Intelligent medicine administration method and system based on internet of things

Abstract

The application relates to an intelligent medicine application method and system based on the Internet of Things, which comprises the following steps: collecting image information of crops in a monitoring area and soil environment characteristic information of the monitoring area, and pre-processing the collected information; establishing a comparison database to store image characteristic information of various information crops in various growth stages, soil environment characteristic information of crop growth, image characteristic information of various pests, and information of medicine application types and proportions of crops in different growth conditions; establishing a deep convolutional neural network model, and performing deep learning and training on the neural network model through training sample data; identifying and judging crop growth conditions, soil environment information, whether pests exist, whether medicine application is needed, and medicine application types in the monitoring area through the deep convolutional neural network model; and selecting medicine application types and proportions for adjustment according to result information after identification and judgment, and applying medicine to the monitoring area through a self-adaptive working mode.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of Internet of Things, the technical field of image recognition and the technical field of agricultural environment detection, and particularly relates to an intelligent pesticide application method and system based on Internet of Things. BACKGROUND

[0002] Grain is the basis of human survival. China's agricultural planting has developed for thousands of years, and has accumulated a lot of agricultural planting experience. Different responses are adopted for different crops. Corresponding drugs are applied to crops suffering from diseases and insect pests to prevent losses caused by diseases and insect pests. When taking measures, manual application of drugs has the disadvantages of uneven spraying, overuse of drugs, and the like, which may pollute the environment and affect the growth of crops. How to better respond to insect pests and diseases without affecting the surrounding environment and the growth of crops is an important problem.

[0003] Nowadays, agricultural science and technology is combined with the experience of planting crops in the past. The growth condition of crops is identified through image recognition technology. Whether there is a potential harm is predicted according to the growth condition of crops and the soil environment condition. Corresponding measures are taken to respond according to the identification and prediction. SUMMARY

[0004] The present application overcomes the defects of the prior art and provides an intelligent pesticide application method and system based on Internet of Things, which mainly aims to improve the efficiency and accuracy of agricultural pesticide application.

[0005] To achieve the above-mentioned purpose, the present application provides an intelligent pesticide application method based on Internet of Things, which comprises the following steps:

[0006] Collecting image information of crops in a monitoring area and environmental characteristic information of the monitoring area, and preprocessing the collected information;

[0007] Establishing a comparison database to store image characteristic information of various crops at different growth stages, soil environment characteristic information of crop growth, image characteristic information of various pests, pesticide type information and proportion information of crops under different growth conditions;

[0008] Establishing a deep convolutional neural network model, and performing deep learning and training on the neural network model through test samples;

[0009] Identifying and judging the growth condition of crops, soil environment information, whether there are pests, whether pesticide application is needed and the type of pesticide application in the monitoring area through the deep convolutional neural network model;

[0010] According to the result information after identification and judgment, selecting pesticide type and proportion for adjustment, and applying pesticide to the monitoring area through self-adaptive working mode;

[0011] In the scheme, the image information of crops in the monitoring area and the environmental feature information of the monitoring area are collected, and the collected information is preprocessed, specifically:

[0012] Real-time image information of various crops located in the monitoring area is collected by a high-definition camera device;

[0013] The soil environmental feature information in the monitoring area is collected by remote sensing technology;

[0014] The environmental feature information includes soil environmental temperature information, soil environmental humidity information, and information about insects and microorganisms in the soil;

[0015] The soil environmental temperature information and the soil environmental humidity information are collected by a high-sensitivity sensor;

[0016] The real-time collected image information and soil environmental feature information are preprocessed by noise reduction, filtering and screening.

[0017] In the scheme, the contrast database is established to store the image feature information of various crops at each growth stage, the soil environmental feature information of crop growth, the image feature information of various pests, the pesticide type information and proportion information of crops in different growth conditions, specifically including:

[0018] Collect the image information of each growth stage of each crop that has occurred historically;

[0019] Collect the pest image feature information that has occurred historically;

[0020] Collect the soil environmental feature information suitable for the growth of each crop;

[0021] Collect the use type and proportion information of pesticides for crops in different growth conditions;

[0022] Establish a contrast database to store the above-mentioned collected information data, which is used for comparison and judgment with the output value of the neural network.

