Intelligent Early Warning System and Method for Crop Diseases

Abstract

The present invention belongs to the technical field of crop disease early warning, and particularly relates to an intelligent crop disease early warning system and method, including: obtaining crop information in the current farmland and establishing a farmland visualization model; detecting the disease situation of the current farmland through farmland monitoring equipment to generate a disease information feature set; collecting relevant meteorological data of the farmland, processing the disease information feature set, establishing a disease spread analysis model, analyzing disease-related factors, and establishing a farmland disease impact factor database; analyzing the development of diseases in the current farmland according to the farmland disease impact factor database, evaluating the disease risks of grids without diseases in the farmland, and feeding back the discrimination results to the farmland visualization model to warn relevant personnel. The accuracy of predicting the disease spread trend and the location where the disease appears under complex and changeable meteorological conditions is improved, the crop losses caused by the lag of the prevention and control time are avoided, and the disease prevention and control effect is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of crop disease early warning, and particularly relates to an intelligent crop disease early warning system and method. Background Art

[0002] With the development of modern agriculture, large-scale centralized planting has become more and more popular. Through reasonable land planning, large farmlands implement efficient mechanized operations and apply Internet of Things technology to achieve real-time monitoring of farmland status. However, centralized planting increases the risk of crop disease transmission.

[0003] When warning against crop diseases in the prior art, historical disease data is usually analyzed based on the crop growth cycle and environmental factors to predict the development trend of diseases. However, traditional methods often rely on empirical judgment and simple statistical models, and these methods may not be able to accurately predict the occurrence of diseases in complex environments, resulting in untimely warning information and delaying the prevention and control timing. At the same time, traditional methods may be difficult to integrate multi-source data, unable to make full use of the advantages of big data to achieve real-time dynamic monitoring, and difficult to track the development trend of diseases. Summary of the Invention

[0004] The purpose of the present invention is to provide an intelligent crop disease early warning system and method to solve the problems raised in the above background art.

[0005] To achieve the purpose of the present invention, the following technical solutions will be adopted for implementation.

[0006] An intelligent crop disease early warning method includes the following steps:

[0007] Step S100: Obtain crop information in the current farmland and establish a farmland visualization model, where the farmland visualization model includes farmland three-dimensional terrain data, crop information, and disease warning information;

[0008] Step S200: Detect the disease situation of the current farmland through farmland monitoring equipment to generate a disease information feature set;

[0009] Step S300: Collect relevant meteorological data of the farmland, process the disease information feature set, establish a disease transmission analysis model, analyze disease-related factors based on the disease transmission analysis model, and establish a farmland disease impact factor database;

[0010] Step S400: Analyze the development of diseases in the current farmland according to the farmland disease impact factor database, evaluate the disease risk of grids without diseases in the farmland, and feedback the discrimination result to the farmland visualization model to warn relevant personnel;

[0011] Accurately predict the spread trend and occurrence location of diseases under complex and changeable meteorological conditions, avoid crop losses caused by the lag of prevention and control time, help farmers effectively prevent and control crop diseases, improve the effect of disease prevention and control, and reduce planting costs.

[0012] As a preferred embodiment of the present invention, the implementation process of step S100 includes the following steps:

[0013] Step S101: Divide the current farmland into several grids and number them according to the farmland images collected by the drone, obtain the crop information in the current farmland through the sensors set in the farmland, and generate a crop information set, which is denoted as , where represents the crop feature set of the j-th grid, J represents the total number of grids in the current farmland, , n represents the total number of crops in the grid; represents the position data of the i-th crop in the j-th grid, represents the growth feature data of the i-th crop in the j-th grid, and the growth feature data includes plant height, stem thickness, flower quantity and fruit quantity, represents the health feature data of the i-th crop in the j-th grid, and the health feature data includes the disease type and the disease occurrence time of the crop;

[0014] Step S102: Establish a farmland visualization model based on virtual reality technology, and display the crop growth conditions of each grid in the farmland according to the crop information set.

[0015] As a preferred embodiment of the present invention, in step S200, the diseases in the current farmland are detected by farmland monitoring equipment, and the corresponding positions in the farmland visualization model are marked according to the disease occurrence locations; access the database of the farmland visualization model to generate a disease information feature set, which is denoted as , where represents the disease information feature of the j-th grid, , represents the disease information feature of the i-th crop in the j-th grid; the disease information feature includes disease type, disease occurrence time and crop performance characteristics.

