Hypericum perforatum field targeted spraying device based on multispectral recognition

Abstract

The invention belongs to the technical field of field targeted spraying, and discloses a hypericum perforatum field targeted spraying device based on multispectral recognition. The device comprises a crop hazard identification and extraction module, a fixed-point pesticide spraying decision optimization module, a pesticide spraying instruction generation and transmission module, a pesticide residue prediction and traceability module and a planting standard conformity evaluation module, hazard factors are analyzed based on a standardized data set, a crop hazard report is obtained, analysis is performed based on the crop hazard report, and a pesticide residue prediction and traceability result is obtained. The method comprises the steps of obtaining a crop pesticide spraying strategy report, a pesticide spraying operation record report and a pesticide residue prediction report, evaluating the standard degree of a planting process, providing corresponding improvement suggestions, and obtaining a planting standard conformity evaluation report. The system has the remarkable advantages of high basic data reliability, high dynamic and accurate regulation and control intelligence capability and good crop full life cycle management effect.

Description

Technical Field

[0001] This invention relates to the field of targeted spraying technology, and more specifically, to a field targeted spraying device for Hypericum perforatum based on multispectral recognition. Background Technology

[0002] As an important and distinctive Chinese medicinal herb, the quality and safety of St. John's wort are of paramount importance. Field pest and disease control is a key link in ensuring its yield and quality. However, traditional manual inspection and full-coverage spraying methods have problems such as low efficiency, high risk of pesticide abuse, and inability to guarantee the traceability of medicinal material quality and safety. Therefore, how to achieve accurate identification of pests and diseases and targeted spraying based on modern technology has become an urgent need for the current Chinese medicinal herb planting industry.

[0003] However, traditional field-targeted spraying devices for Hypericum perforatum suffer from several shortcomings. First, they rely on a single or a few spectral indices to assess crop health, lacking the detailed identification of specific crop damage symptoms such as pests, nutrient deficiencies, and detailed classifications of pests and diseases. This leads to high false alarm and false negative rates, hindering precise targeted spraying. Second, traditional devices often operate solely based on the presence or absence of damage in the planting area, spraying only when damage is present. This lack of precise and effective dosage control based on specific damage conditions and environmental factors can result in excessive spraying and exacerbate environmental pollution. Third, traditional devices are limited to damage assessment and spraying decisions, failing to effectively utilize data on the planting area. This hinders effective management throughout the entire crop lifecycle, from planting to harvest. In summary, effectively addressing the issues of insufficient basic data support, inadequate dynamic and precise intelligent control, and lack of comprehensive lifecycle management in traditional devices is a crucial challenge that current field-targeted spraying devices for Hypericum perforatum need to address.

[0004] In view of this, the present invention proposes a field-targeted spraying device for Hypericum perforatum based on multispectral recognition to solve the above problems. Summary of the Invention

[0005] To overcome the aforementioned deficiencies of the prior art and to achieve the above objectives, the present invention provides the following technical solution, including:

[0006] The multispectral data acquisition module is used to acquire multispectral sensor data based on a sensor network to obtain a multispectral dataset.

[0007] Furthermore, the steps for acquiring multispectral sensor data based on sensor networks include:

[0008] S1.1: Based on the preset acquisition time, generate a multispectral data acquisition command and output the multispectral acquisition command to the sensor network for data acquisition to obtain the original spectral dataset;

[0009] S1.2: Receive the raw spectral dataset transmitted back from the sensor network based on the standard API interface, and perform integrity verification on the raw spectral dataset. Integrity verification is to calculate the checksum value of the raw spectral dataset and compare the calculated checksum value with the transmitted checksum value. If the comparison result is consistent, proceed to step S1.3. If the comparison result is inconsistent, request to retransmit the raw spectral dataset.

[0010] S1.3: Based on the spatiotemporal reference, the data items from different times and different sensor sources in the original spectral dataset are unified to obtain the spectral aligned dataset;

[0011] S1.4: Integrate the spectral alignment dataset with the aligned timestamps to obtain a multispectral dataset;

[0012] S1.5: Output the multispectral dataset to the data extraction and preprocessing module;

[0013] The data extraction and preprocessing module is used to extract features and preprocess multispectral datasets to obtain standardized datasets;

[0014] Furthermore, the steps for feature extraction and preprocessing of the multispectral dataset include:

[0015] S2.1: Perform surface reflectance correction on the multispectral dataset to obtain the corrected dataset. The specific calculation formula for the correction is as follows:

[0016]

[0017] Surface reflectance A is obtained a , where A b For the radiance collected by the sensor network in the multispectral dataset, A c For path radiation, A d θ is the solar irradiance, and θ is the solar zenith angle.