[0023] In the scheme, the deep convolutional neural network model is established, and the neural network model is deeply learned and trained through test samples, specifically including:

[0024] The deep convolutional neural network model is established, including the identification of the deep convolutional neural network model and the prediction of the deep convolutional neural network model;

[0025] The identification of the neural network model is divided into crop identification neural network model and pest identification neural network model

[0026] The prediction of the neural network model is divided into pest and disease prediction neural network model;

[0027] The deep convolutional neural network model is deeply learned and trained through the training sample data;

[0028] Image information of different crops is collected, the image information is preprocessed to obtain growth stage information of the crops, the susceptible disease types of the growth stage are searched, image information of the diseased crops is obtained according to the susceptible disease types, and the image information of the diseased crops is taken as training sample data;

[0029] Through the searched susceptible disease types, soil environment feature information causing the diseases is collected and taken as training sample data;

[0030] The deep convolutional neural network model is deeply learned and trained through the training sample data;

[0031] The established neural network model is deeply learned and trained by using an automatic segmentation method, and network parameters and features are automatically learned;

[0032] Through continuous deep learning and training of the deep convolutional neural network model, the deviation between the calculated value and the expected value is reduced, and a deep convolutional neural network model with a deviation value within an acceptable error range is obtained.

[0033] In the scheme, the deep convolutional neural network model is used to identify and determine the growth of crops in the monitoring area, soil environment information, whether there are pests, whether pesticide application is needed, and the type of pesticide application, and specifically includes:

[0034] The collected soil environment feature information and crop image information are input into the identification deep convolutional neural network model to obtain an output value;

[0035] The output value is calculated for similarity with the data in the comparison database by using a similarity calculation method;

[0036] If the similarity value is greater than a threshold value, any one or several of the identification result information of the poor growth of crops in the monitoring area, the poor soil environment, and the existence of pests in the monitoring area is obtained;

[0037] If the similarity value is less than a threshold value, any one or several of the identification result information of the good growth of crops in the monitoring area, the good soil environment, and the non-existence of pests is obtained;

[0038] The identification result information is input into a prediction deep convolutional neural network model to predict whether there are potential pests in the soil of the monitoring area, and prediction result information is obtained;

[0039] According to the identification result information and the prediction result information, whether pesticide application is needed, the type and proportion of pesticide application are determined, and result information after the determination is obtained.

[0040] In the scheme, the type and proportion of the pesticide are selected according to the result information after the identification and judgment, and the monitoring area is sprayed by the adaptive working mode, and the adaptive working mode comprises the following steps in particular:

[0041] According to the result information after the identification and judgment, the working equipment automatically adjusts the pesticide or fertilizer according to the preset proportion;

[0042] Similar influence factors are classified into the same category by the Euclidean clustering algorithm, and the operation criteria for different situations are set to realize the adaptive working;

[0043] The adaptive working mode is adopted to adapt to various environmental influence factors during pesticide application, and the spraying direction, spraying intensity and spraying time are automatically adjusted;

[0044] After the above steps are completed, the working equipment starts spraying the pesticide in the preset area according to the preset trajectory.

[0045] To achieve the above purpose, the second aspect of the present application provides an intelligent pesticide application system based on the Internet of Things, which comprises:

[0046] The information acquisition module is divided into a first acquisition unit and a second acquisition unit;

[0047] The first acquisition unit is used for acquiring image information of crops in the monitoring area, and is composed of a plurality of high-definition shooting devices;

[0048] The second acquisition unit is used for acquiring soil environment characteristic information of the monitoring area, and is composed of a remote sensing shooting device and a high-sensitivity sensor array.

