A method and device for predicting plant disease rate

Abstract

The present application provides a method and device for predicting plant disease rate, including: obtaining remote sensing data of target plants and vegetation coverage at the jointing stage corresponding to multiple pixel areas respectively; determining the initial prediction time corresponding to multiple pixel areas respectively according to the remote sensing data of target plants corresponding to multiple pixel areas respectively; obtaining daily meteorological data and daily date related information corresponding to multiple pixel areas respectively; determining the total planting quantity, initial infection quantity and daily disease prevalence rate corresponding to multiple pixel areas respectively according to the remote sensing data of target plants, vegetation coverage at the jointing stage, initial prediction time, daily meteorological data and daily date related information corresponding to multiple pixel areas respectively; determining the daily disease rate of target plants corresponding to multiple pixel areas respectively according to the total planting quantity, initial infection quantity, initial prediction time and daily disease prevalence rate corresponding to multiple pixel areas respectively. The present application can realize the prediction of disease rate that is continuous in surface and continuous in time.

Description

Technical Field

[0001] The present application relates to the field of plant disease epidemiology, and in particular to a method and device for predicting plant disease rate. Background Art

[0002] Some current plant diseases are characterized by rapid onset, preventability, controllability, and irreversibility. Their high-precision, high-frequency dynamic prediction is conducive to the accurate and green prevention and control of diseases.

[0003] At present, scholars at home and abroad have established an agronomic-meteorological model between meteorological factors and disease information based on multi-year observation data. This model can analyze the key meteorological factors of the disease epidemic, such as temperature, precipitation, rainy days, wind direction, etc. during the critical growth period, which improves the temporal resolution of the model, but still has the following shortcomings: First, the current point data-driven, agricultural-meteorological combined spatiotemporal extensive-scale early warning model is limited by the spatial distribution of the data, and lacks consideration of the impact of host conditions on the occurrence and development of the disease; Second, the current research on disease prediction combined with meteorological data has failed to analyze the epidemic process of the disease. Summary of the invention

[0004] In view of this, the present application provides a method and device for predicting plant disease rate, which are used to solve the above technical problems. The technical solution is as follows:

[0005] A method for predicting plant disease rate, comprising:

[0006] Obtaining remote sensing data of target plants and vegetation coverage at the jointing stage corresponding to multiple pixel areas, wherein the vegetation coverage at the jointing stage refers to the vegetation coverage under the remote sensing image when the target plants are at the jointing stage;

[0007] According to the remote sensing data of the target plants corresponding to the multiple pixel regions, the greening periods of the target plants corresponding to the multiple pixel regions are determined as the initial prediction times corresponding to the multiple pixel regions;

[0008] Obtaining daily meteorological data and daily date related information corresponding to a plurality of pixel regions respectively, wherein the daily meteorological data and daily date related information refer to the daily meteorological data and daily date related information within a time range consisting of an initial prediction time and a future time to be predicted;

[0009] According to the target plant remote sensing data corresponding to the multiple pixel areas, the vegetation coverage during the jointing period, the initial prediction time, the daily meteorological data and the daily date related information, the total planting quantity, the initial infection quantity and the daily disease prevalence rate of the target plants corresponding to the multiple pixel areas are determined, wherein the target plants are in any one of the healthy, latent, infected and removed infected states, the initial infection quantity refers to the number of target plants in the infected state at the time of initial prediction, the disease prevalence rate includes the infection rate, the first conversion rate and the second conversion rate, the infection rate refers to the probability that the target plant in the healthy state is infected by the target plant in the infected state and becomes the target plant in the latent state, the first conversion rate refers to the conversion rate when the target plant in the latent state is converted to the target plant in the infected state, and the second conversion rate refers to the conversion rate when the target plant in the infected state is converted to the target plant in the removed infected state;

[0010] The daily diseased plant rates of the target plants corresponding to the multiple pixel areas are determined according to the total planting quantity, initial infection quantity, initial prediction time and daily disease prevalence rate of the target plants corresponding to the multiple pixel areas.

[0011] A plant disease rate prediction device, comprising:

[0012] The first basic data acquisition module is used to obtain remote sensing data of target plants and vegetation coverage at the jointing stage corresponding to multiple pixel areas, wherein the vegetation coverage at the jointing stage refers to the vegetation coverage under the remote sensing image when the target plant is at the jointing stage;

[0013] An initial prediction time determination module is used to determine the greening period of the target plants corresponding to the multiple pixel areas respectively according to the remote sensing data of the target plants corresponding to the multiple pixel areas respectively, as the initial prediction time corresponding to the multiple pixel areas respectively;

[0014] The second basic data acquisition module is used to obtain daily meteorological data and daily date related information corresponding to a plurality of pixel areas, wherein the daily meteorological data and daily date related information refer to the daily meteorological data and daily date related information within a time range consisting of an initial prediction time and a future time to be predicted;

[0015] a process data determination module, for determining the total planting quantity, initial infection quantity and daily disease prevalence rate of target plants corresponding to the plurality of pixel areas respectively according to the target plant remote sensing data corresponding to the plurality of pixel areas respectively, the vegetation coverage during the jointing period, the initial prediction time, the daily meteorological data and the relevant information of the daily date, wherein the target plants are in any one of the healthy, latent, infected and removed infected states, the initial infection quantity refers to the quantity of target plants in the infected state at the time of the initial prediction, the disease prevalence rate includes the infection rate, the first conversion rate and the second conversion rate, the infection rate refers to the probability that a target plant in a healthy state is infected by a target plant in an infected state and becomes a target plant in a latent state, the first conversion rate refers to the conversion rate when a target plant in a latent state is converted to a target plant in an infected state, and the second conversion rate refers to the conversion rate when a target plant in an infected state is converted to a target plant in a removed infected state;

[0016] The diseased plant rate prediction module is used to determine the daily diseased plant rate of target plants corresponding to multiple pixel areas according to the total planting quantity, initial infection quantity, initial prediction time and daily disease prevalence rate of target plants corresponding to multiple pixel areas.

[0017] It can be known from the above technical scheme that the plant disease rate prediction method provided by the present application first obtains the remote sensing data of the target plants and the vegetation coverage in the jointing stage corresponding to multiple pixel areas respectively, and then determines the greening period of the target plants corresponding to the multiple pixel areas respectively according to the remote sensing data of the target plants corresponding to the multiple pixel areas respectively, as the initial prediction time, and then obtains the daily meteorological data and daily date-related information corresponding to the multiple pixel areas respectively, and then determines the total planting number, initial infection number and daily disease epidemic rate of the target plants corresponding to the multiple pixel areas respectively according to the remote sensing data of the target plants, the vegetation coverage in the jointing stage, the initial prediction time, the daily meteorological data and the daily date-related information, and finally determines the daily disease rate of the target plants corresponding to the multiple pixel areas respectively according to the total planting number, initial infection number, initial prediction time and daily disease epidemic rate of the target plants corresponding to the multiple pixel areas respectively. The present application combines meteorological data and remote sensing data to predict the disease rate of target plants corresponding to multiple pixel areas, achieving accurate prediction of the disease rate on a continuous surface. In addition, the present application can realize continuous disease rate prediction within a time range consisting of an initial prediction time and a future time to be predicted, thereby analyzing the epidemic process of the disease and improving the prediction accuracy of the disease rate. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0019] Figure 1 A schematic diagram of a process for predicting plant disease rate provided in an embodiment of the present application;

[0020] Figure 2 A schematic diagram of a normal distribution curve for determining vegetation coverage at the jointing stage provided in an embodiment of the present application;

[0021] Figure 3a is α RH , β RH and 1-γ RH Schematic diagram of the response curves to humidity respectively;

[0022] Figure 3b is α T , β T and 1-γ T Schematic diagram of the response curves to temperature respectively;

[0023] Figure 3c is α age , β age and 1-γ age Schematic diagram of the response curves to the growth period respectively;

[0024] Figure 4 Schematic diagram of the first-order sensitivity and total effect sensitivity corresponding to 25 parameters;

[0025] Figure 5 It is a schematic diagram of the optimization process for three groups of sensitivity parameters;

[0026] Figure 6 To use 11 vegetation indices to I Schematic diagram of the average RMSE obtained by initializing the training model;

[0027] Figure 7 is the RMSE and R obtained based on the training set and test set 2 Schematic diagram of

[0028] Figure 8 This is a schematic diagram of the number of target plants under four conditions obtained by continuous prediction using the disease mechanism model;

[0029] Fig. 9 A schematic diagram of the structure of a plant disease rate prediction device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0030] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0031] This application provides a method for predicting the plant disease rate. The following examples will describe the method in detail. Figure 1 , shows a schematic flow chart of a method for predicting a plant disease rate provided in an embodiment of the present application, and the method for predicting a plant disease rate may include:

[0032] Step S101, obtaining remote sensing data of target plants and vegetation coverage at the jointing stage corresponding to a plurality of pixel regions respectively.

