Plant disease treatment prompting method and device, electronic equipment and storage medium
By acquiring and analyzing questionnaire and image data of potted plants, and using a disease identification model to identify and confirm disease types, treatment plans were developed and implemented, solving the problem of inaccurate identification of potted plant diseases and enabling timely treatment.
Patent Information
- Application Number
- CN202510896839.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-06-30
- Publication Date
- 2025-11-18
AI Technical Summary
In existing methods, when people grow potted plants themselves, they cannot accurately identify the cause of diseases, resulting in untimely disease treatment.
By acquiring questionnaire data, overall and partial image data of the target plants, a disease identification model is used to identify candidate disease types and their confidence levels, and questionnaire data is generated to confirm the target disease type, and a treatment plan is formulated and pushed out.
It enables accurate identification and timely treatment of plant diseases, improving the efficiency of disease management for potted plants.
Smart Images
Figure CN120976689A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of data processing, and in particular to a plant disease treatment prompting method and device, electronic equipment and storage medium. BACKGROUND
[0002] The growth environment is a basic factor affecting the quality of plant growth, such as temperature, light, moisture, and terrain soil. During the growth of plants, diseases will occur, which will affect plant growth. The existing method for plant disease treatment process is found by artificial discovery and experience treatment. For personnel self-planting potted plants, the cause of the disease cannot be accurately identified, and therefore, a plant disease treatment prompting method suitable for household potted plants is urgently needed. To solve the problem that the existing method uses artificial discovery and experience treatment for plant disease identification, and for personnel self-planting potted plants, the cause of the disease cannot be accurately identified, resulting in that the plant disease cannot be treated in time. SUMMARY
[0003] The present application provides a plant disease treatment prompting method, which aims to solve the problem that the existing method uses artificial discovery and experience treatment for plant disease identification, and for personnel self-planting potted plants, the cause of the disease cannot be accurately identified, resulting in that the plant disease cannot be treated in time. The present application determines the candidate disease type of the target plant and the confidence of the candidate disease type according to the first questionnaire data, the first image data and the second image data of the target plant. When there are multiple candidate disease types with a confidence greater than a confidence threshold, the second questionnaire data is generated according to the multiple candidate disease types, and the target disease type is determined among the multiple candidate disease types by using the second questionnaire data. The disease treatment scheme of the target plant is determined by the target disease type, and the disease treatment scheme is pushed. The present application can accurately identify the cause of the plant disease, so that the plant disease can be treated in time. The problem that the existing method uses artificial discovery and experience treatment for plant disease identification, and for personnel self-planting potted plants, the cause of the disease cannot be accurately identified, resulting in that the plant disease cannot be treated in time is solved.
[0004] In a first aspect, the present application provides a plant disease treatment prompting method, which comprises the following steps:
[0005] Obtain the first questionnaire data, the first image data and the second image data of the target plant, the first image data is the overall image data of the target plant, and the second image data is the local image data of the target plant;
[0006] determine a candidate disease type of the target plant and a confidence of the candidate disease type based on the first questionnaire data, the first image data and the second image data;
[0007] if there are multiple candidate disease types with a confidence greater than a confidence threshold, generate second questionnaire data based on the multiple candidate disease types;
[0008] determine a target disease type from the multiple candidate disease types based on the second questionnaire data;
[0009] determine a disease treatment scheme for the target plant based on the target disease type, and push the disease treatment scheme.
[0010] Optionally, the first questionnaire data of the target plant is obtained by:
[0011] obtaining current growth environment data and historical growth environment data of the target plant;
[0012] generating the first questionnaire data of the target plant based on the current growth environment and the historical growth environment.
[0013] Optionally, the current growth environment data of the target plant is obtained by:
[0014] obtaining a location of the target plant and current season data;
[0015] determining the current growth environment data of the target plant based on the location of the target plant and the current season data.
[0016] Optionally, the historical growth environment data is obtained by:
[0017] obtaining historical home environment data and historical management data of the target plant, the historical management data including at least one of historical disease data, historical watering data, historical fertilization data and historical illumination data;
[0018] determining the historical growth environment data of the target plant based on the historical home environment data and the historical management data.