[0049] By acquiring the image information of crops in the monitoring area and the soil environment characteristic information of the monitoring area in real time, the various conditions of crops in the monitoring area and the soil environment conditions are understood in real time;

[0050] The data processing module is divided into a first processing unit and a second processing unit;

[0051] The first processing unit is used for pre-processing the acquired image information of crops in the monitoring area and the soil environment characteristic information of the monitoring area, and preliminarily eliminating interference data affecting identification and judgment.

[0052] The second processing unit is used for identifying and judging the pre-processed image information of crops and the soil environment characteristic information of the monitoring area, identifying and judging the growth conditions of crops and the conditions of the soil environment, predicting potential harm conditions in the monitoring area, and judging whether pesticide application is needed and the type and proportion of pesticide application;

[0053] The control and implementation module is divided into a first control unit, a second control unit and an implementation unit;

[0054] The first control unit is used for remote control of equipment operation, and the second control unit is used for on-site control of equipment operation.

[0055] The implementation unit is used to execute various work instructions, collect information, and apply pesticides;

[0056] The allocation and storage module is divided into an allocation unit and a storage unit;

[0057] The dispensing unit receives the judgment result information and automatically selects the appropriate type and proportion of pesticide for dispensing.

[0058] The storage unit is used to store medicines for preventing crop diseases and pests, and facilitates the quick and easy preparation of the required proportions of medicines.

[0059] The visualization module is used to display the judgment results, soil environmental information in the monitoring area, crop growth status in the monitoring area, and the working status of each device.

[0060] This invention discloses a pesticide application method and system based on the Internet of Things (IoT). One IoT-based intelligent pesticide application method includes: collecting image information of crops and environmental features of a monitored area; preprocessing the collected information; establishing a comparison database to store image feature information of various crops at different growth stages, soil environmental features of crop growth, image feature information of various pests, and information on pesticide types and proportions for crops under different growth conditions; establishing a deep convolutional neural network model, and performing deep learning and training on the neural network model using test samples; and identifying and judging pesticides in the monitored area using the deep convolutional neural network model. The system monitors crop growth, soil environment, pest presence, pesticide application requirements, and pesticide types. Based on the identified information, it selects and adjusts pesticide types and proportions, applying pesticides to the monitored area using an adaptive operating mode. Intelligent identification and judgment effectively provide real-time insights into crop growth and soil conditions, improving awareness of potential hazards. The use of IoT technology enhances equipment intelligence, enabling remote control and improving ease of use. The adaptive application method improves accuracy and automatically controls pesticide dosage, significantly reducing the risk of environmental pollution from overuse. Attached Figure Description

[0061] To more clearly illustrate the technical solutions in the embodiments or examples of the present invention, the drawings used in the embodiments or examples 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 according to these drawings without creative effort.

[0062] Figure 1 A flowchart illustrating an intelligent pesticide application method based on the Internet of Things (IoT) according to an embodiment of the present invention;

[0063] Figure 2 A flowchart illustrating information processing in an IoT-based intelligent pesticide application system, as provided in an embodiment of the present invention;

[0064] Figure 3 This is a structural diagram of an IoT-based intelligent pesticide application system provided in an embodiment of the present invention;

[0065] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0066] To better understand the above-mentioned objectives, features, and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in these embodiments can be combined with each other.

[0067] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and therefore the scope of protection of the invention is not limited to the specific embodiments disclosed below.

[0068] Figure 1 A flowchart illustrating an intelligent pesticide application method based on the Internet of Things (IoT) according to an embodiment of the present invention;

[0069] like Figure 1 As shown, the present invention provides a flowchart of an intelligent pesticide application method based on the Internet of Things, including:

[0070] S102, collect image information of crops in the monitoring area and soil environmental characteristics information of the monitoring area, and preprocess the collected information;

[0071] Real-time image information of various crops located in the monitoring area is collected using high-definition camera equipment;

[0072] Remote sensing technology was used to collect information on soil environmental characteristics in the monitoring area;

[0073] The environmental characteristic information includes soil environmental temperature information, soil environmental humidity information, and information on insects and microorganisms in the soil;

[0074] Soil ambient temperature and humidity information are collected using high-sensitivity sensors.

