Disease identification-medication guidance-agricultural service closed-loop system based on applet

Abstract

The invention provides a disease identification-medication guidance-agricultural service closed-loop system based on an applet, and the system comprises a front-end lightweight application layer which is used for calling a user terminal camera to obtain or select a diseased crop picture from a user album based on a WeChat applet as a carrier when an operation request is received through a user interface, and transmitting the diseased crop picture to a user terminal; receiving the context information of the diseased crop uploaded by the user; the rear-end intelligent service layer is used for analyzing the diseased crop pictures based on a crop disease and insect pest model, determining disease and insect pest types, and performing comprehensive analysis in combination with context information of the diseased crops and a built-in agricultural knowledge base to obtain a personalized prevention and control scheme; and the data and business support layer is used for analyzing the personalized prevention and control scheme based on the docking pesticide enterprise system, determining a recommended product and a purchase guide, and triggering an agricultural service reservation request in response to user selection. Diseases can be timely prevented and controlled, economic loss is reduced, pesticide abuse is reduced through precise pesticide application, and the quality safety of agricultural products is guaranteed.

Description

Technical Field

[0001] This invention relates to the field of agricultural artificial intelligence technology, and in particular to a closed-loop system based on a mini-program that integrates "disease identification - pesticide application guidance - agricultural services". Background Technology

[0002] Currently, when faced with crop diseases and pests, agricultural producers typically rely on personal experience or consultation with agricultural technicians for identification and pesticide application. This method is inefficient and its accuracy is difficult to guarantee. Although some independent disease and pest identification apps or pesticide query tools exist, their functions are limited and they suffer from the following technical shortcomings: Information silos: The identification, query, and service purchase processes are fragmented, requiring users to switch between different applications, which is cumbersome and data cannot be shared, thus creating information silos. Low level of decision-making intelligence: Most systems only provide simple information matching and lack the ability to integrate multi-dimensional information for comprehensive reasoning and intelligent decision-making, thus failing to generate truly personalized prevention and control solutions; Service chain disruption: Most existing solutions stop at providing advice and fail to effectively connect the diagnosis results with the implementation links such as pesticide purchase and professional agricultural services, thus failing to form a complete closed loop for solving the problem. High technical threshold: It requires a separate APP and supporting services, and it is difficult to get rid of the constraints of expensive hardware, which also brings inconvenience to users. Therefore, in order to overcome the above-mentioned defects, the present invention provides a closed-loop system based on a mini-program that integrates "disease identification - pesticide application guidance - agricultural services". Summary of the Invention

[0003] This invention provides a closed-loop system based on WeChat mini-programs, integrating "disease identification - pesticide guidance - agricultural services." This system offers convenient crop disease identification services via WeChat mini-programs. Users simply need to take a picture of the diseased crop and upload relevant background information to quickly obtain accurate disease and pest diagnoses and personalized control plans. It is "use and go" without installation. Combining crop disease and pest identification models, a crop disease and pest knowledge base, pesticide company systems, intelligent analysis, and big data technology, it can automatically recommend suitable pesticide products and provide purchase guides. It also supports agricultural service appointments, achieving a one-stop closed-loop service from problem identification to solution. This significantly improves the efficiency and accuracy of agricultural production, helps farmers control diseases in a timely manner, reduces economic losses, reduces pesticide abuse through precise pesticide application, lowers agricultural non-point source pollution, ensures the quality and safety of agricultural products, and improves crop yield and quality. Simultaneously, it promotes the digitalization and intelligent development of agricultural services.

[0004] This invention provides a closed-loop system based on a mini-program that integrates "disease identification - pesticide application guidance - agricultural services", including: The front-end lightweight application layer is used to carry out operations based on WeChat Mini Programs. When an operation request is received through the user interface, it calls the user terminal camera to obtain or select images of diseased crops from the user's album. At the same time, it receives context information of the diseased crops uploaded by the user. The backend intelligent service layer is used to analyze images of diseased crops based on crop disease and pest models, determine the types of diseases and pests, and perform comprehensive analysis by combining the contextual information of the diseased crops and the built-in agricultural knowledge base to obtain personalized prevention and control solutions. The data and business support layer is used to analyze personalized prevention and control solutions based on the connected pesticide company systems, determine recommended products and purchase guides, respond to user selections, and trigger agricultural service appointment requests.

[0005] Preferably, a closed-loop system based on a mini-program, encompassing "disease identification - pesticide application guidance - agricultural services," includes a lightweight front-end application layer comprising: The status monitoring unit is used to monitor the image acquisition operation request submitted by the user based on the user interface, and after detecting the image acquisition operation request, it sends a camera call request or a user album access request to the user terminal based on the image acquisition interface of the WeChat mini program. Image acquisition unit, used for: After authorization is requested, the viewfinder is displayed based on the user interface, and the background configuration data of the user terminal is called to adaptively adjust the working parameters of the camera. The system monitors the user's confirmation action on the user interface in real time, and when a confirmation action is detected, it controls the user terminal's camera to capture an image of the current frame to obtain an image of the diseased crop. Once the user's album access request is approved, images of diseased crops are selected from the user's upload requests.