[0019] Optionally, the candidate disease type of the target plant and the confidence of the candidate disease type are determined based on the first questionnaire data, the first image data and the second image data by:
[0020] performing first feature extraction on the first questionnaire data to obtain first questionnaire features;
[0021] performing second feature extraction on the first image data to obtain first image features;
[0022] performing third feature extraction on the second image data to obtain second image features;
[0023] performing feature fusion processing on the first questionnaire features, the first image features and the second image features to obtain multi-modal fusion features of the target plant;
[0024] performing disease identification on the multi-modal fusion features of the target plant through a preset disease identification model to identify a plurality of candidate disease types of the target plant and a confidence of the candidate disease types.
[0025] Optionally, before the disease identification on the multi-modal fusion features of the target plant through the preset disease identification model to identify the plurality of candidate disease types of the target plant and the confidence of the candidate disease types, the method further comprises:
[0026] obtaining a training data set and an untrained disease identification model, the training data set comprising sample multi-modal fusion features and disease annotation data corresponding to the multi-modal fusion features;
[0027] training the untrained disease identification model through the training data set, in the training process, adjusting parameters of the model through a minimum loss function, and obtaining the preset disease identification model after the training is completed.
[0028] Optionally, the determining, based on the second questionnaire data, a target disease type from the plurality of candidate disease types comprises:
[0029] pushing the second questionnaire data to a user;
[0030] determining the second questionnaire features after the user fills in the second questionnaire data;
[0031] determining, based on the second questionnaire features, the target disease type from the plurality of candidate disease types.
[0032] In a second aspect, an embodiment of the present application provides a plant disease management prompting device, which comprises:
[0033] an obtaining module, configured to obtain first questionnaire data, first image data and second image data of a target plant, the first image data being overall image data of the target plant and the second image data being local image data of the target plant;
[0034] The first determining module is configured to determine a candidate disease type of the target plant and a confidence level of the candidate disease type based on the first questionnaire data, the first image data, and the second image data.
[0035] The generating module is configured to generate second questionnaire data based on the multiple candidate disease types if there are multiple candidate disease types with a confidence level greater than a confidence threshold.
[0036] The second determining module is configured to determine a target disease type from the multiple candidate disease types based on the second questionnaire data.
[0037] The pushing module is configured to determine a disease treatment scheme for the target plant based on the target disease type, and push the disease treatment scheme.
[0038] In a third aspect, an electronic device is provided, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the steps of the plant disease treatment prompting method provided in the embodiments of the present application when executing the computer program.
[0039] In a fourth aspect, a computer readable storage medium is provided, which stores a computer program, and the computer program implements the steps of the plant disease treatment prompting method provided in the embodiments of the present application when executed by a processor.
[0040] In the embodiments of the present application, the first questionnaire data, the first image data, and the second image data of the target plant are acquired, the first image data is the overall image data of the target plant, and the second image data is the local image data of the target plant; the candidate disease type of the target plant and the confidence level of the candidate disease type are determined based on the first questionnaire data, the first image data, and the second image data; the second questionnaire data is generated based on the multiple candidate disease types if there are multiple candidate disease types with a confidence level greater than a confidence threshold; the target disease type is determined from the multiple candidate disease types based on the second questionnaire data; and the disease treatment scheme for the target plant is determined based on the target disease type, and the disease treatment scheme is pushed. The present application can accurately identify the cause of plant diseases, so that the plant diseases can be treated in time, and the problem that the plant diseases cannot be treated in time due to the inability to accurately identify the cause of the diseases for the self-planted potted plants by using the existing method of artificial discovery and experience treatment is solved. BRIEF DESCRIPTION OF DRAWINGS
[0041] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0042] Figure 1 This is a flowchart of a plant disease management prompting method provided in an embodiment of the present invention;
[0043] Figure 2 This is a schematic diagram of the structure of a plant disease management and alerting device provided in an embodiment of the present invention;
[0044] Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation
[0045] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0046] like Figure 1 As shown, Figure 1 This is a flowchart of a plant disease management prompting method provided by an embodiment of the present invention. The plant disease management prompting method includes the following steps:
[0047] 101. Obtain the first questionnaire data, the first image data, and the second image data of the target plant.