[0033] Among them, the vegetation coverage at the jointing stage refers to the vegetation coverage under remote sensing images when the target plants are in the jointing stage.

[0034] In this step, a remote sensing image of the target plant planted in the area to be predicted can be obtained. The data contained in the remote sensing image is the remote sensing data of the target plant. It can be understood that the remote sensing image contains multiple pixels, so the area to be predicted can be divided into multiple pixel areas, and each pixel area corresponds to the remote sensing data of the target plant. For example, if the resolution of the remote sensing image is 20, the pixel area in this step can be 20*20m 2 area.

[0035] It should be noted that the target plant remote sensing data obtained in this step refers to the continuous remote sensing data of the target plant within the growth period, and the target plant remote sensing data obtained in this step refers to the target plant remote sensing data collected most recently in history. Here, the growth period of the target plant includes the following phenological periods: wintering period, greening period, jointing period, heading period and flowering period.

[0036] For example, assuming that the growth period of the target plant is from October 1st of each year to June 1st of the following year, and assuming that the current time is April 1st, 2022, the remote sensing data of the target plant obtained in this step are the remote sensing data from October 1st, 2021 to April 1st, 2022, and the remote sensing data from April 2nd, 2021 to June 1st, 2021.

[0037] The above-mentioned vegetation coverage in the jointing stage is the vegetation coverage under the remote sensing image when the target plant is in the jointing stage. Those skilled in the art should understand that the vegetation coverage can be calculated based on the remote sensing data of the target plant. Since the vegetation coverage calculated based on the remote sensing data of the target plant is already available in the prior art, there is no need to calculate again in this step, but the vegetation coverage in the jointing stage can be directly obtained according to the time of the jointing stage in the historical data.

[0038] Optionally, the target plant can be wheat, and this step can predict the daily incidence of scab of wheat (i.e., diseased plant rate). Of course, the target plant in this embodiment can also be other, and this application does not limit this.

[0039] Step S102: determining the greening period of the target plants corresponding to the plurality of pixel regions respectively according to the remote sensing data of the target plants corresponding to the plurality of pixel regions respectively, as the initial prediction time corresponding to the plurality of pixel regions respectively.

[0040] This step can determine the greening period of the target plant corresponding to each pixel area according to the target value plant remote sensing data corresponding to each pixel area, as the initial prediction time corresponding to the pixel area.

[0041] In an optional embodiment, for each pixel area in the plurality of pixel areas, the process of determining the greening period of the target plant corresponding to the pixel area may include:

[0042] Step a1: Determine the vegetation index time series curve of the target plant corresponding to the pixel area according to the remote sensing data of the target plant corresponding to the pixel area.

[0043] Optionally, the vegetation index in this step may be NDVI (Normalized Difference Vegetation Index), that is, this step may determine an NDVI time series curve of the target plant corresponding to the pixel area, and the NDVI time series curve represents the change of the daily NDVI value over time.

[0044] In this step, the formula To determine the daily NDVI value.

[0045] Where NDVI represents the daily NDVI value, ρ nir represents the daily near-infrared reflectivity, ρ red represents the daily red band reflectance. Here ρ nir and ρ red That is the remote sensing data of the target plant corresponding to the pixel area.

[0046] Step a2: Smoothing the vegetation index time series curve of the target plant corresponding to the pixel area to obtain a smoothed vegetation index time series curve corresponding to the pixel area.

[0047] Optionally, in this step, the Whittaker smoothing algorithm may be used to smooth the MODIS (Moderate Resolution Imaging Spectroradiometer) daily NDVI time series curve of the target plant corresponding to the pixel area.

[0048] Step a3: Fit the smoothed vegetation index time series curve corresponding to the pixel area by using the logistic function to obtain a fitting curve corresponding to the pixel area.

[0049] Here, the logistic function refers to the Logistic function. In this step, the smoothed vegetation index time series curve corresponding to the pixel area can be fitted by the Logistic function to obtain a fitting curve.

[0050] Step a4, calculating the second-order curvature of the fitting curve corresponding to the pixel area, and determining the maximum point of the second-order curvature, and taking the time corresponding to the maximum point as the greening period of the target plant corresponding to the pixel area.

[0051] Specifically, this step can calculate the second-order curvature of the fitting curve, extract the maximum value point of the second-order curvature, and correspond to the greening period of the target plant.

[0052] This step can use the greening period as the initial prediction time, that is, regardless of whether the current time is the greening period of the target plant, this step starts the prediction from the greening period.

[0053] Through the above steps a1 to a4, the greening period of the target plant corresponding to each pixel area can be obtained, thereby obtaining the greening period of the target plant corresponding to multiple pixel areas, that is, obtaining the initial prediction time corresponding to multiple pixel areas.

[0054] Step S103: Obtain daily meteorological data and daily date related information corresponding to a plurality of pixel regions respectively.

[0055] The daily meteorological data and daily date related information refer to the daily meteorological data and daily date related information within a time range consisting of an initial prediction time and a future time to be predicted.

[0056] For example, the greening period of the target plant corresponding to a pixel area is February 1, 2022, and the current time is April 1, 2022. This embodiment needs to predict the daily disease rate of the target plant from April 2, 2022 to June 1, 2022. Then, this step can obtain the daily meteorological data and daily date-related information from February 1, 2022 to June 1, 2022. Among them, for the daily meteorological data from February 1, 2022 to April 1, 2022, the actual meteorological data can be used, and for the daily meteorological data from April 2, 2022 to June 1, 2022, the predicted meteorological data can be used.

[0057] Optionally, the above-mentioned daily date related information can be DOY (day of year) information, and the DOY information can indicate the position of the specified date in a year. For example, the DOY information corresponding to January 1 of each year is 1, the DOY information corresponding to January 2 of each year is 2, and so on.

[0058] Step S104, determining the total planting quantity, initial infection quantity and daily disease prevalence rate of target plants corresponding to the multiple pixel areas respectively according to the remote sensing data of target plants corresponding to the multiple pixel areas respectively, the vegetation coverage during the jointing period, the initial prediction time, the daily meteorological data and the daily date related information.

[0059] In this step, the target plant is in any one of the following states during its entire growth period, wherein the healthy state is recorded as the S state; the latent state is recorded as the E state, and when the target plant is in the latent state, it means that the target plant has been infected by the pathogen, but is in the latent period and is not contagious; the infectious state is recorded as the I state, and when the target plant is in the infectious state, it can infect the target plant in the healthy state; the removed infectious state is recorded as the R state, and when the target plant is in the removed infectious state, it is no longer contagious, for example, the target plant dies, or the pathogen is removed artificially, etc.

[0060] The above-mentioned initial infection number refers to the number of target plants in the infection state at the time of initial prediction, that is, the number of target plants in the infection state at the initial prediction time. For example, this step can start prediction when the target plants are in the greening period, then the initial infection number refers to the number of target plants in the infection state during the greening period.

[0061] The above-mentioned disease prevalence rate includes an infection rate, a first conversion rate and a second conversion rate, wherein the infection rate refers to the probability that a target plant in a healthy state is infected by a target plant in an infected state and becomes a target plant in a latent state, the first conversion rate refers to the conversion rate when a target plant in a latent state is converted into a target plant in an infected state, and the second conversion rate refers to the conversion rate when a target plant in an infected state is converted into a target plant in a state of removing the infection.

[0062] Step S105, determining the daily diseased plant rate of the target plants corresponding to the multiple pixel areas respectively according to the total planting quantity, initial infection quantity, initial prediction time and daily disease prevalence rate of the target plants corresponding to the multiple pixel areas respectively.

[0063] It is understandable that when pathogens exist in the target plants, the disease will usually begin to occur in large numbers after a certain period of time (it is more practical to start prediction after this period), and the initial infection number and total planting number during the initial prediction, as well as the daily disease prevalence rate will all have an impact on the disease situation of the target plants. Therefore, this step is based on the total planting number, initial infection number, initial prediction time and daily disease prevalence rate of the target plants corresponding to multiple pixel areas, respectively, and can accurately determine the daily disease rate of the target plants corresponding to multiple pixel areas.