[0006] Preferably, a closed-loop system based on a mini-program for "disease identification - pesticide application guidance - agricultural services" includes an image acquisition unit comprising: Image preprocessing subunit, used for: The system acquires images of diseased crops and obtains the target time and geographical location of the diseased crop images based on the user terminal. Based on the target time information and geographical location, generate accompanying descriptions for images of diseased crops, and associate and bind the images of diseased crops with the accompanying descriptions; Based on the association binding results, images of diseased crops and accompanying descriptions from different locations are converted into a predetermined format; The data upload subunit is used to upload images of diseased crops and accompanying descriptions from different locations to the data center based on the conversion results, monitor the network status in real time, and initiate a breakpoint resume mechanism when the network is abnormal.

[0007] Preferably, a closed-loop system based on a mini-program, encompassing "disease identification - pesticide application guidance - agricultural services," includes a lightweight front-end application layer comprising: A multimodal interaction unit is used to provide a multimodal interaction window to the user based on the user interface, and to start the corresponding data stream receiving process after the user makes a selection. The multimodal interaction window includes images, text and voice. The context information determination unit is used to receive the context information submitted by the user for the diseased crop based on the start-up result control data flow receiving process, and to encapsulate the obtained context information with the corresponding diseased crop image.

[0008] Preferably, a closed-loop system based on a mini-program, encompassing "disease identification - pesticide application guidance - agricultural services," includes a backend intelligent service layer comprising: Image retrieval unit, used for: The system acquires images of diseased crops and analyzes the configuration parameters of the crop disease and pest model to determine the limiting parameters of the crop disease and pest model for the input images. Based on the specified parameters, the images of diseased crops are standardized and then the standardized images of diseased crops are input into the crop disease and pest model. The pest and disease analysis unit is used for: Based on the crop disease and pest model, historical data with labeled disease and pest types are learned, and the attention weights of different key points of interest in image processing are adapted based on the learning results to obtain the target attention mechanism. Based on crop disease and pest model and target attention mechanism, multi-level feature extraction is performed on input diseased crop images to obtain visual features corresponding to the diseased crop images, and the obtained visual features are matched with the benchmark features of known disease and pest types. Based on the matching results, the images of diseased crops are output as the probability distribution of each known disease and pest type. Based on the probability distribution, the known disease and pest types whose probability values ​​exceed the preset threshold are determined as the final disease and pest types. The personalized prevention and control plan determination unit is used for: The contextual information of diseased crops is analyzed to determine the crop type, growth cycle and geographical region corresponding to the diseased crop, and multi-dimensional conditions are generated based on the type of pest and disease, crop type, growth cycle and geographical region. A global scan of the built-in agricultural knowledge base is performed to extract the structured features of crops, pests and diseases, pesticides and control methods from the built-in agricultural knowledge base. Based on multi-dimensional conditions, the structured features are retrieved to obtain personalized control solutions for diseased crops.

[0009] Preferably, a closed-loop system based on a mini-program, encompassing "disease identification - pesticide application guidance - agricultural services," includes a personalized prevention and control plan determination unit, comprising: The scheme defines sub-units for: When the final pest or disease type is a single type, obtain the conditional retrieval results based on the structured features under multi-dimensional conditions, and retrieve the corresponding basic prevention and control plan based on the conditional retrieval results. Extract the basic pest and disease quantitative indicators corresponding to the basic prevention and control plan, and determine the correction amount of the basic prevention and control plan based on the deviation values ​​between the multi-dimensional conditions and the basic pest and disease quantitative indicators. Based on the correction amount, the basic prevention and control plan is modified to obtain the personalized prevention and control plan for the diseased crop. When the final pest or disease type is not a single type, the basic prevention and control plan corresponding to each type is retrieved based on the conditional retrieval results of the structured features under multi-dimensional conditions. Based on the structured characteristics of crops, pests, pesticides and control methods in the built-in agricultural knowledge base, the correlation characteristics between different types of basic control programs are determined. Based on the correlation characteristics, the basic control programs corresponding to different types are synergistically and compatibly modified to obtain composite and personalized control programs for diseased crops.

[0010] Preferably, a closed-loop system based on a mini-program, encompassing "disease identification - pesticide application guidance - agricultural services," includes a data and business support layer comprising: The scheme analysis unit is used to analyze the obtained personalized prevention and control schemes, determine the key prevention and control elements contained in the personalized prevention and control schemes, and upload the obtained key prevention and control elements to the connected pesticide company system. Product determination unit, used for: Based on the pesticide enterprise system, the key prevention and control elements are prioritized for matching. Based on the results of the prioritization, the integrated product matching engine is called to perform a cyclical and progressive matching of each key prevention and control element with the basic status parameters of each product in the product database. Based on the cyclical progressive matching results, recommended products corresponding to personalized prevention and control solutions are obtained.

[0011] Preferably, a closed-loop system based on a mini-program, encompassing "disease identification - pesticide application guidance - agricultural services," includes a product determination unit comprising: The product information determination subunit is used to trace the source of the recommended products and determine the corresponding product price, inventory, pesticide company information and distribution location. Purchase a tour guide generation unit for: Based on the results of the connection with the pesticide company's system, an online purchase link is generated according to product price, inventory, pesticide company information and distribution location; At the same time, an offline supplier location map is generated based on pesticide company information and distribution locations; By encapsulating online purchase links and offline supplier location maps, a purchase guide for recommended products can be obtained.