[0048] In this embodiment of the invention, the above-described plant disease management prompting method can be applied to a plant management platform. This platform can be built on a server-based or distributed architecture and includes a data interface (for sensor or user uploads), a knowledge database, and a knowledge database construction program. The data interface can be used to acquire first questionnaire data, first image data, and second image data of the target plant. The knowledge database construction program can be used to construct the knowledge database, which is specifically designed to provide additional relational information for identified data entities, thereby enhancing the data recognition system's understanding of the content.
[0049] The target plants mentioned above can be potted plants, such as snake plant, pothos, fiddle-leaf fig, clivia, etc.
[0050] The first questionnaire data can be understood as growth environment data of the target plant collected in the form of a questionnaire, such as planting environment, fertilization, etc.
[0051] The first image data is overall image data of the target plant, and the first image data can be visual information of the whole plant, such as shape, color, etc.
[0052] The second image data is partial image data of the target plant, such as image data of parts such as branches and leaves, fruits, branches, roots, etc.
[0053] The image data can be a photo of the target plant taken by a camera or a mobile phone or the like.
[0054] It should be noted that the first questionnaire data, the first image data, and the second image data are obtained to better understand the growth data of the target plant.
[0055] 102, Based on the first questionnaire data, the first image data, and the second image data, the candidate disease type of the target plant and the confidence of the candidate disease type are determined.
[0056] In the embodiments of the present application, the target plant can be identified based on the first questionnaire data, the first image data, and the second image data, and the candidate disease type of the target plant and the confidence of the candidate disease type are determined. The disease identification can be understood as a recognition process of classifying and identifying diseases of the target plant. The disease type can be understood as the type of disease suffered by the target plant, such as powdery mildew, gray mold, anthracnose, etc.
[0057] The candidate disease type can be understood as an alternative disease type, such as a candidate disease type of powdery mildew, gray mold, and anthracnose.
[0058] The confidence of the candidate disease type can be understood as the confidence of the candidate disease type as the correct disease type, and the confidence is used to measure the confidence of the candidate disease type as the correct disease type. It can be understood that the higher the confidence of the candidate disease type, the higher the confidence as the correct disease type.
[0059] It should be noted that the target plant can be identified by a disease identification model, and a plurality of candidate disease types of the target plant and the confidence of the candidate disease types are identified. The disease identification model can be a disease identification model based on deep learning or machine learning, such as PPLC-Net, YOLOv7, etc. The disease identification model can automatically identify and classify different types of plant diseases by analyzing image data of plant leaves, fruits, etc.
[0060] 103、If there are multiple candidate disease types with a confidence greater than the confidence threshold, the second questionnaire data is generated based on the multiple candidate disease types.
[0061] In the embodiments of the present application, the confidence threshold is a confidence threshold preset by the system.
[0062] Further, if there are multiple candidate disease types with a confidence greater than the confidence threshold, the second questionnaire data is generated based on the multiple disease types.
[0063] The second questionnaire data is generated based on the multiple candidate disease types with a confidence greater than the confidence threshold.
[0064] It should be noted that the second questionnaire data is used to further confirm the specific performance and characteristics of the candidate disease type.
[0065] 104、Based on the second questionnaire data, a target disease type is determined from the multiple candidate disease types.
[0066] In the embodiments of the present application, the target disease type can be understood as the disease type corresponding to the target plant.
[0067] Specifically, the second questionnaire data can be feature extracted to determine the second questionnaire features, and the disease recognition model can be used to recognize the disease of the target plant according to the second questionnaire features, and determine the target disease type of the target plant from the multiple candidate diseases. The disease recognition model can be a disease recognition model based on deep learning or machine learning, such as PPLC-Net, YOLOv7, etc. The disease recognition model can automatically recognize and classify different types of plant diseases by analyzing image data of plant leaves, fruits, etc.
[0068] 105、Based on the target disease type, a disease treatment scheme for the target plant is determined and the disease treatment scheme is pushed.
[0069] In the embodiments of the present application, the disease treatment scheme can be understood as a scheme for developing a strategy to control or eliminate the pathogen according to the target disease type, so as to improve the disease resistance of the plant and change the environmental conditions to be not conducive to the growth of the pathogen.
[0070] Specifically, a corresponding disease treatment scheme can be developed according to the target disease type suffered by the target plant, and the disease treatment scheme is pushed to relevant personnel or system. The pushing can be through short message, email, APP, etc.