[0064] The plant disease rate prediction method provided in the present application first obtains remote sensing data of target plants and vegetation coverage at the jointing stage corresponding to multiple pixel areas, then determines the greening period of target plants corresponding to the multiple pixel areas according to the remote sensing data of target plants corresponding to the multiple pixel areas as the initial prediction time, then obtains daily meteorological data and daily date-related information corresponding to the multiple pixel areas, then determines the total planting number, initial infection number and daily disease epidemic rate of target plants corresponding to the multiple pixel areas according to the remote sensing data of target plants, vegetation coverage at the jointing stage, initial prediction time, daily meteorological data and daily date-related information, finally determines the daily disease rate of target plants corresponding to the multiple pixel areas according to the total planting number, initial infection number, initial prediction time and daily disease epidemic rate of target plants corresponding to the multiple pixel areas. The present application combines meteorological data and remote sensing data to predict the disease rate of target plants corresponding to multiple pixel areas, achieving accurate prediction of the disease rate on a continuous surface. In addition, the present application can realize continuous disease rate prediction within a time range consisting of an initial prediction time and a future time to be predicted, thereby analyzing the epidemic process of the disease and improving the prediction accuracy of the disease rate.

[0065] An embodiment of the present application describes the process of the aforementioned "step S104, determining the total planting quantity, initial infection quantity and daily disease prevalence rate of target plants corresponding to multiple pixel areas respectively according to the remote sensing data of target plants corresponding to multiple pixel areas respectively, the vegetation coverage during the jointing period, the initial prediction time, the daily meteorological data and the daily date-related information".

[0066] Optionally, the daily meteorological data includes daily temperature data and daily humidity data. Based on this, the process of step S104 may include:

[0067] Step b1: determining the total planting quantity of target plants corresponding to the plurality of pixel areas respectively according to the vegetation coverage at the jointing stage corresponding to the plurality of pixel areas respectively.

[0068] Specifically, the process of this step may include the following steps b11 to b12:

[0069] Step b11: determining a linear regression coefficient according to the vegetation coverage at the jointing stage corresponding to each of the plurality of pixel regions.

[0070] For ease of description, the vegetation coverage at the jointing stage is expressed as FVC. jointing express.

[0071] In this embodiment, the target plant has no invalid tillers at the jointing stage, and the field is not closed. After the FVC corresponding to the multiple pixel areas is jointing The statistical results showed that it was normally distributed. Assuming that 95% of FVC jointing Corresponding to the government-guided unit target plant number range (e.g., 300-375 per square meter), this step uses linear regression to calculate the FVC to plant number, which can be calculated based on the 95% FVC jointing The two points in the corresponding normal distribution curve (see Figure 2 A and B shown, Figure 2 The shaded area shown is 95% of FVC jointing ), determine the linear regression coefficient N par1 and N par2 .

[0072] Step b12: determining the total planting quantity of target plants corresponding to the plurality of pixel areas respectively according to the linear regression coefficient and the vegetation coverage at the jointing stage corresponding to the plurality of pixel areas respectively.

[0073] Specifically, this embodiment assumes that in each pixel area, the total number of target plants planted is always a constant value N, that is, no new target plants are planted after the prediction starts, and the number of target plants removed as a whole can be ignored.

[0074] For each pixel region in the plurality of pixel regions, this step can adopt the formula N=(N par1 ×FVC jointing ×N par2 )×s 2 , to determine the total number of target plants corresponding to the pixel area, where FVC jointing It refers to the vegetation coverage at the jointing stage corresponding to the pixel area, s 2 Indicates the area of ​​the pixel region, for example, s 2 Can be 20 2 .

[0075] Step b2: determining the initial infection numbers of the target plants corresponding to the multiple pixel areas respectively according to the remote sensing data and daily temperature data of the target plants corresponding to the multiple pixel areas respectively, combined with the initial prediction times corresponding to the multiple pixel areas respectively.

[0076] Optionally, for each pixel area among the multiple pixel areas, the process of determining the initial infection number of the target plant corresponding to the pixel area in this step may include the following steps b21 to b24:

[0077] Step b21, determining the greening vegetation index corresponding to the pixel area according to the target plant remote sensing data and the initial prediction time corresponding to the pixel area.

[0078] Optionally, the green vegetation index in this step may be an optimized soil-adjusted vegetation index (OSAVI).

[0079] Here, the calculation formula of OSAVI is:

[0080] Therefore, according to the initial prediction time (i.e., the greening period of the target plant), the near-infrared band reflectivity ρ of the greening period can be determined from the remote sensing data of the target plant corresponding to the pixel area. nir and red band reflectivity ρ red , so the above-mentioned OSAVI calculation formula can be used to calculate the green vegetation index corresponding to the pixel area.

[0081] Of course, the green vegetation index in this step may also be other types of vegetation indexes, which is not limited in this application.

[0082] Step b22: Obtain the vegetation coverage in the green period corresponding to the pixel area according to the initial prediction time corresponding to the pixel area.

[0083] Among them, the vegetation coverage in the green period refers to the vegetation coverage under remote sensing images when the target plants are in the green period.

[0084] After the greening period of the target plant corresponding to the pixel area is determined in the aforementioned step, this step can obtain the greening period vegetation coverage corresponding to the pixel area from the existing vegetation coverage data.

[0085] Step b23: Calculate the average temperature data from the wintering period to the greening period of the target plant based on the daily temperature data corresponding to the pixel area, as the average temperature data corresponding to the pixel area.

[0086] Considering that the rate of pathogen infection on the target plant is related to the early growth environment and host of the target plant, especially the overwintering temperature and the growth of the vegetative growth nodes affect the pathogen concentration, based on this, this step can also calculate the average temperature data of the target plant from the overwintering period to the greening period to determine the initial infection number based on the average temperature data.

[0087] Step b24, determining the initial infection number of the target plants corresponding to the pixel areas respectively according to the average temperature data, the green vegetation index and the green vegetation coverage corresponding to the pixel areas respectively.

[0088] Specifically, for each pixel area in multiple pixel areas, this step can use the formula To determine the initial infection number of the target plant corresponding to the pixel area.

[0089] In the formula, Init I Indicates the initial infection number corresponding to the pixel area, VI green Indicates the vegetation index of the green period corresponding to the pixel area, FVC green Indicates the vegetation coverage of the pixel area during the green period, T overwinter Represents the average temperature data corresponding to the pixel area, where n, m and k are constants.

[0090] Step b3: determining the flowering period of the target plants corresponding to the plurality of pixel regions respectively according to the remote sensing data of the target plants and the daily temperature data corresponding to the plurality of pixel regions respectively.

[0091] Optionally, for each pixel area among the multiple pixel areas, the process of determining the flowering period of the target plant corresponding to the pixel area in this step may include the following steps b31 to b34:

[0092] Step b31: Determine the vegetation index time series curve of the target plant corresponding to the pixel area according to the remote sensing data of the target plant corresponding to the pixel area.

[0093] The process of determining the vegetation index time series curve in this step is the same as the process in the aforementioned step a1. For details, please refer to the introduction in the aforementioned steps, which will not be repeated here.

[0094] Step b32: Smoothing the vegetation index time series curve of the target plant corresponding to the pixel area to obtain a smoothed vegetation index time series curve corresponding to the pixel area.

[0095] The process of determining the vegetation index time series curve in this step is the same as the process in the aforementioned step a2. For details, please refer to the introduction in the aforementioned steps, which will not be repeated here.

[0096] Step b33: determine the maximum point from the smoothed vegetation index time series curve corresponding to the pixel area, and the time corresponding to the maximum point is used as the heading period of the target plant corresponding to the pixel area.

[0097] Step b34: Determine the flowering period of the target plant corresponding to the pixel area according to the heading period and daily temperature data of the target plant corresponding to the pixel area.

[0098] Specifically, this step may start with accumulating the temperature from the heading stage of the target plant until a preset temperature threshold is reached to obtain the flowering stage (also referred to as the blooming stage) of the target plant.

[0099] It should be noted that the above temperature threshold can be set according to actual conditions. For example, in this embodiment, the temperature threshold is 105°C.

[0100] Step b4: determine the daily disease prevalence rates of target plants corresponding to the multiple pixel areas respectively according to the daily temperature data, daily humidity data and daily date related information corresponding to the multiple pixel areas respectively, combined with the flowering periods corresponding to the multiple pixel areas respectively.

[0101] Optionally, the process of this step may include the following steps b41 to b46:

[0102] Step b41, based on the daily humidity data corresponding to the multiple pixel areas, and the pre-established response curves of the infection rate, the first conversion rate and the second conversion rate to humidity, determine the response values ​​of the daily infection rate, the first conversion rate and the second conversion rate corresponding to the multiple pixel areas to humidity, as the daily first response value, the second response value and the third response value, respectively.