[0012] Preferably, a closed-loop system based on a mini-program, encompassing "disease identification - pesticide application guidance - agricultural services," includes a data and business support layer comprising: The behavior monitoring unit is used to: monitor the user's selection behavior under the purchase guide in real time based on the user interface, and determine the user's purchase selection needs, including online purchase and offline purchase. Agricultural service units, used for: When the purchase request is selected as online purchase, an online order request is generated based on the recommended products and sent to the connected pesticide company's system to complete the online order placement; When the purchase request is set to offline purchase, the agricultural service appointment request process is initiated, and the user's selected appointment time slot, submitted user information, and request details are received. Based on the reservation time slot, user information, and demand details, a reservation request order is generated and sent to the corresponding agricultural service terminal to complete the reservation.

[0013] Preferably, a closed-loop system based on a mini-program, encompassing "disease identification - pesticide application guidance - agricultural services," includes a data and business support layer comprising: The information aggregation unit is used to summarize the obtained results of pest and disease types, personalized prevention and control plans, recommended products, purchase guides, and agricultural service reservation requests. The information display unit is used to display the summary results on the user interface.

[0014] Compared with the prior art, the beneficial effects of the present invention are as follows: This WeChat mini-program provides a convenient crop disease identification service. Users simply need to take a picture of the diseased crop and upload relevant background information to quickly obtain accurate disease and pest diagnoses and personalized control solutions. It's "use and go" without installation. Combining crop disease and pest identification models, a crop disease and pest knowledge base, pesticide company systems, intelligent analysis, and big data technology, it can automatically recommend suitable pesticide products and provide purchase guides. It also supports agricultural service appointments, achieving a one-stop closed-loop service from problem identification to solution. This significantly improves the efficiency and accuracy of agricultural production, helps farmers control diseases in a timely manner, reduces economic losses, reduces pesticide overuse through precise pesticide application, lowers agricultural non-point source pollution, ensures the quality and safety of agricultural products, and improves crop yield and quality. Simultaneously, it promotes the digitalization and intelligent development of agricultural services.

[0015] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in this application.

[0016] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0017] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings: Figure 1 This is a structural diagram of a closed-loop system based on a mini-program for "disease identification - pesticide application guidance - agricultural services" in an embodiment of the present invention; Figure 2 This is a structural diagram of the front-end lightweight application layer in a closed-loop system based on a mini-program for "disease identification - pesticide guidance - agricultural services" in an embodiment of the present invention. Figure 3 This is a structural diagram of the backend intelligent service layer in a closed-loop system based on a mini-program for "disease identification - pesticide guidance - agricultural services" in an embodiment of the present invention. Detailed Implementation

[0018] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0019] Example 1: This example provides a closed-loop system based on a mini-program that integrates "disease identification - pesticide application guidance - agricultural services," such as... Figure 1 As shown, it includes: The front-end lightweight application layer is used to carry out operations based on WeChat Mini Programs. When an operation request is received through the user interface, it calls the user terminal camera to obtain or select images of diseased crops from the user's album. At the same time, it receives context information of the diseased crops uploaded by the user. The backend intelligent service layer is used to analyze images of diseased crops based on crop disease and pest models, determine the types of diseases and pests, and perform comprehensive analysis by combining the contextual information of the diseased crops and the built-in agricultural knowledge base to obtain personalized prevention and control solutions. The data and business support layer is used to analyze personalized prevention and control solutions based on the connected pesticide company systems, determine recommended products and purchase guides, respond to user selections, and trigger agricultural service appointment requests.

[0020] In this embodiment, the front-end lightweight application layer refers to the user interface part developed based on the WeChat mini program, which is responsible for calling the camera to obtain crop images and collecting contextual information input by the user (such as crop type, growth stage, etc.).

[0021] In this embodiment, the user interface is an information interface that can interact with the user, using a WeChat mini program as the carrier.

[0022] In this embodiment, the contextual information of the diseased crop refers to the species, growth cycle, and growth characteristics of the diseased crop.

[0023] In this embodiment, the backend intelligent service layer refers to the server-side processing part that uses crop disease and pest models to perform image analysis and combines them with an agricultural knowledge base to generate personalized prevention and control solutions.

[0024] In this embodiment, the built-in agricultural knowledge base is pre-set and stores structured knowledge in the fields of crops, pests and diseases, pesticides, and control methods.

[0025] In this embodiment, the personalized prevention and control plan refers to the prevention and control measures that are suitable for the current type of pest or disease, determined through analysis.

[0026] In this embodiment, the data and business support layer refers to the backend module that connects to the pesticide company's system, which is responsible for parsing prevention and control plans, recommending products, and processing service reservation requests.

[0027] In this embodiment, the pesticide enterprise system refers to a platform with analysis and drug matching recommendations.

[0028] In this embodiment, the agricultural service reservation request includes reserving professional agricultural services, such as drone spraying.