[0071] In a possible embodiment, for example, the target disease type of Ficus microcarpa L. is the symptom of powdery mildew, and the disease treatment method and drug of powdery mildew can be determined by analyzing the disease condition and cause of powdery mildew, such as the disease treatment scheme of improving the growth environment, regularly checking and removing the plant leaves infected with the pathogen, using appropriate fungicides, and preventing the spread of the pathogen, and the disease treatment scheme is pushed to the relevant personnel.
[0072] In the embodiment of the present application, the first questionnaire data, the first image data and the second image data of the target plant are obtained, the first image data is the overall image data of the target plant, and the second image data is the local image data of the target plant; based on the first questionnaire data, the first image data and the second image data, the candidate disease type of the target plant and the confidence of the candidate disease type are determined; if there are multiple candidate disease types with a confidence greater than a confidence threshold, the second questionnaire data is generated based on the multiple candidate disease types; based on the second questionnaire data, the target disease type is determined from the multiple candidate disease types; based on the target disease type, the disease treatment scheme of the target plant is determined, and the disease treatment scheme is pushed. The present application can accurately identify the cause of plant disease, so that the plant disease can be treated in time, and the problem that the plant disease cannot be treated in time due to the inability to accurately identify the cause of the disease of the potted plant planted by the personnel themselves is solved.
[0073] It can be understood that in the specific embodiments of the present application, plant data, image data, knowledge data, seasonal data, environmental data and other related data are involved, and when the embodiments of the present application are applied to specific products or technologies, the permission or consent of the user needs to be obtained, and the collection, use and processing of related data, as well as the training, deployment and calling of algorithm models, need to comply with relevant laws, regulations and standards of countries and regions.
[0074] Optionally, in the step of obtaining the first questionnaire data of the target plant, the current growth environment data and the historical growth environment data of the target plant can be obtained; based on the current growth environment and the historical growth environment, the first questionnaire data of the target plant is generated.
[0075] In the embodiment of the present application, the above-mentioned current growth environment data can be understood as the position of the target plant at present, such as the room, the window, the balcony and the like, and the surrounding light, temperature, humidity and the like.
[0076] The above-mentioned historical growth environment data can be understood as the record of the past growth of the target plant, including the growth speed, the disease occurrence and the like.
[0077] The first questionnaire data can be collected in the form of a questionnaire about the current growth environment data and the historical growth environment data of the target plant.
[0078] The target plant can be a household potted plant, such as a green vine, a cactus, a dracaena, etc.
[0079] Optionally, in the step of obtaining the current growth environment data of the target plant, the location of the target plant and the current season data can be obtained; and the current growth environment data of the target plant is determined based on the location of the target plant and the current season data.
[0080] In the embodiment of the present application, the location of the target plant can be understood as the place or environment where the target plant grows, including the geographical position, indoor, outdoor, balcony, garden, etc. The location of the target plant is obtained according to the regional level position and the household level position of the target plant. The regional level position can be understood as the regional position, such as city, province, country, etc. The household level position can be understood as the household position, such as the type of house, orientation, floor, ventilation state, etc.
[0081] The current season data includes the season in which the current time period is located and the current climate data. The season can be spring, summer, autumn and winter. The climate data includes the average value and the change range of temperature, precipitation, humidity, etc. The climate data reflects the basic characteristics of the cold, warm, dry and wet of the region. The climate data is collected by weather stations, satellites, etc.
[0082] The current growth environment data is obtained according to the location of the target plant and the current season data, including the room, the window nearby, the balcony, etc. where the target plant is located, and the current season, the current climate data, etc.
[0083] Optionally, in the step of obtaining the historical growth environment data, the historical household environment data and the historical management data of the target plant can be obtained; and the historical growth environment data of the target plant is determined based on the historical household environment data and the historical management data.
[0084] In the embodiment of the present application, the historical household environment data can be understood as the geographical position, the climate data and other environment data affecting the growth of the plant in the past time period.
[0085] The historical management data can be understood as the care and management of the target plant in the past period of time, and the historical management data includes at least one of historical disease data, historical watering data, historical fertilization data, and historical light data. The historical disease data can be understood as information such as the type, frequency, and severity of the plant diseases and insect pests suffered by the target plant in the past period of time. The historical watering data can be understood as information such as the watering frequency, watering amount, and time of the target plant in the past period of time. The historical fertilization data can be understood as information such as the type and amount of fertilizer applied to the target plant in the growth process in the past period of time. The historical light data can be understood as information such as the light intensity and duration received by the target plant in the past period of time.