[0103] Optionally, a Logistic curve may be used to simulate the response curves of the infection rate, the first conversion rate, and the second conversion rate to humidity, respectively.

[0104] In this embodiment, in order to ensure the positive driving effect of humidity on the prevalence of diseases, the infection rate α RH and the first conversion rate β RH Both are positively correlated with relative humidity RH, and the second conversion rate γ RHIt is inversely correlated with relative humidity RH, that is, 1-γ RH It is positively correlated with relative humidity RH.

[0105] In this step, the formula for the humidity response curve is: RH (x) = 1 / e (-x+δ) / τ .

[0106] Among them, x represents humidity data, for the infection rate α RH , f RH (x) represents the response curve of infection rate to humidity. For the first conversion rate β RH , f RH (x) represents the response curve of the first conversion rate to humidity. RH , f RH (x) represents the response curve of the second conversion rate to humidity; δ and τ represent two parameters in the Logistic curve. RH and β RH ,δ=δ αβ , τ=τ αβ ; For 1-γ RH ,δ=δ γ , τ=τ γ .

[0107] See also Figure 3a As shown, α RH , β RH and 1-γ RH Schematic diagram of the response curves to humidity. For given δ and τ, this step can input the daily humidity data corresponding to each pixel area into the humidity response curve corresponding to the pixel area, and obtain the daily response value of the infection rate to humidity corresponding to the pixel area as the first response value of the day; similarly, the daily response value of the first conversion rate to humidity corresponding to the pixel area can be obtained as the second response value of the day, and the daily response value of the second conversion rate to humidity corresponding to the pixel area can be obtained as the third response value of the day.

[0108] Step b42: According to the daily temperature data corresponding to the multiple pixel areas, and the pre-established response curves of the infection rate, the first conversion rate and the second conversion rate to the temperature, the response values ​​of the daily infection rate, the first conversion rate and the second conversion rate corresponding to the multiple pixel areas to the temperature are determined as the fourth response value, the fifth response value and the sixth response value of the day, respectively.

[0109] Optionally, a Gaussian curve may be used to simulate the response curves of the infection rate, the first conversion rate, and the second conversion rate to temperature. It is worth noting that in this embodiment, different variances are used on both sides of the mean of the Gaussian curve.

[0110] In this embodiment, the infection rate α T , the first conversion rate β T The relationship between the relative temperature T and the second conversion rate γ T The relationship with the relative temperature T shows an opposite trend, that is, the infection rate α T , the first conversion rate β T , 1-γ T The relationship with the relative temperature T shows the same trend.

[0111] In this step, the formula for the temperature response curve is:

[0112] Among them, x represents temperature data, for the infection rate α T , f T (x) represents the response curve of infection rate to temperature. For the first conversion rate β T , f T (x) represents the response curve of the first conversion rate to temperature. T , f T (x) represents the response curve of the second conversion rate to temperature; μ, σ T1 and σ T2 Represents the three parameters in the Gaussian curve. For α T and β T , μ=μ αβ , σ T1 =σ T1_αβ , σ T2 =σ T2_αβ ; For 1-γ T , μ=μ γ , σ T1 =σ T1_γ , σ T2 =σ T2_γ .

[0113] See also Figure 3b As shown, α T , β T and 1-γ T Schematic diagram of the response curve to temperature. For a given μ, σ T1 and σ T2 In this step, the daily temperature data corresponding to each pixel area can be input into the temperature response curve corresponding to the pixel area to obtain the daily response value of the infection rate to the temperature corresponding to the pixel area as the fourth response value of the day; similarly, the daily response value of the first conversion rate to the temperature corresponding to the pixel area can be obtained as the fifth response value of the day, and the daily response value of the second conversion rate to the temperature corresponding to the pixel area can be obtained as the sixth response value of the day.

[0114] Step b43, based on the daily date-related information and flowering period corresponding to the multiple pixel areas, and the pre-established response curves of the infection rate, the first conversion rate and the second conversion rate to the growth period, determine the response values ​​of the daily infection rate, the first conversion rate and the second conversion rate corresponding to the multiple pixel areas to the growth period, as the seventh response value, the eighth response value and the ninth response value of the day, respectively.

[0115] Optionally, a Gaussian curve may be used to simulate the response curves of the infection rate, the first conversion rate and the second conversion rate to the growth period, respectively. It is worth noting that in this embodiment, different variances are used on both sides of the mean of the Gaussian curve.

[0116] In this embodiment, the infection rate α age , the first conversion rate β age The relationship between the reproductive age and the second conversion rate γ age The relationship with the reproductive age is in the opposite trend, that is, the infection rate α age , the first conversion rate β age , 1-γ age The relationship with reproductive age shows the same trend.

[0117] In this step, the formula for the growth period response curve is:

[0118] Where t0 = DOY anthesis +k,DOY anthesis represents the flowering period, k represents the time bias, t0 represents the mean under the growth period response curve; x represents date related information (such as DOY), for the infection rate α age , f age (x) represents the response curve of infection rate to reproductive period. For the first conversion rate β age , f age (x) represents the response curve of the first conversion rate to the growth period. age , f age (x) represents the response curve of the second conversion rate to the growth period; k, σ A1 and σ A2 Represents the three parameters in the Gaussian curve. For α age , k=k α , σ A1 =σ A1_α , σ A2 =σ A2_α ; For β age , k=k β , σ A1 =σ A1_β , σ A2 =σ A2_β ; For 1-γ age, k=k γ , σ A1 =σ A1_γ , σ A2 =σ A2_γ .

[0119] See also Figure 3c As shown, α age , β age and 1-γ age Schematic diagram of the response curves for the growth period. A1 and σ A2 In this step, the daily growth period data and flowering period DOY corresponding to each pixel area can be anthesis Input it into the growth period response curve corresponding to the pixel area, and obtain the daily infection rate response value to the growth period corresponding to the pixel area as the seventh response value of the day; similarly, the daily first conversion rate response value to the growth period corresponding to the pixel area can be obtained as the eighth response value of the day, and the daily second conversion rate response value to the growth period corresponding to the pixel area can be obtained as the ninth response value of the day.

[0120] Step b44: determining the daily infection rates of the target plants corresponding to the plurality of pixel regions respectively according to the daily first response value, the fourth response value and the seventh response value.

[0121] Optionally, for each pixel region in the plurality of pixel regions, this step may adopt the formula α(t)=α0×α T (t)×α RH (t)×α age (t) to determine the infection rate of the target plant on day t corresponding to the pixel area.

[0122] Among them, α(t) represents the infection rate on day t, α RH (t) represents the first response value on the tth day, α T (t) represents the fourth response value on day t, α age (t) represents the seventh response value on the tth day, α0 is a constant, and represents the infection rate adjustment value (optimal infection rate) under the optimal state.

[0123] Step b45: determining the daily first conversion rates of the target plants corresponding to the plurality of pixel areas respectively according to the daily second response value, the fifth response value and the eighth response value.

[0124] Optionally, for each pixel region in the plurality of pixel regions, this step may use the formula To determine the first conversion rate of the target plant corresponding to the pixel area on the tth day.

[0125] Among them, β(t) represents the first conversion rate on day t, βRH (t) represents the second response value on day t, β T (t) represents the fifth response value on day t, β age (t) represents the eighth response value on the tth day, β0 is a constant, and represents the first conversion rate adjustment value (optimal EI conversion rate) under the optimal state.

[0126] Step b46: Determine the daily second conversion rates of the target plants corresponding to the plurality of pixel areas respectively according to the daily third response value, the sixth response value and the ninth response value.

[0127] Optionally, for each pixel region in the plurality of pixel regions, this step may use the formula To determine the second conversion rate of the target plant corresponding to the pixel area on the tth day.

[0128] Among them, γ(t) represents the second conversion rate on the tth day, γ RH (t) represents the third response value on day t, γ T (t) represents the sixth response value on day t, γ age (t) represents the ninth response value on the tth day, γ0 is a constant, and represents the second conversion rate adjustment value (optimal IR conversion rate) in the optimal state.

[0129] In summary, through the above steps b41 to b46, the daily infection rate, the first conversion rate and the second conversion rate of the target plants corresponding to the multiple pixel areas can be obtained, that is, the daily disease prevalence rate of the target plants corresponding to the multiple pixel areas can be obtained.

[0130] Another embodiment of the present application describes the process of the aforementioned "step S105, determining the daily disease rate of target plants corresponding to the multiple pixel areas respectively according to the total planting quantity, initial infection quantity, initial prediction time and daily disease prevalence rate of the target plants respectively corresponding to the multiple pixel areas."