[0075] The real-time acquired image information and soil environmental feature information are preprocessed by noise reduction, filtering and screening.

[0076] Furthermore, the preprocessing of the real-time acquired image information and soil environmental feature information, such as noise reduction and filtering, improves the stability, accuracy and clarity of the acquired data, making it easier to compare and identify.

[0077] S104. Establish a comparison database to store image feature information of various crops at various growth stages, soil environment feature information of various crops, image feature information of various pests, and information on the types and proportions of pesticides applied to crops under different growth conditions.

[0078] Collect image information of various crops at different growth stages throughout history;

[0079] Collect image feature information of historically occurring pests;

[0080] Collect information on soil environmental characteristics suitable for the growth of various crops;

[0081] Collect information on the types and proportions of pesticides used on crops under different growth conditions;

[0082] Establish a comparison database to store the information data collected above, which will be used to compare and judge with the output value of the neural network;

[0083] Furthermore, the collected image information of crops at various growth stages is used to compare and identify the crop image information acquired in real time, thereby identifying the growth stage and condition of crops in the monitoring area.

[0084] Furthermore, the collected pest image feature information is used to identify pests that may be present in real-time crop image information, facilitating timely application of pesticides.

[0085] Furthermore, the collected information on crop soil environmental characteristics is compared and calculated with the real-time collected information on soil environmental characteristics to determine the condition of the soil environment in the monitoring area, predict whether there are potential hazards in the environment, and thus take preventive measures in advance.

[0086] S106, Establish a deep convolutional neural network model, and perform deep learning and training on the neural network model using training sample data;

[0087] Establish deep convolutional neural network models, including identifying and predicting deep convolutional neural network models;

[0088] Recognition neural network models are divided into crop recognition neural network models and pest recognition neural network models.

[0089] Predictive neural network models are divided into disease and pest prediction neural network models;

[0090] The above deep convolutional neural network model is learned and trained using training sample data.

[0091] Collect image information of different crops, preprocess the image information to obtain crop growth stage information, retrieve the susceptible species of the growth stage, obtain image information of diseased crops based on the susceptible species, and use the image information of diseased crops as training sample data.

[0092] By retrieving susceptible species, we collect soil environmental characteristics that lead to these diseases and use them as training sample data.

[0093] The deep convolutional neural network model is deeply learned and trained using training sample data;

[0094] The established neural network model is trained using an automatic segmentation method to automatically learn network parameters and features;

[0095] By continuously performing deep learning and training on the deep convolutional neural network model, the deviation between the calculated value and the expected value is reduced, resulting in a deep convolutional neural network model with a deviation value within an acceptable error range.

[0096] Furthermore, the training sequence for training the deep convolutional neural network model is as follows: first, train the recognition neural network model to obtain a recognition neural network model that meets expectations; then, use the output value of the trained neural network model and the training sample data to train the prediction neural network model, and finally obtain a neural network model with prediction accuracy within an acceptable error range.

[0097] S108 uses a deep convolutional neural network model to identify and determine the crop growth, soil environment, presence of pests, need for pesticide application, and type of pesticide in the monitoring area.

[0098] The collected soil environmental features and crop image information are input into a recognition deep convolutional neural network model to obtain the output value;

[0099] The similarity calculation method is used to calculate the similarity between the output value and the data in the comparison database;

[0100] If the similarity value is greater than the threshold, the identification result information of any one or more of the following in the monitoring area is obtained: poor crop growth, poor soil environment, and the presence of pests in the monitoring area.

[0101] If the similarity value is less than the threshold, the identification result information of any one or more of the following in the monitoring area is obtained: good crop growth, poor soil environment, and absence of pests.

[0102] The identification result information is input into a value prediction deep convolutional neural network model to predict whether there are potential pests in the soil of the monitoring area.