[0086] The historical growth data is obtained according to the historical home environment data and the historical management data, and includes information such as growth speed and disease occurrence.
[0087] Optionally, in the step of determining the candidate disease type of the target plant and the confidence of the candidate disease type based on the first questionnaire data, the first image data, and the second image data, the first questionnaire data can be subjected to first feature extraction to obtain first questionnaire features; the first image data can be subjected to second feature extraction to obtain first image features; the second image data can be subjected to third feature extraction to obtain second image features; the first questionnaire features, the first image features, and the second image features can be subjected to feature fusion processing to obtain multi-modal fusion features of the target plant; and the multi-modal fusion features of the target plant can be subjected to disease recognition by a preset disease recognition model to recognize multiple candidate disease types of the target plant and the confidence of the candidate disease types.
[0088] In the embodiment of the present application, the first questionnaire features can be collected in the form of a questionnaire about the current growth environment data and the historical growth environment data of the target plant.
[0089] The first feature extraction can be understood as a processing process of extracting useful information or features from the first questionnaire data.
[0090] The first questionnaire features are information extracted from the first questionnaire data, such as growth environment, maintenance condition, and the like.
[0091] The first image data is the overall image data of the target plant, such as the overall shape, overall color, and the like.
[0092] The second feature extraction can be understood as a processing process of extracting useful information or features from the first image data.
[0093] The first image feature is feature information extracted from the overall image data of the target plant, such as overall shape, overall color, and the like.
[0094] The second image data is local image data of the target plant, such as image data of branches and leaves, fruits, stems, root systems, and the like.
[0095] The third feature extraction can be understood as a processing process of extracting useful information or features from the second image data.
[0096] The second image feature is feature information extracted from the local image data of the target plant, such as branches and leaves, fruits, stems, root systems, and the like.
[0097] The feature fusion can be understood as a processing process of forming a new comprehensive feature by integrating the first questionnaire feature, the first image feature, and the second image feature through data integration technology. The data integration technology is a data integration method of constructing a unified view by extracting, converting, and loading multi-source heterogeneous data.
[0098] The multi-modal fusion feature can be understood as a more comprehensive representation of the target plant feature formed by integrating the first questionnaire feature, the first image feature, and the second image feature.
[0099] The preset disease identification model can be a disease identification model constructed based on deep learning or machine learning, such as PPLC-Net, YOLOv7, and the like. The disease identification model can automatically identify and classify different types of plant diseases by analyzing the multi-modal fusion feature of the target plant.
[0100] The disease identification can be understood as a recognition process of classifying and identifying diseases of the target plant by the preset disease identification model.
[0101] The candidate disease type can be understood as a candidate disease type, such as powdery mildew, gray mold, anthracnose, and the like.
[0102] The confidence of the candidate disease type can be understood as the confidence degree of the candidate disease type as the correct disease type. The confidence is used to measure the confidence degree of the candidate disease type as the correct disease type. It can be understood that the higher the confidence of the candidate disease type, the higher the confidence degree as the correct disease type.
[0103] Optionally, before the step of identifying the multiple candidate disease types of the target plant and the confidence of the candidate disease types by identifying the multi-modal fusion features of the target plant through the preset disease identification model, the training data set and the untrained disease identification model can also be obtained; the untrained disease identification model is trained through the training data set, and in the training process, the model is adjusted in parameters through the minimum loss function, the training is completed, and the preset disease identification model is obtained.
[0104] In the embodiment of the application, the training data set includes sample multi-modal fusion features and disease annotation data corresponding to the multi-modal fusion features.
[0105] The untrained disease identification model can be a disease identification model constructed based on deep learning or machine learning, such as PPLC-Net, YOLOv7, etc.
[0106] The training can be supervised training, which is a method of training a model using a set of data with known labels, and the model can predict the labels of new data or make decisions based on the characteristics of existing data by optimizing the model parameters.
[0107] The loss function is used to evaluate and optimize the performance of the model. The loss function can be a mean square error loss function, a cross-entropy loss function, etc.