[0029] In this embodiment, the process of generating a personalized prevention and control plan also includes: Natural language-based medication consultation questions are received from user input via the user interface. The intent recognition and entity extraction of pesticide consultation questions are performed to determine at least one key semantic entity from the crop category, suspected pest or disease name, or pesticide ingredient corresponding to the natural language pesticide consultation question; Generate structured knowledge query requests based on key semantic entities; The knowledge query request is input into the question-answering model, which then drives the model to perform targeted retrieval and multi-step reasoning on the built-in agricultural knowledge base. Based on the results of targeted retrieval and multi-step reasoning, a preliminary medication suggestion text is generated according to the built-in agricultural knowledge base; The initial medication recommendation text is verified for compliance and regional applicability to obtain standardized medication guidance information after verification; The standardized medication guidance information will be returned to the user interface for display.

[0030] The beneficial effects of the above technical solution are as follows: It provides convenient crop disease identification services through WeChat mini-programs. Users only need to take pictures of diseased crops and upload relevant background information to quickly obtain accurate disease and pest diagnoses and personalized control plans, achieving "use and go" without installation. Combining crop disease and pest identification models, crop disease and pest knowledge bases, pesticide company systems, intelligent analysis, and big data technology, it can automatically recommend suitable pesticide products and provide purchase guides. It also supports agricultural service appointments, realizing a one-stop closed-loop service from problem identification to solution. This significantly improves the efficiency and accuracy of agricultural production, helps farmers control diseases in a timely manner, reduces economic losses, reduces pesticide abuse through precise pesticide application, lowers agricultural non-point source pollution, ensures the quality and safety of agricultural products, and improves crop yield and quality. Simultaneously, it promotes the digital and intelligent development of agricultural services.

[0031] Example 2: Based on Example 1, this example provides a closed-loop system based on a mini-program for "disease identification - pesticide application guidance - agricultural services", such as... Figure 2 As shown, the lightweight front-end application layer includes: The status monitoring unit is used to monitor the image acquisition operation request submitted by the user based on the user interface, and after detecting the image acquisition operation request, it sends a camera call request or a user album access request to the user terminal based on the image acquisition interface of the WeChat mini program. Image acquisition unit, used for: After authorization is requested, the viewfinder is displayed based on the user interface, and the background configuration data of the user terminal is called to adaptively adjust the working parameters of the camera. The system monitors the user's confirmation action on the user interface in real time, and when a confirmation action is detected, it controls the user terminal's camera to capture an image of the current frame to obtain an image of the diseased crop. Once the user's album access request is approved, images of diseased crops are selected from the user's upload requests.

[0032] In this embodiment, the status monitoring unit refers to a functional module used to listen for the user's photo-taking command on the mini-program interface and initiate the camera call process accordingly.

[0033] In this embodiment, the image acquisition unit refers to a complete image acquisition function module responsible for retrieving the camera, adaptively adjusting parameters, displaying the viewfinder, and completing photo capture upon user confirmation. Adaptive adjustment refers to the ability to automatically adjust camera parameters based on lighting conditions and distance.

[0034] In this embodiment, the background configuration data refers to the preset system parameters related to camera performance in the user terminal device, which are used to automatically optimize settings to ensure that the captured images are clear and usable.

[0035] The beneficial effects of the above technical solution are: it can automatically listen to user operations, quickly call up and intelligently adjust the camera, so that users only need to simply align and confirm to efficiently complete the acquisition of diseased crop images and select diseased crop images from the user's album according to the user's upload requirements, which significantly reduces the operation threshold, provides a high-quality image foundation for subsequent accurate disease identification, and greatly improves user experience and system efficiency.

[0036] Example 3: Based on Example 2, this example provides a closed-loop system for "disease identification - pesticide application guidance - agricultural services" based on a mini-program. The image acquisition unit includes: Image preprocessing subunit, used for: The system acquires images of diseased crops and obtains the target time and geographical location of the diseased crop images based on the user terminal. Based on the target time information and geographical location, generate accompanying descriptions for images of diseased crops, and associate and bind the images of diseased crops with the accompanying descriptions; Based on the association binding results, images of diseased crops and accompanying descriptions from different locations are converted into a predetermined format; The data upload subunit is used to upload images of diseased crops and accompanying descriptions from different locations to the data center based on the conversion results, monitor the network status in real time, and initiate a breakpoint resume mechanism when the network is abnormal.

[0037] In this embodiment, the supplementary description refers to the metadata associated with the image of the diseased crop, which mainly includes the specific time and geographical location information of the time the image was taken.

[0038] In this embodiment, the predetermined format refers to the uniform image data specifications that the system pre-sets for easy storage, transmission and processing.

[0039] In this embodiment, the breakpoint resume mechanism refers to a technical solution that allows uploading to continue from the point of interruption when the connection is restored after an unexpected interruption in network transmission, without having to restart the entire transmission process.

[0040] The beneficial effects of the above technical solution are: it can automatically add time and location information to the collected crop images and format them in a unified manner, thereby ensuring data standardization and contextual integrity. At the same time, it can cope with network fluctuations during the upload process, ensuring that the image data is transmitted to the server stably and reliably, providing a high-quality data foundation for subsequent analysis.