[0016] As a preferred embodiment of the present invention, the implementation process of step S300 includes the following steps:

[0017] Step S301: Analyze and extract the central crops of the k-th type of disease in the farmland according to the disease information feature set:

[0018] ;

[0019] Where Represents the position of the central crop of the k-th type of disease, and the central crop of the k-th type of disease is the source of transmission of the k-th type of disease; Represents the total number of crops with the disease information characteristic of the k-th type of disease;

[0020] Step S302: Analyze the central crops corresponding to the diseased crops according to the following formula:

[0021] ;

[0022] When the above equation holds, it is determined that the position of the central crop corresponding to the i-th crop in the j-th grid is ;

[0023] Among them, ; Represents the similarity between the i-th crop in the j-th grid and the central crop of the k-th type of disease, B represents the total number of disease categories, Represents the disease information characteristic of the central crop of the k-th type of disease;

[0024] Step S303: Collect the relevant meteorological data of the farmland through the sensor network deployed in the farmland, and record the relevant meteorological data of the farmland according to the time sequence; generate a sample set for each central crop of various diseases and the diseased crops corresponding to the central crop respectively, process the disease information feature set based on the recorded relevant meteorological data of the farmland, calculate the correlation between the relevant meteorological data of the farmland and the occurrence of diseases using the Pearson correlation coefficient, and the relevant meteorological data of the farmland includes temperature, wind direction and precipitation; generate a disease transmission analysis model, and the disease transmission analysis model generates a correlation coefficient matrix between features according to the processing of the corresponding disease information feature sets of each sample set, and the element in the a-th row and m-th column of the matrix is denoted as L am , L am Represents the correlation coefficient between the a-th sample set and the m-th element in the relevant meteorological data of the farmland; linearly expand the matrix to establish a farmland disease impact factor database, and the farmland disease impact factor database includes the correlation between the disease types and each element in the relevant meteorological data of the farmland.

[0025] According to the above method, the diseased crops in the farmland are integrated, the disease center is traced, the source of transmission of various diseases is determined, and then the correlation between the diseased crops and the meteorological data is analyzed. The transmission impact is reflected through the correlation coefficient matrix between features, and the transmission impact includes positive impact and negative impact. The positive impact means that the meteorological element helps to accelerate the spread of the disease, and the negative impact means that the meteorological element helps to control the spread of the disease.

[0026] As a preferred solution of the present invention, the implementation process of step S400 includes the following steps:

[0027] Step S401: Obtain the meteorological data related to the farmland in the next unit time through a third-party API; calculate the disease risk index of the grids without diseases in the current farmland according to the following formula:

[0028] ;

[0029] where BH j represents the disease risk index of the j-th grid without diseases in the current farmland; represents the shortest distance between the j-th grid without diseases and the crops with the k-th type of disease in the farmland, represents the correlation coefficient between the k-th type of disease and the m-th element in the meteorological data related to the farmland, which is obtained from the farmland disease impact factor database; , M represents the total number of elements in the meteorological data related to the farmland;

[0030] Step S402: Calculate the disease risk index of each grid in the farmland, preset a disease risk index threshold. If the disease risk index of the j-th grid without diseases in the current farmland is greater than the disease risk index threshold, it is determined that there is a disease risk in the j-th grid of the farmland, and the discrimination result is fed back to the farmland visualization model to give a warning to relevant personnel.

[0031] The intelligent early warning system for crop diseases includes: a visualization model construction module, a disease detection module, an impact factor analysis module, and a disease early warning module;

[0032] The visualization model construction module is used to obtain the crop information in the current farmland and establish a farmland visualization model;

[0033] The disease detection module is used to detect the diseases in the current farmland through farmland monitoring equipment and mark the corresponding positions in the farmland visualization model according to the positions where the diseases appear;

[0034] The impact factor analysis module is used to collect the meteorological data related to the farmland, process the disease information feature set, establish a disease transmission analysis model, analyze the disease-related factors based on the disease transmission analysis model, and establish a farmland disease impact factor database;

[0035] The disease early warning module is used to predict the development of diseases in the current farmland according to the farmland disease impact factor database, evaluate the disease risks of each area in the farmland, and feed back the evaluation results to the farmland visualization model to give a warning to relevant personnel.

[0036] Further, the visualization model construction module includes a farmland data acquisition unit, a three-dimensional model establishment unit, and a farmland data update unit;

[0037] The farmland data acquisition unit is used to divide the current farmland into several grids and number them according to the farmland images collected by the drone, and obtain the crop information in the current farmland through the sensors set in the farmland to generate a crop information set;

[0038] The 3D model establishment unit establishes a farmland visualization model based on virtual reality technology and displays the crop growth conditions of each grid in the farmland according to the crop information set;

[0039] The farmland data update unit is used to update the disease warning information in the farmland visualization model according to the detection results of the disease detection module.