[0103] If the similarity value is greater than the threshold, then the prediction result information of potential hazards in the soil of the monitoring area is obtained;

[0104] If the similarity value is less than the threshold, the prediction result information is that there is no potential hazard in the soil of the monitored area;

[0105] Based on the identification and prediction results, it is determined whether medication is needed, as well as the type and proportion of medication to be administered, thus obtaining the result information.

[0106] S110, based on the results of identification and judgment, selects the type and proportion of pesticides to be applied and adjusts them, and applies pesticides to the monitoring area through an adaptive working mode;

[0107] Based on the identification and judgment results, the working equipment automatically mixes pesticides or fertilizers according to a preset ratio;

[0108] By using Euclidean clustering algorithm, similar influencing factors are grouped into the same category, and operational benchmarks are set for different situations to achieve the adaptive operation.

[0109] It adopts an adaptive working mode to adapt to various environmental factors during pesticide application, and automatically adjusts the spraying direction, spraying intensity, and spraying time.

[0110] After the above steps are completed, the working equipment will automatically start spraying pesticides in the preset area according to the preset trajectory.

[0111] Furthermore, the monitoring area is divided into multiple sub-regions. The degree of pests and diseases in crops within the monitoring area is determined by identifying and judging the results and predicting the results. The deviation of the degree of pests and diseases in the sub-regions is calculated. Based on the different degrees of pests and diseases in each sub-region, the sub-regions are classified, and the average degree of pests and diseases in the same category of sub-regions is calculated. The average degree of pests and diseases in the same category of sub-regions is obtained. The average value is used as the average benchmark for pesticide application in the same category of sub-regions. Based on the average benchmark, the optimal pesticide application plan is formulated, and the optimal pesticide type and ratio, spray droplet size and rate are selected. This improves the utilization rate of pesticide application, avoids crop losses caused by over-application, and reduces the risk of environmental pollution.

[0112] Furthermore, from the classified sub-regions, areas with severe disease are selected, and environmental features of these areas are extracted. The areas near these areas are defined as high-risk areas, and environmental feature parameters of these high-risk areas are extracted. The Manhattan distance between the environmental feature parameters of the severely affected areas and the high-risk areas is calculated. The calculated Manhattan distance is used as the environmental feature similarity value between the severely affected areas and the high-risk areas. This similarity value is compared with a preset threshold. The relationship between the similarity value and the preset threshold is used to determine the probability of disease in the high-risk areas. Based on the determination results, the corresponding optimal medication plan is adopted to effectively prevent the disease severity in other areas from increasing.

[0113] Furthermore, by calculating the proportion of diseased features in crop images collected within the region to the overall basic features of the crop, such as calculating the percentage of diseased leaf area in diseased crops to the total leaf area of ​​the crop, the severity of pests and diseases within the region can be obtained. The severity of pests and diseases is divided into severe, moderate, and very low. Based on the severity of pests and diseases, the severity and probability of disease in sub-regions can be determined, and the optimal pesticide application plan can be adopted for diseased areas and high-risk areas.

[0114] It should be noted that by collecting real-time crop image information and soil environment information in the monitoring area, the crop growth status and soil environment status of the monitoring area can be identified and judged. Based on the identification and judgment results, predictions can be made to prevent potential hazards and take countermeasures in advance.

[0115] It should be noted that after obtaining the identification and prediction results, the system matches the corresponding solutions based on the results information and automatically adjusts the pesticides according to the types and proportions in the solution. Through an adaptive working mode, the system controls the flow rate and spraying direction of pesticides based on different environmental factors when spraying pesticides, which greatly improves the intelligence of pesticide application and also ensures environmental protection.

[0116] Figure 2 A flowchart illustrating information processing in an IoT-based intelligent pesticide application system, as provided in an embodiment of the present invention;

[0117] like Figure 2 As shown, the present invention provides a flowchart of information processing for an IoT-based intelligent pesticide application system, including:

[0118] S202, Obtain crop image information and soil environment information in the monitored area;

[0119] S204, preprocessing the collected information data;

[0120] The collected information data is preprocessed with noise reduction, filtering and other methods to improve the stability and clarity of the collected data, making it easier to compare and identify.