[0131] Taking any one of the multiple pixel areas as an example, the following provides a process of "determining the daily disease rate of the target plant corresponding to the pixel area according to the total planting number, initial infection number, initial prediction time and daily disease prevalence rate of the target plant corresponding to the pixel area."

[0132] Optionally, the process of determining the daily disease rate of the target plant corresponding to the pixel area may include:

[0133] Step c1: Determine the number of target plants in healthy, latent, infectious and removed infectious states in the pixel area on a daily basis according to the total planting number, initial infection number, initial prediction time and daily disease prevalence rate of the target plants corresponding to the pixel area.

[0134] Optionally, this step can be based on the total planted number of target plants corresponding to the pixel area, the initial infection number, the initial prediction time and the daily disease prevalence rate, combined with the following formulas (1) to (4), to determine the number of target plants in the pixel area that are in the healthy, latent, infectious and removed infection states on a daily basis.

[0135]

[0136]

[0137]

[0138]

[0139] In the above formula, represents the change in the number of target plants in a healthy state on day t, which represents the difference in the number of target plants in a healthy state on day t and day t+1; represents the change in the number of target plants in the latent state on day t, which represents the difference in the number of target plants in the latent state on day t and day t+1; represents the change in the number of target plants in the infectious state on day t, which represents the difference in the number of target plants in the infectious state on day t and day t+1; It represents the change in the number of target plants in the infection removal state on the t day, which represents the difference in the number of target plants in the infection removal state on the t day and the t+1 day.

[0140] In the above formula, S(t) represents the number of target plants in a healthy state on the tth day, E(t) represents the number of target plants in a latent state on the tth day, I(t) represents the number of target plants in an infectious state on the tth day, α(t) represents the infection rate on the tth day, β(t) represents the first conversion rate on the tth day, γ(t) represents the second conversion rate on the tth day, and N represents the total number of target plants planted corresponding to the pixel area.

[0141] After obtaining the change in the number of target plants that are respectively in the healthy, latent, infectious and removed infection states on the t day, this embodiment adds the change in number to the number of target plants that are respectively in the healthy, latent, infectious and removed infection states on the t day, and the number of target plants that are respectively in the healthy, latent, infectious and removed infection states on the t+1 day can be obtained.

[0142] That is, the prediction process of this step is predicted in the order of days. For example, assuming that the target plant in each pixel area is in a healthy state or an infectious state during the initial prediction, the initial infection number Init is determined.I After summing the total number of plants planted, N, we can get the number of target plants in a healthy state corresponding to each pixel area, that is, Init S =N-Init I Assuming that the initial prediction time is February 1, 2022, if you want to predict the number of target plants in healthy, latent, infectious, and removed infection states on May 1, 2022, first set Init S 、Init I , N, and the infection rate, first conversion rate, and second conversion rate corresponding to the initial prediction time (February 1, 2022) are substituted into formulas (1) to (4) to obtain the changes in the number of target plants in the healthy, latent, infectious, and removed infection states on February 1, 2022, thereby obtaining the number of target plants in the healthy, latent, infectious, and removed infection states on February 2, 2022; then, the number of target plants in the healthy, latent, infectious, and removed infection states on February 2, 2022, N, and the infection rate, first conversion rate, and second conversion rate corresponding to February 2, 2022 are substituted into formulas (1) to (4) to obtain the changes in the number of target plants in the healthy, latent, infectious, and removed infection states on February 2, 2022, thereby obtaining the number of target plants in the healthy, latent, infectious, and removed infection states on February 3, 2022; and so on, finally obtaining the number of target plants in the healthy, latent, infectious, and removed infection states on May 1, 2022.

[0143] Step c2: determining the daily disease rate of the target plants corresponding to the pixel area according to the number of target plants in the healthy, latent, infectious and removed infectious states in the pixel area on a daily basis.

[0144] Optionally, this step may use formula (5) to determine the daily disease rate of the target plant corresponding to the pixel area.

[0145]

[0146] Where DI(t) represents the disease rate of the target plant on day t.

[0147] Through the above steps c1 to c2, the daily diseased plant rates of the target plants corresponding to the multiple pixel areas can be obtained.

[0148] In another embodiment of the present application, a disease mechanism model can be constructed to implement the above-mentioned step b24 based on the disease mechanism model: "determine the initial infection number of target plants corresponding to multiple pixel areas respectively according to the average temperature data, green-return vegetation index and green-return vegetation coverage corresponding to multiple pixel areas respectively", "step b4, determine the daily disease epidemic rate of target plants corresponding to multiple pixel areas respectively according to the daily temperature data, daily humidity data and daily date-related information corresponding to multiple pixel areas respectively, combined with the flowering period corresponding to multiple pixel areas respectively", and "step S105, determine the daily diseased plant rate of target plants corresponding to multiple pixel areas respectively according to the total planting number, initial infection number, initial prediction time and daily disease epidemic rate of target plants corresponding to multiple pixel areas respectively".

[0149] Here, the disease mechanism model is a SEIR (Susceptible-Exposed-Infectious-Removed) model, which uses training daily temperature data, training daily humidity data, training daily date-related information, training average temperature data, training green vegetation index, training green vegetation coverage, training flowering period, total number of training target plants planted and training initial prediction time corresponding to multiple pixel areas as training samples, and is trained with the diseased plant rate corresponding to the labeled training samples as training labels.

[0150] Therefore, the process of "determining the initial infection number of target plants corresponding to multiple pixel areas respectively according to the average temperature data, green vegetation index and green vegetation coverage corresponding to multiple pixel areas respectively, "step b4, determining the daily disease epidemic rate of target plants corresponding to multiple pixel areas respectively according to the daily temperature data, daily humidity data and daily date related information corresponding to multiple pixel areas respectively, combined with the flowering period corresponding to multiple pixel areas respectively", and "step S105, determining the daily disease rate of target plants corresponding to multiple pixel areas respectively according to the total planting number, initial infection number, initial prediction time and daily disease epidemic rate of target plants corresponding to multiple pixel areas respectively" in step b24 may include: using the pre-acquired disease mechanism model to process the daily temperature data, daily humidity data, daily date related information, average temperature data, green vegetation index, green vegetation coverage, flowering period, total planting number and initial prediction time corresponding to multiple pixel areas respectively, to obtain the daily disease rate of target plants contained in the multiple pixel areas output by the disease mechanism model.

[0151] In an optional embodiment, the process of obtaining the disease mechanism model (SEIR model) in this embodiment may include the following steps d1 to d4:

[0152] Step d1: Obtain task processing models corresponding to a plurality of pre-built vegetation indices.

[0153] Optionally, see Table 1 below, which shows multiple vegetation indices in this step, their types and calculation formulas.

[0154] Table 1 Multiple vegetation indices, types and calculation formulas

[0155]

[0156]

[0157] In the above Table 1, ρ λ Corresponding to the reflectivity of the band containing wavelength λ or near wavelength λ, ρ nir ,ρ red and ρ green is the reflectance of near infrared, red band and green band. From Table 1 above, this step can obtain the task processing models corresponding to 11 vegetation indices.

[0158] Step d2: For each task processing model in the task processing models corresponding to multiple vegetation indices, a global sensitivity analysis method is used to analyze the sensitivity of the parameters in the task processing model, so as to divide the parameters in the task processing model into a first sensitivity parameter group, a second sensitivity parameter group whose sensitivity is lower than the first sensitivity parameter group, and a third sensitivity parameter group whose sensitivity is lower than the second sensitivity parameter group.

[0159] Among them, the parameters in the task processing model include parameters in the response curves of the infection rate, the first conversion rate and the second conversion rate to humidity, respectively, parameters in the response curves of the infection rate, the first conversion rate and the second conversion rate to temperature, respectively, parameters in the response curves of the infection rate, the first conversion rate and the second conversion rate to the growth period, and parameters in the initial infection number calculation formula.

[0160] In this embodiment, the model parameters are mainly derived from the epidemiological research results of the target plant diseases. However, due to the differences in target plant varieties, pathogen species, research scales, research objects, etc., the reference values ​​in the literature cannot be directly applied to the study area and have certain deviations. Therefore, it is necessary to evaluate the range of parameter values ​​based on existing knowledge and perform parameter calibration, that is, to train the model. The parameter training problem of the model is converted into an optimization problem of bounded parameters. In order to successfully and effectively apply the optimization method, the sensitivity of the parameters to the model output results is first analyzed, and the parameters are divided into several groups accordingly.

[0161] Specifically, the parameters in each task processing model and related information of the parameters are shown in Table 2 below.