[0041] Example 4: Based on Example 1, this example provides a closed-loop system based on a mini-program for "disease identification - pesticide application guidance - agricultural services," with a lightweight front-end application layer including: A multimodal interaction unit is used to provide a multimodal interaction window to the user based on the user interface, and to start the corresponding data stream receiving process after the user makes a selection. The multimodal interaction window includes images, text and voice. The context information determination unit is used to receive the context information submitted by the user for the diseased crop based on the start-up result control data flow receiving process, and to encapsulate the obtained context information with the corresponding diseased crop image.

[0042] In this embodiment, the multimodal interaction window refers to the functional interface in the unit that the user can select to receive image, text, or voice input respectively.

[0043] In this embodiment, the data stream receiving process corresponds one-to-one with the multimodal interaction window, which is a way to effectively receive different types of data.

[0044] In this embodiment, encapsulation refers to the process of combining the image of the diseased crop with its corresponding background information into a complete data packet.

[0045] The beneficial effects of the above technical solution are: by providing multiple input methods such as images, text, and voice, users can flexibly and conveniently submit background information related to diseases, and automatically integrate and bind it with crop images, ensuring the integrity and relevance of the data, providing sufficient basis for accurate backend analysis, and greatly improving the convenience of operation and the accuracy of information collection.

[0046] Example 5: Based on Example 1, this example provides a closed-loop system based on a mini-program for "disease identification - pesticide application guidance - agricultural services", such as... Figure 3 As shown, the backend intelligent service layer includes: Image retrieval unit, used for: The system acquires images of diseased crops and analyzes the configuration parameters of the crop disease and pest model to determine the limiting parameters of the crop disease and pest model for the input images. Based on the specified parameters, the images of diseased crops are standardized and then the standardized images of diseased crops are input into the crop disease and pest model. The pest and disease analysis unit is used for: Based on the crop disease and pest model, historical data with labeled disease and pest types are learned, and the attention weights of different key points of interest in image processing are adapted based on the learning results to obtain the target attention mechanism. Based on crop disease and pest model and target attention mechanism, multi-level feature extraction is performed on input diseased crop images to obtain visual features corresponding to the diseased crop images, and the obtained visual features are matched with the benchmark features of known disease and pest types. Based on the matching results, the images of diseased crops are output as the probability distribution of each known disease and pest type. Based on the probability distribution, the known disease and pest types whose probability values ​​exceed the preset threshold are determined as the final disease and pest types. The personalized prevention and control plan determination unit is used for: The contextual information of diseased crops is analyzed to determine the crop type, growth cycle and geographical region corresponding to the diseased crop, and multi-dimensional conditions are generated based on the type of pest and disease, crop type, growth cycle and geographical region. A global scan of the built-in agricultural knowledge base is performed to extract the structured features of crops, pests and diseases, pesticides and control methods from the built-in agricultural knowledge base. Based on multi-dimensional conditions, the structured features are retrieved to obtain personalized control solutions for diseased crops.

[0047] In this embodiment, the limiting parameters refer to the standardized requirements that the crop disease and pest model imposes on the input image in terms of size, format, etc.

[0048] In this embodiment, the target attention mechanism refers to an algorithm model that enables the neural network to adaptively focus on key areas (such as lesions and insect bodies) related to pests and diseases when processing images.

[0049] In this embodiment, the baseline feature refers to a standard pattern for feature comparison that is learned and stored from known types of pest and disease sample images.

[0050] In this embodiment, multi-dimensional conditions refer to a comprehensive query condition composed of identified pest and disease types, combined with information such as crop type, growth cycle, and geographical region provided by the user.

[0051] In this embodiment, structured features refer to the data format that is easy for computers to retrieve, formed by organizing professional knowledge (such as crop characteristics, pesticide information, and control methods) in the agricultural knowledge base according to specific relationships and organizational forms.

[0052] The beneficial effects of the above technical solution are as follows: by standardizing the processing of crop images uploaded by users and automatically extracting image features using deep learning models, it is possible to efficiently and accurately identify crop pest and disease types. Furthermore, by combining multi-dimensional information such as crop type, growth cycle, and geographical location provided by users, it is possible to quickly retrieve and generate highly matched personalized prevention and control plans from a massive agricultural knowledge base. This realizes intelligent decision-making from image recognition to precise pesticide application guidance, significantly improves the accuracy of disease diagnosis and the scientific nature of prevention and control measures, and effectively helps farmers reduce production losses.

[0053] Example 6: Based on Example 5, this example provides a closed-loop system based on a mini-program for "disease identification - pesticide application guidance - agricultural services," including a personalized prevention and control plan determination unit, comprising: The scheme defines sub-units for: When the final pest or disease type is a single type, obtain the conditional retrieval results based on the structured features under multi-dimensional conditions, and retrieve the corresponding basic prevention and control plan based on the conditional retrieval results. Extract the basic pest and disease quantitative indicators corresponding to the basic prevention and control plan, and determine the correction amount of the basic prevention and control plan based on the deviation values ​​between the multi-dimensional conditions and the basic pest and disease quantitative indicators. Based on the correction amount, the basic prevention and control plan is modified to obtain the personalized prevention and control plan for the diseased crop. When the final pest or disease type is not a single type, the basic prevention and control plan corresponding to each type is retrieved based on the conditional retrieval results of the structured features under multi-dimensional conditions. Based on the structured characteristics of crops, pests, pesticides and control methods in the built-in agricultural knowledge base, the correlation characteristics between different types of basic control programs are determined. Based on the correlation characteristics, the basic control programs corresponding to different types are synergistically and compatibly modified to obtain composite and personalized control programs for diseased crops.