[0040] Furthermore, the impact factor analysis module includes a disease characteristic analysis unit, a disease transmission analysis unit, and a farmland disease impact factor database establishment unit;

[0041] The disease characteristic analysis unit is used to analyze the central crops of various diseases according to the disease information characteristic set;

[0042] The disease transmission analysis unit is used to analyze the disease transmission law according to the central crops of various diseases;

[0043] The farmland disease impact factor database establishment unit is used to establish a farmland disease impact factor database according to the output of the disease transmission analysis unit. Beneficial effects

[0044] (1) By establishing a farmland visualization model and updating the crop status in real time, the present invention makes it easier for managers to understand and analyze the farmland conditions;

[0045] (2) By collecting and analyzing the relevant meteorological data of the farmland, establishing a disease transmission analysis model, establishing a farmland disease impact factor database, and accumulating disease prevention and control knowledge, the present invention provides support for long-term management; accurately predicting the disease transmission trend and the location where the disease appears under complex and changeable meteorological conditions, avoiding crop losses caused by the lag of prevention and control time, helping farmers effectively prevent and control crop diseases, taking preventive measures in advance, reducing the occurrence and spread of crop diseases, improving the disease prevention and control effect, and reducing the planting cost. Description of the drawings

[0046] Figure 1 is a schematic flowchart of the intelligent early warning system for crop diseases of the present invention. Detailed implementation manners

[0047] Next, the technical solutions of the present invention will be clearly and completely described in conjunction with the embodiments of the present invention and the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0048] As an embodiment of the present invention, as Figure 1 shown, the present invention provides an intelligent early warning method for crop diseases, including the following steps:

[0049] Step S100: Obtain the crop information in the current farmland and establish a farmland visualization model, where the farmland visualization model includes farmland three-dimensional terrain data, crop information, and disease warning information;

[0050] Step S101: Divide the current farmland into several grids and number them according to the farmland images collected by the unmanned aerial vehicle. Obtain the crop information in the current farmland through the sensors set in the farmland, and generate a crop information set, denoted as where, represents the crop feature set of the j-th grid, J represents the total number of grids in the current farmland, and n represents the total number of crops in the grid; represents the position data of the i-th crop in the j-th grid, represents the growth feature data of the i-th crop in the j-th grid, and the growth feature data includes plant height, stem thickness, flower quantity, and fruit quantity, represents the health feature data of the i-th crop in the j-th grid, and the health feature data includes the disease type and the disease occurrence time of the crop;

[0051] Step S102: Establish a farmland visualization model based on virtual reality technology, and display the crop growth conditions of each grid in the farmland according to the crop information set.

[0052] Step S200: Detect the diseases in the current farmland through the farmland monitoring equipment, and mark the corresponding positions in the farmland visualization model according to the disease occurrence positions; access the database of the farmland visualization model, and generate a disease information feature set, denoted as where, represents the disease information feature of the j-th grid, and represents the disease information feature of the i-th crop in the j-th grid; the disease information feature includes the disease type, the disease occurrence time, and the crop performance characteristics.

[0053] Step S300: Collect meteorological data related to farmland, process the disease information feature set, establish a disease transmission analysis model, analyze disease-related factors based on the disease transmission analysis model, and establish a database of farmland disease impact factors;

[0054] Step S301: Analyze and extract the central crop of the k-th type of disease in the farmland according to the disease information feature set:

[0055] ;

[0056] where, represents the position of the central crop of the k-th type of disease, and the central crop of the k-th type of disease is the source of transmission of the k-th type of disease; represents the total number of crops with the k-th type of disease information feature;

[0057] Step S302: Analyze the central crop corresponding to the diseased crop according to the following formula:

[0058] ;

[0059] When the above equation holds, it is determined that the position of the central crop corresponding to the i-th plant in the j-th grid is ;

[0060] where, ; represents the similarity between the i-th plant in the j-th grid and the central crop of the k-th type of disease, B represents the total number of disease categories, represents the disease information feature of the central crop of the k-th type of disease;

[0061] Step S303: Collect meteorological data related to farmland through the sensor network deployed in the farmland, record the meteorological data related to the farmland according to the time sequence; generate a sample set for each central crop of various diseases and the diseased crops corresponding to the central crop respectively, process the disease information feature set based on the recorded meteorological data related to the farmland, calculate the correlation between the meteorological data related to the farmland and the occurrence of diseases using the Pearson correlation coefficient, and the meteorological data related to the farmland includes temperature, wind direction and precipitation; generate a disease transmission analysis model, and the disease transmission analysis model generates a correlation coefficient matrix between features according to the processing of the disease information feature set corresponding to each sample set, which can be displayed using a heat map, and the element in the a-th row and m-th column of the matrix is denoted as L am , L amrepresents the correlation coefficient between the a-th sample set and the m-th element in the meteorological data related to farmland; linearly expand the matrix, and insert new values into the matrix through interpolation to smoothly expand the size of the matrix, so that the farmland disease impact factor database can include daily meteorological values; establish a farmland disease impact factor database, which includes the correlation between disease types and each element in the meteorological data related to farmland.