[0162] Table 2 Parameters in the task processing model and related information of the parameters

[0163]

[0164]

[0165] As shown in Table 2 above, each task processing model in this embodiment includes 25 parameters. This step can use the global sensitivity analysis method Sobol's to analyze the sensitivity of the parameters in the task processing model, and divide the 25 parameters into three groups. For example, the high sensitivity group (the first sensitivity parameter group) includes: α0, β0, k α , σ A1_α , σ A2_α , k β , σ A1_β , σ A2_β , k γ , σ A1_γ , σ A2_γ , τ αβ , σ T1_αβ ; The medium sensitivity group (second sensitivity parameter group) includes: α0, β0, γ0, δ αβ , τ αβ , μ αβ , σ T1_αβ , σ T2_αβ ; Low sensitivity group (third sensitivity parameter group) includes: α0, β0, γ0, δ γ , τ γ , μ γ , σ T1_γ , σ T2_γ , n, m, k.

[0166] Optionally, the process of dividing into three parameter groups may include the following steps d21 to d22:

[0167] Step d21: for each parameter in the task processing model, input the training sample into the task processing model to obtain the diseased plant rate output by the task processing model, and calculate the variance of the diseased plant rate output by the task processing model, and determine the first-order sensitivity, second-order sensitivity and total effect sensitivity corresponding to the parameter based on the parameter and the variance, so as to obtain the first-order sensitivity, second-order sensitivity and total effect sensitivity corresponding to all the parameters in the task processing model.

[0168] Specifically, in the field of plant disease epidemiology, the Sobol analysis method has been widely used to analyze the key factors affecting the spread of pathogens. Based on the Sobol analysis method, the SEIR model is considered as a function:

[0169] Among them, y is the model output, that is, the disease rate DI, x i represents the i-th parameter of the model, f ij For x i and x j The equation, d = 25, then the Sobol analysis method output is the following formula (6) to formula (8):

[0170]

[0171]

[0172] s Ti =s i +s ij(i≠j) +…+s 1…i…d Formula (8)

[0173] Where D(y) is the variance of y, D i is the function f i The variance, D ij f ij The partial variance, s i For x i The first-order sensitivity of , reflecting the i-th parameter x i The size of the decision on the output result, s ij For x i and x j The second-order sensitivity of s reflects the impact of the interaction between the two on the output results. Ti For x i The total effect sensitivity reflects x i The effect of all interactions with other parameters on the results.

[0174] In this step, the above formula (1) to formula (5) can be used to obtain the diseased plant rate corresponding to the training sample output by the task processing model, and the variance of the diseased plant rate can be calculated. Then, the first-order sensitivity, second-order sensitivity and total effect sensitivity corresponding to all parameters can be calculated according to formulas (6) to (8). Figure 4 As shown, it is a schematic diagram of the first-order sensitivity and total effect sensitivity corresponding to the 25 parameters calculated in this embodiment. For example, in this step, Sobol's sequence method can be used to generate 100,000 groups of samples to analyze the sensitivity of the 25 parameters to the dependent variable.

[0175] Step d22: Divide all parameters in the task processing model into a first sensitivity parameter group, a second sensitivity parameter group and a third sensitivity parameter group according to the first-order sensitivity, second-order sensitivity and total effect sensitivity corresponding to all parameters in the task processing model respectively.

[0176] Specifically, this step can first divide the parameters into three groups based on the total effect sensitivity and the first-order sensitivity, and then adjust the three groups of parameters in combination with the second-order sensitivity and the epidemiological significance of the parameters to ensure that the parameters with high interactions or affecting the same driving factors are in the same group.

[0177] Step d3, for each task processing model corresponding to multiple vegetation indices, based on the diseased plant rate corresponding to the training samples and the labeled training samples, the cross-validation method and the sequential least squares planner are used to optimize the first sensitivity parameter group, the second sensitivity parameter group and the third sensitivity parameter group in the task processing model with the minimum root mean square error as the goal, and obtain the optimized model corresponding to the task processing model; so as to obtain the optimized models corresponding to the multiple vegetation indices.

[0178] Specifically, in order to optimize the model parameters, the cross-validation method is combined with the sequential least squares programming optimizer (SLSQP) to train the parameters within a range of variation. Here, SLSQP is one of the most robust methods for solving nonlinear bounded parameter optimization problems.

[0179] In this embodiment, the root mean square error (RSME) between the diseased plant rate DI output by the model and the true DI of the sample can be selected as the minimization objective function of the SLSQP method. First, the default values ​​of the parameters shown in Table 2 are used as the starting point of the SLSQP optimizer, and the three sets of parameters obtained by the Sobol sensitivity analysis method are adjusted in sequence. When training one set of parameters, the values ​​of the other parameters are set to the results obtained from the last training, and the process is repeated several times, such as Figure 5 ( Figure 5 In the example above, “high” represents the first sensitivity parameter group, “medium” represents the second sensitivity parameter group, and “low” represents the third sensitivity parameter group). To avoid overfitting, a cross-validation method is used. The samples are then divided into K folds (optional, K=3 in this embodiment), each subset is used as a test set, and the other K-1 subsets are used as training sets to execute the SLSQP optimizer. The RMSE of each test set is calculated using the model optimized by the corresponding training data, and the average RMSE of the K experiments is calculated. The average RMSE is recorded as RMSE CV The cross validation method is repeated N times (optionally, in this embodiment, N=10) to obtain N RMSECVs. In this embodiment, the training parameters corresponding to the smallest RMSE in the experiment corresponding to the smallest RMSECV are determined as the final model parameters, thereby obtaining the optimized model.

[0180] Step d4: Determine the disease mechanism model from the optimized models corresponding to the multiple vegetation indices.

[0181] See also Figure 6 As shown in the figure, 11 vegetation indices are used to calculate the vegetation index of Init.I The average RMSE obtained from the initialization training model shows that the result of OSAVI is the best. Therefore, optionally, in this embodiment, the optimized model corresponding to OSAVI can be determined as the disease mechanism model.

[0182] To evaluate the optimal model obtained from training, that is, using OSAVI to estimate InitI, the optimal parameters of the model obtained by the SLSQP and cross-validation methods are used to calculate the RMSE and the correlation coefficient R 2 It can be seen Figure 7 as shown, where (a) is the RMSE and R obtained under the training set 2 , and (b) is the RMSE and R obtained under the test set 2 .

[0183] In this embodiment, the samples can also be divided into different levels according to the diseased plant rate (for example, level 1 is DI ≤ 0.3, level 2 is 0.3 < DI ≤ 0.6, and level 3 is DI > 0.6), and the classification accuracy is obtained.

[0184] In addition, based on the training set and the test set, this embodiment uses the disease mechanism model to predict the diseased plant rate of wheat scab, and calculates the missed detection rate (MDR) and the accuracy rate (Acc) for the prediction results, that is, the proportion of samples that are actually at a high level but are misclassified as the lowest level. The results are shown in Table 3.

[0185] Table 3 Calculation results of evaluation indicators for the training set and the test set

[0186]

[0187] In addition, it can also be seen Figure 8 as shown, which is a schematic diagram of the number of target plants in four states obtained by continuous prediction using the disease mechanism model in this embodiment. Thus, it can be seen that this embodiment can analyze the epidemic process of the disease.

[0188] Based on the above results, the disease mechanism model provided in this embodiment can simulate the diseased plant rate with an error less than 0.2, and has high accuracy and precision, and can analyze the epidemic process of the disease.

[0189] This application embodiment also provides a device for predicting the diseased plant rate of plants. Next, the device for predicting the diseased plant rate of plants provided in this application embodiment will be described. The device for predicting the diseased plant rate of plants described below can be correspondingly referred to the plant diseased plant rate prediction method described above.

[0190] Please refer to Fig. 9 , which shows the structural schematic diagram of the device for predicting the diseased plant rate of plants provided in this application embodiment. As Fig. 9As shown, the plant disease rate prediction device may include: a first basic data acquisition module 901, an initial prediction time determination module 902, a second basic data acquisition module 903, a process data determination module 904 and a disease rate prediction module 905.

[0191] The first basic data acquisition module 901 is used to obtain remote sensing data of target plants and vegetation coverage at the jointing stage corresponding to multiple pixel areas, wherein the vegetation coverage at the jointing stage refers to the vegetation coverage under the remote sensing image when the target plant is at the jointing stage.

[0192] The initial prediction time determination module 902 is used to determine the greening period of the target plants corresponding to the multiple pixel areas respectively according to the remote sensing data of the target plants corresponding to the multiple pixel areas respectively, as the initial prediction time corresponding to the multiple pixel areas respectively.