[0054] In this embodiment, the basic prevention and control plan refers to the standard prevention and control methods stored in the agricultural knowledge base for a specific disease or pest.

[0055] In this embodiment, the basic pest and disease quantitative indicators refer to the key reference data related to the severity of pests and diseases, pesticide dosage, and timing of prevention and control in the basic prevention and control plan.

[0056] In this embodiment, the associated features refer to the interrelationships and influences between the control schemes of various pests and diseases in terms of pesticide compatibility, application order, and overall effect.

[0057] The beneficial effects of the above technical solution are: it can intelligently call and optimize basic prevention and control plans based on whether the identified pest or disease type is a single case; for a single disease, it can accurately adjust the pesticide application strategy according to the specific situation; for a complex disease, it can coordinate multiple plans to avoid conflicts and form a comprehensive prevention and control measure, ensuring the pertinence and feasibility of the provided plan and effectively improving the prevention and control effect.

[0058] Example 7: Based on Example 1, this example provides a closed-loop system based on a mini-program for "disease identification - pesticide application guidance - agricultural services," including a data and business support layer: The scheme analysis unit is used to analyze the obtained personalized prevention and control schemes, determine the key prevention and control elements contained in the personalized prevention and control schemes, and upload the obtained key prevention and control elements to the connected pesticide company system. Product determination unit, used for: Based on the pesticide enterprise system, the key prevention and control elements are prioritized for matching. Based on the results of the prioritization, the integrated product matching engine is called to perform a cyclical and progressive matching of each key prevention and control element with the basic status parameters of each product in the product database. Based on the cyclical progressive matching results, recommended products corresponding to personalized prevention and control solutions are obtained.

[0059] In this embodiment, key control elements refer to the core measures information extracted from the personalized control plan, such as the required pesticide type, mode of action, or specific physical control methods.

[0060] In this embodiment, matching priority division refers to assigning different matching orders to prevention and control elements based on their importance or urgency in the plan.

[0061] In this embodiment, the product matching engine refers to a dedicated software component or algorithm responsible for comparing and filtering prevention and control elements with records in the product database.

[0062] In this embodiment, the basic state parameters refer to the key attributes used in the product database to describe pesticide products, such as active ingredients, target pests, formulations, and toxicity.

[0063] In this embodiment, cyclic progressive matching refers to a process of using various prevention and control elements in sequence according to priority to screen products in multiple rounds and gradually refinement.

[0064] The beneficial effects of the above technical solution are: it can automatically analyze the core elements in the prevention and control plan, and intelligently match the core elements with the product database of pesticide companies. Through priority sorting and multi-round progressive screening, it can quickly lock in the most suitable recommended products, which significantly improves the accuracy of pesticide application guidance and the efficiency of product selection.

[0065] Example 8: Based on Example 7, this example provides a closed-loop system based on a mini-program for "disease identification - pesticide application guidance - agricultural services," including a product determination unit comprising: The product information determination subunit is used to trace the source of the recommended products and determine the corresponding product price, inventory, pesticide company information and distribution location. Purchase a tour guide generation unit for: Based on the results of the connection with the pesticide company's system, an online purchase link is generated according to product price, inventory, pesticide company information and distribution location; At the same time, an offline supplier location map is generated based on pesticide company information and distribution locations; By encapsulating online purchase links and offline supplier location maps, a purchase guide for recommended products can be obtained.

[0066] In this embodiment, traceability refers to determining relevant detailed information about the recommended product, such as price, inventory, manufacturer details, and distributor location.

[0067] In this embodiment, the online purchase link refers to a URL that directly redirects to the pesticide company's or partner e-commerce platform's page, where users can place an order online by clicking on it.

[0068] In this embodiment, the offline supplier location map refers to a navigation interface that marks the specific addresses of distributors in the form of a visual map to help users find the nearest physical purchase point.

[0069] In this embodiment, the purchase guide refers to comprehensive guidance information that includes online purchase links and offline supplier location maps, used to guide users to complete product purchases.

[0070] The beneficial effects of the above technical solution are: it can automatically obtain the price, inventory and sales channel information of recommended products, and generate online purchase links and offline supplier maps, providing users with convenient online and offline purchase guidance, simplifying the purchase process, and improving the efficiency and experience of purchasing medicines.

[0071] Example 9: Based on Example 1, this example provides a closed-loop system based on a mini-program for "disease identification - pesticide application guidance - agricultural services," including a data and business support layer: The behavior monitoring unit is used to monitor the user's selection behavior under the purchase guide in real time based on the user interface, and to determine the user's purchase selection needs, including online purchase and offline purchase. Agricultural service units, used for: When the purchase request is selected as online purchase, an online order request is generated based on the recommended products and sent to the connected pesticide company's system to complete the online order placement; When the purchase request is set to offline purchase, the agricultural service appointment request process is initiated, and the user's selected appointment time slot, submitted user information, and request details are received. Based on the reservation time slot, user information, and demand details, a reservation request order is generated and sent to the corresponding agricultural service terminal to complete the reservation.