[0062] According to the above method, integrate the diseased crops in the farmland, trace the disease center, determine the transmission sources of various diseases, and then analyze the correlation between the diseased crops and the meteorological data. The transmission impact is reflected by the correlation coefficient matrix between features, indicating the transmission impact of different meteorologies on different disease types; the transmission impact includes positive impact and negative impact. The positive impact means that meteorological elements help to accelerate the spread of diseases, and the negative impact means that meteorological elements help to control the spread of diseases.

[0063] Step S400: Analyze the development of diseases in the current farmland according to the farmland disease impact factor database, evaluate the disease risks of the grids without diseases in the farmland, and feedback the discrimination results to the farmland visualization model to alert relevant personnel;

[0064] Step S401: Obtain the meteorological data related to the farmland in the next unit time through a third-party API; calculate the disease risk index of the grids without diseases in the current farmland according to the following formula:

[0065] ;

[0066] where BH j represents the disease risk index of the j-th grid without diseases in the current farmland; represents the shortest distance between the j-th grid without diseases and the crops with the k-th type of disease in the farmland, represents the correlation coefficient between the k-th type of disease and the m-th element in the meteorological data related to the farmland, obtained from the farmland disease impact factor database; , M represents the total number of elements in the meteorological data related to the farmland;

[0067] Step S402: Calculate the disease risk index of each grid in the farmland, preset the disease risk index threshold. If the disease risk index of the j-th grid without diseases in the current farmland is greater than the disease risk index threshold, it is determined that there is a disease risk in the j-th grid of the farmland, and the discrimination result is feedback to the farmland visualization model to alert relevant personnel.

[0068] The technical solutions of the present invention have been described in detail above in conjunction with the embodiments / figures. However, the present invention is not limited to the above technical solutions. For those of ordinary skill in the art, after learning the content described in the present invention, without departing from the principle of the present invention, several equivalent transformations and substitutions can still be made, and these equivalent transformations and substitutions should also be regarded as belonging to the protection scope of the present invention.

Claims

1. An intelligent early warning method for crop diseases, characterized in that: The following steps are involved: Step S100: obtaining crop information in the current farmland and establishing a farmland visualization model, wherein the farmland visualization model includes three-dimensional farmland terrain data, crop information and disease warning information; Step S200: Detect the disease condition of the current farmland through farmland monitoring equipment to generate a disease information feature set; Step S300: Collect farmland-related meteorological data, process the disease information feature set, establish a disease propagation analysis model, analyze disease-related factors based on the disease propagation analysis model, and establish a database of farmland disease influencing factors; Step S400: Analyze the disease development of the current farmland according to the farmland disease influencing factor database, evaluate the disease risk of the grids in the farmland where no disease occurs, and feed back the judgment results to the farmland visualization model to warn relevant personnel; The implementation process of step S300 includes the following steps: Step S301: Extract the center crop of the kth type of disease in the farmland according to the disease information feature set analysis: Among them, X k represents the location of the center crop of the kth disease, |C k | indicates the total number of crops with the disease information characteristic of the kth type of disease; Represents the location data of the i-th crop in the j-th grid; Step S302: Analyze the central crop corresponding to the crop with disease according to the following formula: When the above equation holds true, the position of the central crop corresponding to the i-th crop in the j-th grid is determined to be X k ; in, represents the similarity between the i-th crop in the j-th grid and the central crop of the k-th disease, B represents the total number of disease categories, R(X k ) represents the disease information characteristics of the central crop of the kth disease; Represents the disease information characteristics of the i-th crop in the j-th grid; Step S303: Collect farmland-related meteorological data through the sensor network deployed in the farmland, and record the farmland-related meteorological data in chronological order; generate a sample set for each central crop of various diseases and the diseased crop corresponding to the central crop, process the disease information feature set based on the recorded farmland-related meteorological data, use the Pearson correlation coefficient to calculate the correlation between the farmland-related meteorological data and the occurrence of the disease, and generate a disease propagation analysis model. The disease propagation analysis model generates a correlation coefficient matrix between features based on the processing of the disease information feature set corresponding to each sample set, and the element in the a-th row and m-th column of the matrix is ​​denoted as L am , L am Represents the correlation coefficient between the a-th sample set and the m-th element in the farmland-related meteorological data; linearly expands the matrix to establish a database of farmland disease influencing factors, wherein the database of farmland disease influencing factors includes the correlation between the disease type and each element in the farmland-related meteorological data.