[0108] The parameter adjustment can be understood as updating the model parameters through an optimization algorithm during the model training process to minimize the loss function, thereby improving the performance of the model. The optimization algorithm can be gradient descent, stochastic gradient descent, etc.
[0109] The preset disease identification model can automatically identify and classify different types of plant diseases by analyzing the multi-modal fusion features of the target plant.
[0110] It should be noted that the fusion features of the target plant can be identified by the preset disease identification model to identify multiple candidate disease types of the target plant.
[0111] Optionally, in the step of determining the target disease type from the multiple candidate disease types based on the second questionnaire data, the second questionnaire data can be pushed to the user; after the user fills out the second questionnaire data, the second questionnaire features are determined; and based on the second questionnaire features, the target disease type is determined from the multiple candidate disease types.
[0112] In the embodiment of the application, the second questionnaire data is generated according to the multiple candidate disease types when the confidence of the candidate disease types is greater than the confidence threshold.
[0113] The second questionnaire feature is information extracted from the second questionnaire data. Specifically, the second questionnaire feature can be obtained by performing feature extraction on the completed second questionnaire data.
[0114] It should be noted that after obtaining the second questionnaire data, the second questionnaire data can be pushed to the user, and after the user completes the questionnaire, the completed second questionnaire data is extracted to obtain the second questionnaire feature, and the target disease type of the target plant is determined from the plurality of disease types according to the second questionnaire feature.
[0115] In the embodiments of the present application, the target disease type of the plant can be identified in real time according to the questionnaire data of the user, which helps the user better understand and handle the disease problems of the plant.
[0116] As shown in Figure 2 The plant disease management prompting device provided by the embodiments of the present application comprises:
[0117] The acquisition module 201 is configured to acquire first questionnaire data, first image data and second image data of a target plant, wherein the first image data is overall image data of the target plant, and the second image data is local image data of the target plant.
[0118] The first determination module 202 is configured to determine candidate disease types of the target plant and confidence levels of the candidate disease types based on the first questionnaire data, the first image data and the second image data.
[0119] The generation module 203 is configured to generate second questionnaire data based on a plurality of candidate disease types with confidence levels greater than a confidence threshold, if there are a plurality of candidate disease types with confidence levels greater than the confidence threshold.
[0120] The second determination module 204 is configured to determine a target disease type from the plurality of candidate disease types based on the second questionnaire data.
[0121] The push module 205 is configured to determine a disease management scheme of the target plant based on the target disease type, and push the disease management scheme.
[0122] Optionally, the acquisition module 201 is further configured to acquire current growth environment data and historical growth environment data of the target plant, and generate the first questionnaire data of the target plant based on the current growth environment and the historical growth environment.
[0123] Optionally, the acquisition module 201 is further configured to acquire a location where the target plant is located and current season data, and determine current growth environment data of the target plant based on the location where the target plant is located and the current season data.
[0124] Optionally, the acquisition module 201 is further configured to acquire historical home environment data of the target plant and historical management data, the historical management data including at least one of historical disease data, historical watering data, historical fertilization data and historical illumination data, and determine historical growth environment data of the target plant based on the historical home environment data and the historical management data.
[0125] Optionally, the first determination module 202 is further configured to perform first feature extraction on the first questionnaire data to obtain first questionnaire features, perform second feature extraction on the first image data to obtain first image features, perform third feature extraction on the second image data to obtain second image features, perform feature fusion processing on the first questionnaire features, the first image features and the second image features to obtain multi-modal fusion features of the target plant, and perform disease identification on the multi-modal fusion features of the target plant through a preset disease identification model to identify a plurality of candidate disease types of the target plant and a confidence of the candidate disease types.
[0126] Optionally, the apparatus is further configured to acquire a training data set and an untrained disease identification model, the training data set including sample multi-modal fusion features and disease label data corresponding to the multi-modal fusion features, and train the untrained disease identification model through the training data set, in which, in a training process, the model is adjusted in parameters through a minimum loss function, and after training is completed, a preset disease identification model is obtained.
[0127] Optionally, the second determination module 204 is further configured to push the second questionnaire data to a user, determine the second questionnaire features after the user fills in the second questionnaire data, and determine a target disease type from the plurality of candidate disease types based on the second questionnaire features.