[0018] S2.2: Based on the calibration dataset, the spectral characteristics of vegetation are calculated to obtain a feature dataset, which includes vegetation features, red edge features, and soil-modified vegetation features. The specific formula set for calculation is as follows:

[0019]

[0020] Vegetation characteristics B were obtained respectively. a Red-edge feature B b and soil-regulated vegetation characteristics B c Among them, Bd B represents the surface reflectance of vegetation in the near-infrared band. e B represents the surface reflectance of vegetation in the red light band. f κ represents the surface reflectance of vegetation in the red-edge band, and κ is the soil adjustment coefficient.

[0021] S2.3: Calculate the feature dataset using the Z-score standardization formula to obtain the standardized dataset;

[0022] S2.4: Output the standardized dataset to the crop hazard identification and extraction module;

[0023] The crop hazard identification and extraction module is used to analyze hazard factors based on standardized datasets and generate crop hazard reports.

[0024] Furthermore, the steps for analyzing hazard factors based on standardized datasets include:

[0025] S3.1: Retrieve a pre-trained hazard classification model from the database, input the standardized dataset into the hazard classification model, and output the hazard probability dataset;

[0026] S3.2: Retrieve the hazard assessment decision rule table from the database and compare it with the hazard probability dataset to obtain a hazard level report;

[0027] S3.3: Package the hazard level report, the coordinate data of all data items in the hazard level report, and the comparison timestamp to obtain the crop hazard report;

[0028] S3.4: Output crop damage reports to the targeted spraying decision optimization module;

[0029] The targeted spraying decision optimization module is used to analyze crop damage reports and obtain crop spraying strategy reports.

[0030] Furthermore, the steps for analyzing crop damage reports include:

[0031] S4.1: Based on the DBSCAN density spatial clustering algorithm, the connected hazard areas in the crop hazard report are identified, and the connected hazard areas with an area less than or equal to the preset area are filtered out to obtain the hazard area.

[0032] S4.2: The priority assessment value is calculated based on the crop damage report and the area of ​​the damaged area. The specific calculation method is to obtain the priority assessment value by weighted summation of the damage level report, the area of ​​the damaged area and the distance from the healthy area.

[0033] S4.3: Calculate the spraying amount based on the area of ​​the affected area to obtain the spraying amount data. The specific formula for calculation is as follows:

[0034] C a =C b ×(1+γ1×C c +γ2×C d )×C e ;

[0035] Obtain spraying amount data C a , where C b For hazard level reporting, γ1 and γ2 are weighting coefficients, C c C represents the distance from the healthy area. d C represents canopy coverage. e The area of ​​the affected region;

[0036] S4.4: Package the boundary coordinates of the affected area, priority assessment values, and pesticide application data to obtain a crop spraying strategy report;

[0037] S4.5: Output the crop spraying strategy report to the spraying instruction generation and transmission module;

[0038] The spraying instruction generation and transmission module is used to convert crop spraying strategy reports into executable spraying instructions and transmit them to obtain a spraying operation record report.

[0039] Furthermore, the steps of converting crop spraying strategy reports into executable spraying instructions and transmitting them include:

[0040] S5.1: Based on the crop spraying strategy report, generate the first spraying instruction set, which includes spraying operation area boundary instructions and spraying quantity instructions;

[0041] S5.2: Calculate the optimal spraying path based on the crop spraying strategy report to obtain the optimal spraying operation path;

[0042] S5.3: Combine the boundary instructions for the spraying operation area, the spraying quantity instructions, and the optimal spraying operation path to obtain a comprehensive spraying instruction set;

[0043] S5.4: Output the integrated spraying command set to the intelligent controller, monitor the intelligent controller's reception results, and obtain a command reception report, which includes reception success and reception failure.