[0193] The second basic data acquisition module 903 is used to obtain daily meteorological data and daily date related information corresponding to multiple pixel areas respectively, wherein the daily meteorological data and daily date related information refer to the daily meteorological data and daily date related information within the time range composed of the initial prediction time and the future time to be predicted.

[0194] The process data determination module 904 is used to determine the total planting quantity, initial infection quantity and daily disease prevalence rate of the target plants corresponding to the multiple pixel areas respectively according to the target plant remote sensing data corresponding to the multiple pixel areas respectively, the vegetation coverage during the jointing period, the initial prediction time, the daily meteorological data and the daily date related information, wherein the target plants are in any one of the healthy, latent, infected and removed infected states, the initial infection quantity refers to the number of target plants in the infected state at the time of initial prediction, the disease prevalence rate includes the infection rate, the first conversion rate and the second conversion rate, the infection rate refers to the probability that the target plant in the healthy state is infected by the target plant in the infected state and becomes the target plant in the latent state, the first conversion rate refers to the conversion rate when the target plant in the latent state is converted to the target plant in the infected state, and the second conversion rate refers to the conversion rate when the target plant in the infected state is converted to the target plant in the removed infected state.

[0195] The diseased plant rate prediction module 905 is used to determine the daily diseased plant rate of the target plants corresponding to the multiple pixel areas according to the total planting quantity, initial infection quantity, initial prediction time and daily disease prevalence rate of the target plants corresponding to the multiple pixel areas.

[0196] In summary, the working principle of the plant disease rate prediction device disclosed in this embodiment is the same as the working principle of the plant disease rate prediction method disclosed in the above embodiment, and will not be repeated here.

[0197] Finally, it should be noted that, in this article, relational terms such as and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprises" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the statement "comprises a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0198] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0199] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for predicting plant disease rate, characterized in that: include: Acquire target plant remote sensing data and jointing vegetation coverage corresponding to a plurality of pixel regions, wherein the jointing vegetation coverage refers to the vegetation coverage under the remote sensing image when the target plant is in the jointing stage; the target plant remote sensing data indicates continuous remote sensing data of the target plant in the growth period; the growth period includes wintering period, greening period, jointing period, heading period and flowering period; Determine, based on the target plant remote sensing data corresponding to the plurality of pixel regions, the greening periods of the target plants corresponding to the plurality of pixel regions, as the initial prediction times corresponding to the plurality of pixel regions; Acquire daily meteorological data and daily date related information corresponding to the plurality of pixel areas respectively, wherein the daily meteorological data and the daily date related information refer to the daily meteorological data and the daily date related information within a time range consisting of the initial prediction time and the future time to be predicted; According to the target plant remote sensing data corresponding to the multiple pixel areas, the vegetation coverage during the jointing period, the initial prediction time, the daily meteorological data and the daily date-related information, the total planting quantity, the initial infection quantity and the daily disease prevalence rate of the target plants corresponding to the multiple pixel areas are determined, wherein the target plants are in any one of the healthy, latent, infected and removed infected states, the initial infection quantity refers to the number of target plants in the infected state at the time of initial prediction, the disease prevalence rate includes the infection rate, the first conversion rate and the second conversion rate, the infection rate refers to the probability that the target plant in the healthy state is infected by the target plant in the infected state and becomes the target plant in the latent state, the first conversion rate refers to the conversion rate when the target plant in the latent state is converted to the target plant in the infected state, and the second conversion rate refers to the conversion rate when the target plant in the infected state is converted to the target plant in the removed infected state; Determine the daily diseased plant rate of the target plants corresponding to the multiple pixel areas respectively according to the total planting quantity, initial infection quantity, initial prediction time and daily disease prevalence rate of the target plants corresponding to the multiple pixel areas respectively; Wherein, determining the greening period of the target plants corresponding to the plurality of pixel regions respectively according to the remote sensing data of the target plants corresponding to the plurality of pixel regions respectively comprises: Determine a vegetation index time series curve of the target plant corresponding to each pixel area according to the target plant remote sensing data corresponding to the plurality of pixel areas; Smoothing the vegetation index time series curve of the target plant corresponding to each pixel area to obtain the smoothed vegetation index time series curve corresponding to each pixel area; The smoothed vegetation index time series curve corresponding to each pixel area is fitted by the logistic function to obtain the fitting curve corresponding to each pixel area; The second-order curvature of the fitting curve corresponding to each pixel area is calculated, and the maximum value point of each second-order curvature is determined. The time corresponding to each maximum value point is used as the greening period of the target plant corresponding to each pixel area.

2. The method for predicting plant disease rate according to claim 1, characterized in that: The daily meteorological data includes daily temperature data and daily humidity data; The method of determining the total planting quantity, initial infection quantity and daily disease prevalence rate of target plants corresponding to the plurality of pixel areas respectively according to the target plant remote sensing data corresponding to the plurality of pixel areas respectively, the vegetation coverage at the jointing stage, the initial prediction time, the daily meteorological data and the daily date related information comprises: Determine the total planting quantity of the target plants corresponding to the plurality of pixel areas respectively according to the vegetation coverage at the jointing stage corresponding to the plurality of pixel areas respectively; Determine the initial infection number of the target plants corresponding to the plurality of pixel areas respectively according to the remote sensing data of the target plants and the daily temperature data corresponding to the plurality of pixel areas respectively, in combination with the initial prediction time corresponding to the plurality of pixel areas respectively; Determine the flowering period of the target plants corresponding to the plurality of pixel regions respectively according to the remote sensing data of the target plants and the daily temperature data corresponding to the plurality of pixel regions respectively; According to the daily temperature data, daily humidity data and daily date-related information respectively corresponding to the multiple pixel areas, combined with the flowering periods respectively corresponding to the multiple pixel areas, the daily disease prevalence rates of the target plants respectively corresponding to the multiple pixel areas are determined.

3. The method for predicting plant disease rate according to claim 2, characterized in that: The determining the total planting quantity of target plants corresponding to the plurality of pixel areas respectively according to the vegetation coverage at the jointing stage corresponding to the plurality of pixel areas respectively comprises: Determining a linear regression coefficient according to the vegetation coverage at the jointing stage corresponding to each of the plurality of pixel regions; The total planting quantity of the target plants respectively corresponding to the plurality of pixel areas is determined according to the linear regression coefficient and the vegetation coverage at the jointing stage respectively corresponding to the plurality of pixel areas.

4. The method for predicting plant disease rate according to claim 3, characterized in that: The method of determining the initial infection number of the target plants corresponding to the plurality of pixel regions respectively according to the remote sensing data of the target plants and the daily temperature data respectively corresponding to the plurality of pixel regions and the initial prediction time respectively corresponding to the plurality of pixel regions comprises: For each pixel region in the plurality of pixel regions: Determine the vegetation index of the green period corresponding to the pixel area according to the remote sensing data of the target plant corresponding to the pixel area and the initial prediction time; According to the initial prediction time corresponding to the pixel area, the vegetation coverage in the green period corresponding to the pixel area is obtained, wherein the vegetation coverage in the green period refers to the vegetation coverage under the remote sensing image when the target plant is in the green period; According to the daily temperature data corresponding to the pixel area, the average temperature data within the range from the wintering period to the greening period of the target plant is calculated as the average temperature data corresponding to the pixel area; To obtain the average temperature data, green vegetation index and green vegetation coverage corresponding to the multiple pixel areas respectively; The initial infection numbers of the target plants corresponding to the multiple pixel areas are determined according to the average temperature data, the green vegetation index and the green vegetation coverage corresponding to the multiple pixel areas.

5. The method for predicting plant disease rate according to claim 4, characterized in that: The determining of the daily disease prevalence rate of the target plants corresponding to the plurality of pixel areas respectively according to the daily temperature data, daily humidity data and daily date related information respectively corresponding to the plurality of pixel areas in combination with the flowering periods respectively corresponding to the plurality of pixel areas comprises: According to the daily humidity data respectively corresponding to the plurality of pixel regions and the pre-established response curves of the infection rate, the first conversion rate and the second conversion rate to humidity, respectively, determining the response values ​​of the daily infection rate, the first conversion rate and the second conversion rate respectively corresponding to the plurality of pixel regions to humidity as the daily first response value, the second response value and the third response value, respectively; According to the daily temperature data corresponding to the plurality of pixel regions, and the pre-established response curves of the infection rate, the first conversion rate, and the second conversion rate to the temperature, respectively, determining the response values ​​of the daily infection rate, the first conversion rate, and the second conversion rate corresponding to the plurality of pixel regions to the temperature, respectively as the fourth response value, the fifth response value, and the sixth response value of the daily response value; According to the daily date-related information and flowering period corresponding to the multiple pixel regions, and the pre-established response curves of the infection rate, the first conversion rate, and the second conversion rate to the growth period, respectively, determining the response values ​​of the daily infection rate, the first conversion rate, and the second conversion rate corresponding to the multiple pixel regions to the growth period, respectively, as the seventh response value, the eighth response value, and the ninth response value of the daily response value; Determining daily infection rates of target plants corresponding to the plurality of pixel regions respectively according to the daily first response value, the fourth response value, and the seventh response value; Determine daily first conversion rates of target plants corresponding to the plurality of pixel regions respectively according to the daily second response value, the fifth response value and the eighth response value; According to the third response value, the sixth response value and the ninth response value of each day, the second conversion rates of the target plants respectively corresponding to the plurality of pixel areas are determined.