[0128] As shown in Figure 3 the embodiment of the present application further provides an electronic device, which includes a processor, and the processor can execute any one of the plant disease management prompting methods.
[0129] Specifically, the electronic device includes a processor 301 and a memory 302, and a computer program for executing the plant disease management prompting method stored in the memory 302 and capable of running on the processor 301, wherein:
[0130] The processor 301 runs a computer program of a plant disease treatment suggestion method stored in the memory 302, and performs the following steps:
[0131] Obtain first questionnaire data of a target plant, first image data of the target plant, and second image data of the target plant, wherein the first image data is overall image data of the target plant, and the second image data is local image data of the target plant;
[0132] Based on the first questionnaire data, the first image data, and the second image data, determine a candidate disease type of the target plant and a confidence level of the candidate disease type;
[0133] If there are multiple candidate disease types with a confidence level greater than a confidence threshold, generate second questionnaire data based on the multiple candidate disease types;
[0134] Based on the second questionnaire data, determine a target disease type from the multiple candidate disease types;
[0135] Based on the target disease type, determine a disease treatment plan for the target plant, and push the disease treatment plan.
[0136] Optionally, the processor 301 performs the obtaining of the first questionnaire data of the target plant, comprising:
[0137] Obtain current growth environment data and historical growth environment data of the target plant;
[0138] Based on the current growth environment and the historical growth environment, generate the first questionnaire data of the target plant.
[0139] Optionally, the processor 301 performs the obtaining of the current growth environment data of the target plant, comprising:
[0140] Obtain a location of the target plant and current season data;
[0141] Based on the location of the target plant and the current season data, determine the current growth environment data of the target plant.
[0142] Optionally, the processor 301 performs the obtaining of the historical growth environment data, comprising:
[0143] Obtain historical home environment data and historical management data of the target plant, wherein the historical management data comprises at least one of historical disease data, historical watering data, historical fertilization data, and historical lighting data;
[0144] Based on the historical home environment data and the historical management data, determine the historical growth environment data of the target plant.
[0145] Optionally, the processor 301 executes the method of determining a candidate disease type of the target plant and a confidence of the candidate disease type based on the first questionnaire data, the first image data, and the second image data, comprising:
[0146] performing first feature extraction on the first questionnaire data to obtain first questionnaire features;
[0147] performing second feature extraction on the first image data to obtain first image features;
[0148] performing third feature extraction on the second image data to obtain second image features;
[0149] performing feature fusion processing on the first questionnaire features, the first image features, and the second image features to obtain multi-modal fusion features of the target plant;
[0150] performing disease recognition on the multi-modal fusion features of the target plant through a preset disease recognition model to recognize a plurality of candidate disease types of the target plant and a confidence of the candidate disease types.
[0151] Optionally, before the processor 301 performs the disease recognition on the multi-modal fusion features of the target plant through the preset disease recognition model to recognize a plurality of candidate disease types of the target plant and a confidence of the candidate disease types, the processor 301 further executes the method, comprising:
[0152] obtaining a training data set and an untrained disease recognition model, wherein the training data set comprises sample multi-modal fusion features and disease annotation data corresponding to the multi-modal fusion features;
[0153] training the untrained disease recognition model through the training data set, wherein in the training process, the model is adjusted in parameters through a minimum loss function, and after the training is completed, a preset disease recognition model is obtained.
[0154] Optionally, the processor 301 executes the method of determining a target disease type from a plurality of candidate disease types based on the second questionnaire data, comprising:
[0155] pushing the second questionnaire data to a user;
[0156] determining the second questionnaire features after the user fills out the second questionnaire data;
[0157] determining the target disease type from the plurality of candidate disease types based on the second questionnaire features.
[0158] The embodiment of the present application further provides a computer readable storage medium, and the computer readable storage medium stores a computer program. The computer program is executed by a processor to implement each process of the plant disease treatment prompting method provided by the embodiment of the present application, and the same technical effects can be achieved. To avoid repetition, details are not described herein.
[0159] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by a computer program to instruct related hardware. The program can be stored in a computer readable storage medium, and when the program is executed, the processes of the above-mentioned embodiment methods can be included. The storage medium can be a magnetic disc, an optical disc, a read-only memory (ROM) or a random access memory (RAM).
