily average sunshine duration and precipitation; A processing module, used for inputting the daily tea garden environmental parameters into a growth state prediction model to obtain a daily tea tree growth rate; The processing module is further used to analyze the daily tea garden environmental parameters and the daily tea tree growth rate using a time series analysis algorithm to obtain the daily incidence of tea garden pests and diseases; The processing module is further used to determine the observation time of the drone according to the daily incidence of pests and diseases in the tea garden, and control the drone to carry a spectrometer to collect spectral images of the tea garden at the observation time of the drone; The processing module is also used to use a pest and disease identification model to identify pests and diseases on the spectral image to obtain tea tree health status data.

9. A backend server, characterized in that: include: processor; as well as, A memory, configured to store executable instructions of the processor; The processor is configured to perform the method of any one of claims 1 to 7 by executing the executable instructions.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, which are used to implement the method according to any one of claims 1 to 7 when executed by a processor.

Images