[0121] S206, Identification and judgment are performed by recognizing a neural network model, and the results are compared and calculated with data in the comparison database;

[0122] The similarity calculation method is used to calculate the similarity between the output value and the data in the comparison database;

[0123] If the similarity value is greater than the threshold, the identification result information of any one or more of the following in the monitoring area is obtained: poor crop growth, poor soil environment, and the presence of pests in the monitoring area.

[0124] If the similarity value is less than the threshold, the identification result information of any one or more of the following in the monitoring area is obtained: good crop growth, poor soil environment, and absence of pests.

[0125] The identification result information is input into a value prediction deep convolutional neural network model to predict whether there are potential pests in the soil of the monitoring area.

[0126] If the similarity value is greater than the threshold, then the prediction result information of potential hazards in the soil of the monitoring area is obtained;

[0127] If the similarity value is less than the threshold, the prediction result information is that there is no potential hazard in the soil of the monitored area;

[0128] Based on the identification and prediction results, it is determined whether medication is needed, as well as the type and proportion of medication to be administered, to obtain the result information after the determination.

[0129] S208, the recognition result information is obtained and input into the predictive neural network model;

[0130] The identification result information is input into a value prediction deep convolutional neural network model to predict whether there are potential pests in the soil of the monitoring area.

[0131] If the similarity value is greater than the threshold, then the prediction result information of potential hazards in the soil of the monitoring area is obtained;

[0132] If the similarity value is less than the threshold, the prediction result information is that there is no potential hazard in the soil of the monitored area;

[0133] S210, obtain the prediction result information;

[0134] S212, determine the types and proportions of drugs to be used by identifying and predicting results;

[0135] Based on the identification and prediction results, it is determined whether medication is needed, as well as the type and proportion of medication to be administered, to obtain the result information after the determination.

[0136] S214, obtain the judgment result information;

[0137] It should be noted that the recognition results of the neural network model are crucial information for judging intelligent pesticide application and predicting potential hazards. By collecting image information, the model identifies the growth status of crops, whether they are infested with pests and the types of pests, as well as the condition of the soil environment. Based on the recognition results, predictions and judgments can be made to predict potential hazards, determine the appropriate response, and automatically match the type and proportion of pesticides used. This greatly improves the accuracy of pesticide application, ensuring targeted treatment and avoiding losses.

[0138] Figure 3 This is a structural diagram of an IoT-based intelligent pesticide application system provided in an embodiment of the present invention;

[0139] like Figure 3 As shown in the diagram, this invention provides a structural diagram of an intelligent pesticide application system based on the Internet of Things, comprising:

[0140] Information acquisition module, data processing module, control and implementation module, allocation and storage module, visualization module;

[0141] The information acquisition module is divided into a first acquisition unit and a second acquisition unit;

[0142] The first acquisition unit is used to acquire image information of crops in the monitoring area and consists of multiple high-definition imaging devices;

[0143] The second acquisition unit is used to collect soil environmental characteristic information in the monitoring area, and consists of remote sensing imaging equipment and a high-sensitivity sensor array.

[0144] By collecting real-time image information of crops and soil environmental characteristics in the monitoring area, we can understand the various conditions of crops and soil environment in the monitoring area in real time.

[0145] The data processing module is divided into a first processing unit and a second processing unit;

[0146] The first processing unit is used to preprocess the image information of crops in the monitoring area and the soil environmental characteristics information of the monitoring area to initially eliminate interference data that affects identification and judgment.

[0147] The second processing unit is used to identify and judge the pre-processed crop image information and the soil environmental characteristics information of the monitoring area, to identify and judge the growth status of the crops and the condition of the soil environment, to predict potential hazards in the monitoring area, and to determine whether pesticides need to be applied and the type and proportion of pesticides to be applied.