[0072] In this embodiment, the purchase selection demand refers to the intention expressed by the user through operation, which is specifically divided into two types: "online purchase" and "offline purchase".

[0073] In this embodiment, an online order request refers to standardized electronic order data containing recommended product information generated by the system to complete an online transaction.

[0074] In this embodiment, the agricultural service reservation request process refers to a series of steps initiated when a user chooses to purchase offline, which involves collecting and submitting their reservation time, personal information, and specific service requirements.

[0075] In this embodiment, the reservation request order refers to a standardized order formed by the system after integrating the reservation time period, personal information and demand details submitted by the user, which is used to initiate a reservation with the service provider.

[0076] The beneficial effects of the above technical solution are: it can automatically respond to users' purchase choices. When users choose to purchase online, an order will be generated directly and the online transaction will be completed, realizing fast and convenient pesticide procurement. When users choose to purchase offline, they will be guided to submit a service appointment and the order containing time, user information and specific needs will be automatically sent to the server. This effectively connects online diagnosis and offline services, significantly improving the integrity of agricultural services and user experience.

[0077] Example 10: Based on Example 1, this example provides a closed-loop system based on a mini-program for "disease identification - pesticide application guidance - agricultural services," including a data and business support layer: The information aggregation unit is used to summarize the obtained results of pest and disease types, personalized prevention and control plans, recommended products, purchase guides, and agricultural service reservation requests. The information display unit is used to display the summary results on the user interface.

[0078] In this embodiment, the summary result refers to all the final output information that has been integrated and includes pest and disease types, personalized prevention and control plans, recommended products, purchase guides, and agricultural service appointment status.

[0079] The beneficial effects of the above technical solution are: it can automatically integrate all information from disease identification and medication plans to product purchase and service appointment, and display it clearly on the same interface, enabling users to quickly obtain complete decision support and significantly improving the convenience of information acquisition and the continuity of service experience.

[0080] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A closed-loop system based on a mini-program for "disease identification - pesticide application guidance - agricultural services", characterized in that, include: The front-end lightweight application layer is used to carry out operations based on WeChat Mini Programs. When an operation request is received through the user interface, it calls the user terminal camera to obtain or select images of diseased crops from the user's album. At the same time, it receives context information of the diseased crops uploaded by the user. The backend intelligent service layer is used to analyze images of diseased crops based on crop disease and pest models, determine the types of diseases and pests, and perform comprehensive analysis by combining the contextual information of the diseased crops and the built-in agricultural knowledge base to obtain personalized prevention and control solutions. The data and business support layer is used to analyze personalized prevention and control solutions based on the connected pesticide company systems, determine recommended products and purchase guides, respond to user selections, and trigger agricultural service appointment requests.

2. The closed-loop system based on a mini-program for "disease identification - pesticide application guidance - agricultural services" as described in claim 1, characterized in that, The lightweight front-end application layer includes: The status monitoring unit is used to monitor the image acquisition operation request submitted by the user based on the user interface, and after detecting the image acquisition operation request, it sends a camera call request or a user album access request to the user terminal based on the image acquisition interface of the WeChat mini program. Image acquisition unit, used for: After authorization is requested, the viewfinder is displayed based on the user interface, and the background configuration data of the user terminal is called to adaptively adjust the working parameters of the camera. The system monitors the user's confirmation action on the user interface in real time, and when a confirmation action is detected, it controls the user terminal's camera to capture an image of the current frame to obtain an image of the diseased crop. Once the user's album access request is approved, images of diseased crops are selected from the user's upload requests.

3. The closed-loop system based on a mini-program for "disease identification - pesticide application guidance - agricultural services" as described in claim 2, characterized in that, The image acquisition unit includes: Image preprocessing subunit, used for: The system acquires images of diseased crops and obtains the target time and geographical location of the diseased crop images based on the user terminal. Based on the target time information and geographical location, generate accompanying descriptions for images of diseased crops, and associate and bind the images of diseased crops with the accompanying descriptions; Based on the association binding results, images of diseased crops and accompanying descriptions from different locations are converted into a predetermined format; The data upload subunit is used to upload images of diseased crops and accompanying descriptions from different locations to the data center based on the conversion results, monitor the network status in real time, and initiate a breakpoint resume mechanism when the network is abnormal.

4. The closed-loop system based on a mini-program for "disease identification - pesticide application guidance - agricultural services" as described in claim 1, characterized in that, The lightweight front-end application layer includes: A multimodal interaction unit is used to provide a multimodal interaction window to the user based on the user interface, and to start the corresponding data stream receiving process after the user makes a selection. The multimodal interaction window includes images, text and voice. The context information determination unit is used to receive the context information submitted by the user for the diseased crop based on the start-up result control data flow receiving process, and to encapsulate the obtained context information with the corresponding diseased crop image.