2. The intelligent early warning method for crop diseases according to claim 1, characterized in that: The implementation process of step S100 includes the following steps: Step S101: divide the current farmland into several grids according to the farmland image collected by the drone and number them, obtain the crop information in the current farmland through the sensor set in the farmland, and generate a crop information set, which is recorded as ZW={Z j |j∈[1,J]}, where Z j represents the crop feature set of the jth grid, J represents the total number of grids in the current farmland, i∈[1,n], n represents the total number of crops in the grid; represents the growth characteristic data of the i-th crop in the j-th grid, Represents the health characteristic data of the i-th crop in the j-th grid; Step S102: Establish a farmland visualization model based on virtual reality technology, and display the crop growth conditions of each grid of the farmland according to the crop information set.

3. The intelligent early warning method for crop diseases according to claim 1, characterized in that: In step S200, the current farmland disease is detected by farmland monitoring equipment, and the corresponding position in the farmland visualization model is marked according to the location where the disease occurs; the database of the farmland visualization model is accessed to generate a disease information feature set, which is recorded as H = {H (R j )|j∈[1,J]}, where H(R j ) represents the disease information characteristics of the jth grid, 4. The intelligent early warning method for crop diseases according to claim 1, characterized in that: The implementation process of step S400 includes the following steps: Step S401: Obtain farmland-related meteorological data within the next unit of time through a third-party API; The disease risk index of the grid without disease in the current farmland is calculated according to the following formula: Among them, BH j represents the disease risk index of the jth disease-free grid in the current farmland; d jk represents the shortest distance between the jth disease-free grid and the crop with the kth disease in the farmland, L km represents the correlation coefficient between the k-th disease and the m-th element in the farmland-related meteorological data, which is obtained from the farmland disease influencing factor database; m∈[1,M], M represents the total number of elements in the farmland-related meteorological data; Step S402: Calculate the disease risk index of each grid of the farmland, preset the disease risk index threshold, if the disease risk index of the jth grid without disease in the current farmland is greater than the disease risk index threshold, then it is determined that the jth grid of the farmland has a disease risk, and the judgment result is fed back to the farmland visualization model to warn relevant personnel.

5. An intelligent early warning system for crop diseases, executing the intelligent early warning method for crop diseases according to any one of claims 1 to 4, characterized in that: include: Visual model building module, disease detection module, influencing factor analysis module and disease early warning module; The visualization model building module is used to obtain crop information in the current farmland and build a farmland visualization model; The disease detection module is used to detect diseases in the current farmland through farmland monitoring equipment, and mark the corresponding positions in the farmland visualization model according to the locations where the diseases appear; The influencing factor analysis module is used to collect farmland-related meteorological data, process the disease information feature set, establish a disease propagation analysis model, analyze disease-related factors based on the disease propagation analysis model, and establish a farmland disease influencing factor database; The disease warning module is used to predict the current farmland disease development based on the farmland disease influencing factor database, evaluate the disease risk of each area of ​​the farmland, and feed the evaluation results back to the farmland visualization model to warn relevant personnel.

6. The intelligent early warning system for crop diseases according to claim 5 is characterized in that: The visualization model building module includes a farmland data acquisition unit, a three-dimensional model building unit and a farmland data updating unit; The farmland data acquisition unit is used to divide the current farmland into a number of grids and number them according to the farmland images collected by the drone, obtain the crop information in the current farmland through the sensors set in the farmland, and generate a crop information set; The three-dimensional model building unit builds a farmland visualization model based on virtual reality technology, and displays the crop growth conditions of each grid of the farmland according to the crop information set; The farmland data updating unit is used to update the disease warning information in the farmland visualization model according to the detection results of the disease detection module.

7. The intelligent early warning system for crop diseases according to claim 5 is characterized in that: The influencing factor analysis module includes a disease characteristic analysis unit, a disease propagation analysis unit and a farmland disease influencing factor database establishment unit; The disease feature analysis unit is used to analyze the central crops of various diseases according to the disease information feature set; The disease spread analysis unit is used to analyze the disease spread rules according to the central crops of various diseases; The farmland disease influencing factor database establishing unit is used to establish a farmland disease influencing factor database according to the output of the disease propagation analysis unit.

Images