[0148] The control and implementation module is divided into a first control unit, a second control unit, and an implementation unit;

[0149] The first control unit is used for remote control of equipment operation, and the second control unit is used for on-site control of equipment operation.

[0150] The implementation unit is used to execute various work instructions, collect information, and apply pesticides;

[0151] Furthermore, the first control unit is located in the terminal device, which can be a smart device with control functions such as a computer, mobile phone or console. The working status of the device can be remotely controlled through the first control unit in the terminal device.

[0152] Furthermore, the second control unit is a control unit set on the device side, used to control the normal operation of each module on the device side, and at the same time execute the terminal's real-time work instructions;

[0153] Furthermore, the normal operation of each module on the device can be controlled by operating the second control unit on the device side. The second control unit is responsible for the normal operation of all modules on the device side, and the first control unit remotely controls the second control unit to make the device side operate normally.

[0154] Furthermore, the implementation module consists of various working devices that perform daily tasks by executing various operating instructions.

[0155] The allocation and storage module is divided into an allocation unit and a storage unit;

[0156] The dispensing unit receives the judgment result information and automatically selects the appropriate type and proportion of pesticide for dispensing.

[0157] The storage unit is used to store medicines for preventing crop diseases and pests, and facilitates the quick and easy preparation of the required proportions of medicines.

[0158] The visualization module is divided into terminal display and device display, which is used to display the judgment results, soil environmental information in the monitoring area, crop growth status in the monitoring area, and working status of each device.

[0159] Furthermore, the terminal display involves showing the prediction results, the identification and judgment results, and various real-time collected data on a terminal device, which can be a smart device with display capabilities, such as a computer, mobile phone, or console.

[0160] Furthermore, the device-side display involves showing the prediction results, identification and judgment results, and various real-time collected data on a display screen installed on the device. The display screen is a screen that can display information data, and it can be a screen with touch function or a screen with only display function.

[0161] It should be noted that, based on the control unit and display module of the present invention, users can control the operation of the device from any end, and can also query the real-time collected data, identification and judgment results and prediction results from any end. This makes it convenient for users to understand the crop growth status and soil environment of the monitoring area in real time, and to make secondary judgments on the monitoring area based on their own experience, which greatly improves the convenience and ease of use.

[0162] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods, such as: multiple units or components can be combined, or integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the various components shown or discussed can be through some interfaces, and the indirect coupling or communication connection between devices or units can be electrical, mechanical, or other forms.

[0163] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary hardware platforms. Through a system combining software and hardware, intelligent drug application methods can be realized in a more intelligent, systematic, and precise manner. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a combination of software products and hardware facilities. The system program is stored on a storage medium such as ROM / RAM, magnetic disk, optical disk, or other media that can store program code. The system program is implemented through a terminal device, which can be a computer, cloud server, or network device, etc., executing the software control embodiment of the present invention. The actual operation of the acquisition and driving embodiment needs to be implemented through hardware devices.

[0164] Alternatively, if the aforementioned modules of this invention are implemented as a complete or partial product and sold or used as independent products, they can also be stored in a computer-readable storage medium. Based on this understanding, the technical solutions of the embodiments of this invention, or the parts that contribute to the prior art, can be embodied in the form of software products and hardware. The software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, ROM / RAM, magnetic disks, or optical disks.

[0165] The above are merely preferred embodiments of the present invention and do not limit the scope of the patent. Any equivalent structural or procedural transformations made based on the description and drawings of the present invention, or direct or indirect applications in other related technical fields, are similarly included within the scope of patent protection of the present invention.