5. A closed-loop system based on a mini-program for "disease identification - pesticide application guidance - agricultural services" as described in claim 1, characterized in that, The backend intelligent service layer includes: Image retrieval unit, used for: The system acquires images of diseased crops and analyzes the configuration parameters of the crop disease and pest model to determine the limiting parameters of the crop disease and pest model for the input images. Based on the specified parameters, the images of diseased crops are standardized and then the standardized images of diseased crops are input into the crop disease and pest model. The pest and disease analysis unit is used for: Based on the crop disease and pest model, historical data with labeled disease and pest types are learned, and the attention weights of different key points of interest in image processing are adapted based on the learning results to obtain the target attention mechanism. Based on crop disease and pest model and target attention mechanism, multi-level feature extraction is performed on input diseased crop images to obtain visual features corresponding to the diseased crop images, and the obtained visual features are matched with the benchmark features of known disease and pest types. Based on the matching results, the images of diseased crops are output as the probability distribution of each known disease and pest type. Based on the probability distribution, the known disease and pest types whose probability values ​​exceed the preset threshold are determined as the final disease and pest types. The personalized prevention and control plan determination unit is used for: The contextual information of diseased crops is analyzed to determine the crop type, growth cycle and geographical region corresponding to the diseased crop, and multi-dimensional conditions are generated based on the type of pest and disease, crop type, growth cycle and geographical region. A global scan of the built-in agricultural knowledge base is performed to extract the structured features of crops, pests and diseases, pesticides and control methods from the built-in agricultural knowledge base. Based on multi-dimensional conditions, the structured features are retrieved to obtain personalized control solutions for diseased crops.

6. A closed-loop system based on a mini-program for "disease identification - pesticide application guidance - agricultural services" as described in claim 5, characterized in that, The unit for determining personalized prevention and control plans includes: The scheme defines sub-units for: When the final pest or disease type is a single type, obtain the conditional retrieval results based on the structured features under multi-dimensional conditions, and retrieve the corresponding basic prevention and control plan based on the conditional retrieval results. Extract the basic pest and disease quantitative indicators corresponding to the basic prevention and control plan, and determine the correction amount of the basic prevention and control plan based on the deviation values ​​between the multi-dimensional conditions and the basic pest and disease quantitative indicators. Based on the correction amount, the basic prevention and control plan is modified to obtain the personalized prevention and control plan for the diseased crop. When the final pest or disease type is not a single type, the basic prevention and control plan corresponding to each type is retrieved based on the conditional retrieval results of the structured features under multi-dimensional conditions. Based on the structured characteristics of crops, pests, pesticides and control methods in the built-in agricultural knowledge base, the correlation characteristics between different types of basic control programs are determined. Based on the correlation characteristics, the basic control programs corresponding to different types are synergistically and compatibly modified to obtain composite and personalized control programs for diseased crops.

7. A closed-loop system based on a mini-program for "disease identification - pesticide application guidance - agricultural services" as described in claim 1, characterized in that, The data and business support layer includes: The scheme analysis unit is used to analyze the obtained personalized prevention and control schemes, determine the key prevention and control elements contained in the personalized prevention and control schemes, and upload the obtained key prevention and control elements to the connected pesticide company system. Product determination unit, used for: Based on the pesticide enterprise system, the key prevention and control elements are prioritized for matching. Based on the results of the prioritization, the integrated product matching engine is called to perform a cyclical and progressive matching of each key prevention and control element with the basic status parameters of each product in the product database. Based on the cyclical progressive matching results, recommended products corresponding to personalized prevention and control solutions are obtained.

8. A closed-loop system based on a mini-program for "disease identification - pesticide application guidance - agricultural services" as described in claim 7, characterized in that, The product identification unit includes: The product information determination subunit is used to trace the source of the recommended products and determine the corresponding product price, inventory, pesticide company information and distribution location. Purchase a tour guide generation unit for: Based on the results of the connection with the pesticide company's system, an online purchase link is generated according to product price, inventory, pesticide company information and distribution location; At the same time, an offline supplier location map is generated based on pesticide company information and distribution locations; By encapsulating online purchase links and offline supplier location maps, a purchase guide for recommended products can be obtained.

9. A closed-loop system based on a mini-program for "disease identification - pesticide application guidance - agricultural services" as described in claim 1, characterized in that, The data and business support layer includes: The behavior monitoring unit is used to: monitor the user's selection behavior under the purchase guide in real time based on the user interface, and determine the user's purchase selection needs, including online purchase and offline purchase. Agricultural service units, used for: When the purchase request is selected as online purchase, an online order request is generated based on the recommended products and sent to the connected pesticide company's system to complete the online order placement; When the purchase request is set to offline purchase, the agricultural service appointment request process is initiated, and the user's selected appointment time slot, submitted user information, and request details are received. Based on the reservation time slot, user information, and demand details, a reservation request order is generated and sent to the corresponding agricultural service terminal to complete the reservation.

10. A closed-loop system based on a mini-program for "disease identification - pesticide application guidance - agricultural services" as described in claim 1, characterized in that, The data and business support layer includes: The information aggregation unit is used to summarize the obtained results of pest and disease types, personalized prevention and control plans, recommended products, purchase guides, and agricultural service reservation requests. The information display unit is used to display the summary results on the user interface.

Images