6. The method for predicting plant disease rate according to claim 5, characterized in that: Determining the daily diseased plant rate of the target plants corresponding to the plurality of pixel areas respectively according to the total planting quantity, the initial infection quantity, the initial prediction time and the daily disease prevalence rate of the target plants respectively corresponding to the plurality of pixel areas comprises: For each pixel region in the plurality of pixel regions: Determine the number of target plants in the healthy, latent, infectious and removed infectious states in the pixel area on a daily basis, according to the total planting number, initial infection number, initial prediction time and daily disease prevalence rate of the target plants corresponding to the pixel area; Determine the daily disease rate of the target plants corresponding to the pixel area according to the number of target plants in the healthy, latent, infectious and removed infectious states in the pixel area on a daily basis; In order to obtain the daily diseased plant rates of the target plants corresponding to the multiple pixel areas respectively.

7. The method for predicting the plant disease rate according to any one of claims 4 to 6, characterized in that: The method comprises: determining the initial infection number of target plants corresponding to the plurality of pixel areas respectively according to the average temperature data, green vegetation index and green vegetation coverage respectively corresponding to the plurality of pixel areas; determining the daily disease prevalence rate of target plants corresponding to the plurality of pixel areas respectively according to the daily temperature data, daily humidity data and daily date related information respectively corresponding to the plurality of pixel areas, in combination with the flowering period respectively corresponding to the plurality of pixel areas; and determining the daily disease plant rate of target plants corresponding to the plurality of pixel areas respectively according to the total planting number, initial infection number, initial prediction time and daily disease prevalence rate of target plants corresponding to the plurality of pixel areas respectively, including: Using the disease mechanism model obtained in advance, the daily temperature data, daily humidity data, daily date-related information, average temperature data, green vegetation index, green vegetation coverage, flowering period, total planting quantity of target plants and initial prediction time corresponding to the multiple pixel areas are processed to obtain the daily diseased plant rate of the target plants respectively contained in the multiple pixel areas output by the disease mechanism model; Among them, the disease mechanism model uses the training daily temperature data, training daily humidity data, training daily date-related information, training average temperature data, training green vegetation index, training green vegetation coverage, training flowering period, total planting number of training target plants and training initial prediction time corresponding to the multiple pixel areas as training samples, and is trained using the diseased plant rate corresponding to the labeled training samples as training labels.

8. The method for predicting plant disease rate according to claim 7, characterized in that: The process of obtaining the disease mechanism model includes: Obtaining task processing models corresponding to multiple pre-built vegetation indices; For each of the task processing models corresponding to the multiple vegetation indices, a global sensitivity analysis method is used to analyze the sensitivity of the parameters in the task processing model, so as to divide the parameters in the task processing model into a first sensitivity parameter group, a second sensitivity parameter group with lower sensitivity than the first sensitivity parameter group, and a third sensitivity parameter group with lower sensitivity than the second sensitivity parameter group, wherein the parameters in the task processing model include parameters in the response curves of the infection rate, the first conversion rate, and the second conversion rate to humidity, respectively, and parameters in the response curves of the infection rate, the first conversion rate, and the second conversion rate to temperature, respectively, and parameters in the response curves of the infection rate, the first conversion rate, and the second conversion rate to the growth period, respectively, and parameters in the initial infection number calculation formula; For each of the task processing models corresponding to the multiple vegetation indices, based on the training samples and the diseased plant rates corresponding to the labeled training samples, a cross-validation method and a sequential least squares planner are used to optimize the first sensitivity parameter group, the second sensitivity parameter group and the third sensitivity parameter group in the task processing model with the minimum root mean square error as the goal, so as to obtain the optimized model corresponding to the task processing model; so as to obtain the optimized models corresponding to the multiple vegetation indices; The disease mechanism model is determined from the optimized models corresponding to the multiple vegetation indices.

9. The method for predicting plant disease rate according to claim 8, characterized in that: The method of using a global sensitivity analysis method to analyze the sensitivity of the parameters in the task processing model to divide the parameters in the task processing model into a first sensitivity parameter group, a second sensitivity parameter group having lower sensitivity than the first sensitivity parameter group, and a third sensitivity parameter group having lower sensitivity than the second sensitivity parameter group includes: For each parameter in the task processing model, the training sample is input into the task processing model to obtain the diseased plant rate output by the task processing model, and the variance of the diseased plant rate output by the task processing model is calculated, and according to the parameter and the variance, the first-order sensitivity, second-order sensitivity and total effect sensitivity corresponding to the parameter are determined to obtain the first-order sensitivity, second-order sensitivity and total effect sensitivity corresponding to all parameters in the task processing model respectively; According to the first-order sensitivity, second-order sensitivity and total effect sensitivity respectively corresponding to all the parameters in the task processing model, all the parameters in the task processing model are divided into the first sensitivity parameter group, the second sensitivity parameter group and the third sensitivity parameter group.

10. A plant disease rate prediction device, characterized in that: include: The first basic data acquisition module is used to obtain remote sensing data of target plants and vegetation coverage at the jointing stage corresponding to multiple pixel areas, wherein the vegetation coverage at the jointing stage refers to the vegetation coverage under the remote sensing image when the target plant is at the jointing stage; the remote sensing data of the target plant indicates continuous remote sensing data of the target plant within the growth period; the growth period includes the overwintering period, the greening period, the jointing period, the heading period and the flowering period; An initial prediction time determination module is used to determine the greening period of the target plants corresponding to the plurality of pixel areas respectively according to the remote sensing data of the target plants corresponding to the plurality of pixel areas respectively, as the initial prediction time corresponding to the plurality of pixel areas respectively; A second basic data acquisition module is used to acquire daily meteorological data and daily date related information corresponding to the multiple pixel areas respectively, wherein the daily meteorological data and the daily date related information refer to the daily meteorological data and the daily date related information within a time range consisting of the initial prediction time and the future time to be predicted; a process data determination module, for determining the total planting quantity, initial infection quantity and daily disease prevalence rate of the target plants corresponding to the plurality of pixel areas respectively according to the target plant remote sensing data, the vegetation coverage degree at the jointing stage, the initial prediction time, the daily meteorological data and the daily date related information respectively corresponding to the plurality of pixel areas, wherein the target plants are in any one of the healthy, latent, infected and removed infected states, the initial infection quantity refers to the quantity of the target plants in the infected state at the time of initial prediction, the disease prevalence rate includes an infection rate, a first conversion rate and a second conversion rate, the infection rate refers to the probability that the target plant in the healthy state is infected by the target plant in the infected state and becomes the target plant in the latent state, the first conversion rate refers to the conversion rate when the target plant in the latent state is converted to the target plant in the infected state, and the second conversion rate refers to the conversion rate when the target plant in the infected state is converted to the target plant in the removed infected state; A diseased plant rate prediction module, used to determine the daily diseased plant rate of the target plants corresponding to the multiple pixel areas respectively according to the total planting quantity, initial infection quantity, initial prediction time and daily disease prevalence rate of the target plants corresponding to the multiple pixel areas respectively; The initial prediction time determination module is used to: Determine a vegetation index time series curve of the target plant corresponding to each pixel area according to the target plant remote sensing data corresponding to the plurality of pixel areas; Smoothing the vegetation index time series curve of the target plant corresponding to each pixel area to obtain the smoothed vegetation index time series curve corresponding to each pixel area; The smoothed vegetation index time series curve corresponding to each pixel area is fitted by the logistic function to obtain the fitting curve corresponding to each pixel area; The second-order curvature of the fitting curve corresponding to each pixel area is calculated, and the maximum value point of each second-order curvature is determined. The time corresponding to each maximum value point is used as the greening period of the target plant corresponding to each pixel area.

Images