Claims

1. A smart pesticide application method based on the Internet of Things, characterized in that, include: Image information of crops in the monitoring area and soil environmental characteristics of the monitoring area are collected, and the collected information is preprocessed. Establish a comparative database to store image feature information of various crops at different growth stages, soil environment feature information of various crops, image feature information of various pests, and information on the types and proportions of pesticides applied to crops under different growth conditions. A deep convolutional neural network model is established, and the deep convolutional neural network model is deeply learned and trained using training sample data; The deep convolutional neural network model is used to identify and determine the crop growth, soil environment, presence of pests, need for pesticide application, and type of pesticide in the monitoring area. Based on the results of the identification and judgment, the type and proportion of pesticides are selected and adjusted, and pesticides are applied to the monitoring area through an adaptive working mode. The feature is that it uses a deep convolutional neural network model to identify and determine the crop growth status, soil environmental information, presence of pests, need for pesticide application, and type of pesticide within the monitoring area, specifically including: The collected soil environmental features and crop image information are input into a recognition deep convolutional neural network model to obtain the output value; The similarity calculation method is used to calculate the similarity between the output value and the data in the comparison database; If the similarity value is greater than the threshold, the identification result information of any one or more of the following in the monitoring area is obtained: poor crop growth, poor soil environment, and the presence of pests in the monitoring area. If the similarity value is less than the threshold, the identification result information of any one or more of the following in the monitoring area is obtained: good crop growth, good soil environment, and absence of pests. The identification result information is input into the prediction depth convolutional neural network model to predict whether there are potential pests in the soil of the monitoring area, and the prediction result information is obtained. Based on the identification and prediction results, it is determined whether medication is needed, as well as the type and proportion of medication to be administered, thus obtaining the result information.

2. The intelligent pesticide application method based on the Internet of Things according to claim 1, characterized in that, Image information of crops and environmental features in the monitoring area are collected, and the collected information is preprocessed, specifically including: Collect real-time image information of various crops located in the monitoring area; Collect information on soil environmental characteristics in the monitoring area; The environmental characteristic information includes soil temperature information, soil moisture information, and information on insects and microorganisms in the soil; The real-time acquired image information and environmental feature information are preprocessed by noise reduction, filtering and screening.

3. The intelligent pesticide application method based on the Internet of Things according to claim 1, characterized in that, Establish a comparative database to store image feature information of various crops at different growth stages, environmental feature information suitable for the growth of various crops, image feature information of various pests, and information on the types and proportions of pesticides applied to crops under different growth conditions. Specifically, this includes: Collect image information of various crops at different growth stages throughout history; Collect image feature information of historically occurring pests; Collect information on soil environmental characteristics suitable for the growth of various crops; Collect information on the proportion of pesticides applied to crops under different growth conditions and the types of pesticides used; Establish a comparison database to store the information data collected above.

4. The intelligent pesticide application method based on the Internet of Things according to claim 1, characterized in that, A deep convolutional neural network model is established, and the model is subjected to deep learning and training using test samples. Specifically, this includes: Establish deep convolutional neural network models, including identifying and predicting deep convolutional neural network models; Collect soil environmental characteristics information and image information of diseased crops as training sample data; The deep convolutional neural network model is deeply learned and trained using training sample data; The established neural network model is trained using an automatic segmentation method, which automatically learns the network parameters and features.

5. The intelligent pesticide application method based on the Internet of Things according to claim 1, characterized in that, Based on the identification and judgment results, the type and proportion of pesticide are selected and adjusted, and the pesticide is applied to the monitoring area through an adaptive working method, specifically including: Based on the identification and judgment results, the working equipment automatically mixes pesticides or fertilizers according to a preset ratio; By using Euclidean clustering algorithm, similar influencing factors are grouped into the same category, and operational benchmarks are set for different situations to achieve the adaptive working mode. It adopts an adaptive working mode to adapt to various environmental factors during pesticide application, and automatically adjusts the spraying direction, spraying intensity, and spraying time. After the above steps are completed, the working equipment will automatically start spraying pesticides in the preset area according to the preset trajectory.

6. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes an Internet of Things (IoT)-based smart drug application method program, which, when executed by a processor, implements the steps of the IoT-based smart drug application method as described in any one of claims 1 to 5.

Images