[0160] The above only describes the preferred embodiments of the present application, and of course cannot limit the scope of the present application. Therefore, equivalent changes made according to the claims of the present application are still within the scope of the present application.
Claims
1. A method for providing guidance on plant disease management, characterized in that, The method includes the following steps: Acquire first questionnaire data, first image data, and second image data of the target plant, wherein the first image data is overall image data of the target plant, and the second image data is partial image data of the target plant; Based on the first questionnaire data, the first image data, and the second image data, the candidate disease types of the target plant and the confidence level of the candidate disease types are determined. If there are multiple candidate disease types with confidence levels greater than the confidence threshold, then a second questionnaire is generated based on the multiple candidate disease types. Based on the second questionnaire data, the target disease type is determined from the multiple candidate disease types; Based on the target disease type, a disease control plan for the target plant is determined, and the disease control plan is pushed out.
2. The plant disease control and prevention method as described in claim 1, characterized in that, The acquisition of the first questionnaire data for the target plant includes: Obtain current and historical growth environment data for the target plant; Based on the current and historical growth environments, the first questionnaire data for the target plant is generated.
3. The plant disease management and guidance method as described in claim 2, characterized in that, The acquisition of the target plant's current growth environment data includes: Obtain the location and current season data of the target plant; Based on the location of the target plant and the current season data, the current growth environment data of the target plant is determined.
4. The plant disease management and guidance method as described in claim 2, characterized in that, The acquisition of historical growth environment data includes: Obtain historical home environment data and historical management data of the target plant. The historical management data includes at least one of historical disease data, historical watering data, historical fertilization data, and historical light data. Based on the historical home environment data and the historical management data, the historical growth environment data of the target plant is determined.
5. The plant disease management and guidance method as described in claim 1, characterized in that, The step of determining the candidate disease types of the target plant and the confidence level of the candidate disease types based on the first questionnaire data, the first image data, and the second image data includes: The first feature is extracted from the first questionnaire data to obtain the first questionnaire feature; The first image data is subjected to second feature extraction to obtain the first image features; The second image data is subjected to a third feature extraction to obtain the second image features; The first questionnaire features, the first image features, and the second image features are subjected to feature fusion processing to obtain the multimodal fusion features of the target plant; By using a preset disease identification model, the multimodal fusion features of the target plant are used to identify diseases, thereby identifying multiple candidate disease types of the target plant and the confidence level of the candidate disease types.
6. The plant disease management and notification method as described in claim 5, characterized in that, Before identifying multiple candidate disease types and their confidence levels by using a preset disease identification model to perform disease identification on the multimodal fusion features of the target plant, the method further includes: Obtain a training dataset and an untrained disease identification model. The training dataset includes sample multimodal fusion features and disease annotation data corresponding to the multimodal fusion features. The untrained disease identification model is trained using the training dataset. During the training process, the parameters of the model are adjusted using a minimum loss function. Once training is complete, a preset disease identification model is obtained.
7. The plant disease management and notification method as described in claim 1, characterized in that, The step of determining the target disease type from multiple candidate disease types based on the second questionnaire data includes: The second questionnaire data will be sent to the user; After the user completes the second questionnaire data, the characteristics of the second questionnaire are determined; Based on the features of the second questionnaire, the target disease type is determined from among the multiple candidate disease types.
8. A plant disease management alert device, characterized in that, The plant disease management alert device includes: The acquisition module is used to acquire first questionnaire data, first image data and second image data of the target plant, wherein the first image data is the overall image data of the target plant and the second image data is the partial image data of the target plant; The first determining module is used to determine the candidate disease types of the target plant and the confidence level of the candidate disease types based on the first questionnaire data, the first image data and the second image data. The generation module is used to generate second questionnaire data based on multiple candidate disease types if there are multiple candidate disease types with confidence scores greater than a confidence threshold. The second determining module is used to determine the target disease type from among the multiple candidate disease types based on the second questionnaire data; The push module is used to determine the disease treatment plan for the target plant based on the target disease type, and to push the disease treatment plan.
9. An electronic device, characterized in that, include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the steps of the plant disease management prompting method as described in any one of claims 1 to 7.
10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the steps of the plant disease management prompting method as described in any one of claims 1 to 7.