[0044] S5.5: Package the integrated spraying instruction set and instruction reception report to obtain the spraying operation record report;

[0045] S5.6: Store the spraying operation record report in the database and output it to the pesticide residue prediction and traceability module;

[0046] The pesticide residue prediction and tracing module is used to analyze historical crop damage reports and spraying operation records to obtain pesticide residue prediction reports;

[0047] Furthermore, the steps for analyzing historical crop damage reports and pesticide spraying operation records include:

[0048] S6.1: Based on the meteorological release API interface, collect meteorological data of the planting area to obtain a meteorological dataset, which includes real-time temperature data and real-time rainfall data;

[0049] S6.2: Based on the pre-trained pesticide residue prediction model retrieved from the database, historical crop damage reports, historical spraying operation records, and historical meteorological datasets are input into the pesticide residue prediction model, and the predicted pesticide residue data is output. The specific formula for prediction is as follows:

[0050] D a，t =D b ×e -k×t ;

[0051] Obtain the predicted pesticide residue data D at time point t a，t , where D b is the initial pesticide residue, and k is the degradation rate coefficient;

[0052] S6.3: Based on the predicted pesticide residue data, calculate the predicted pesticide residue data at the harvest time point and divide it by the maximum allowable residue level to obtain the pesticide residue risk value;

[0053] S6.4: Based on the residue risk threshold range (Q1, Q2), when the pesticide residue risk value is less than Q1, a low-risk label is generated; when the pesticide residue risk value is greater than or equal to Q1 and less than Q2, a medium-risk label is generated; when the pesticide residue risk value is greater than or equal to Q2, a high-risk label is generated.

[0054] Based on risk labels, generate a risk level distribution map;

[0055] S6.5: Integrate the predicted pesticide residue data into the risk level distribution map to obtain a pesticide residue prediction report;

[0056] S6.6: Output the pesticide residue prediction report to the planting standard compliance assessment module;

[0057] The planting standard compliance assessment module is used to evaluate the standardization of the planting process, provide corresponding improvement suggestions, and generate a planting standard compliance assessment report.

[0058] Furthermore, the steps provided for assessing the standardization of the planting process and offering corresponding improvement suggestions include:

[0059] S7.1: Calculate the ratio of low-risk pesticide residues to total risk pesticide residues based on pesticide residue prediction reports, and multiply by 100% to obtain crop safety.

[0060] The number of times pesticides were sprayed within a specified area is retrieved from the database and compared with the standard number of times pesticides were sprayed. When the number of times pesticides were sprayed was less than or equal to the standard number of times pesticides were sprayed, it was recorded as compliant spraying.

[0061] Calculate the ratio of compliant spraying times to the total number of spraying times, and multiply by 100% to obtain the pesticide use safety level;

[0062] S7.2: A comprehensive compliance score is obtained by comprehensively calculating the crop safety and pesticide use safety. The specific formula for the calculation is as follows:

[0063]

[0064] The overall compliance score was D. c Where N is the number of samples, ω n E represents the weight coefficient of the nth sample. n The safety score of the nth sample;

[0065] S7.3: Based on crop safety, pesticide use safety, and comprehensive compliance score, generate a planting standard assessment report. The planting standard assessment report includes a crop safety compliance report, a pesticide use safety compliance report, a comprehensive compliance score compliance report, a risk warning report, and a traceability query report.

[0066] S7.4: Package the crop safety level, pesticide use safety level, comprehensive compliance score and planting standard assessment report to obtain the planting standard compliance assessment report;

[0067] Furthermore, S1: Collect multispectral sensor data based on sensor networks to obtain a multispectral dataset;

[0068] S2: Perform feature extraction and preprocessing on the multispectral dataset to obtain a standardized dataset;

[0069] S3: Analyze the hazard factors based on standardized datasets to obtain crop hazard reports;

[0070] S4: Analyze crop damage reports to obtain crop spraying strategy reports;

[0071] S5: Convert the crop spraying strategy report into executable spraying instructions and transmit them to obtain a spraying operation record report;

[0072] S6: Analyze historical crop damage reports and pesticide spraying operation records to obtain a pesticide residue prediction report;

[0073] S7: Assess the standardization of the planting process and provide corresponding improvement suggestions to obtain a planting standard compliance assessment report.

[0074] The technical effects and advantages of the field-targeted spraying device for Hypericum perforatum based on multispectral recognition of the present invention are as follows:

[0075] This invention acquires multispectral sensor data via a sensor network to obtain a multispectral dataset. Feature extraction and preprocessing of the multispectral dataset yield a standardized dataset. Analysis of hazard factors based on the standardized dataset generates a crop hazard report. Further analysis of the crop hazard report generates a crop spraying strategy report. This report is then converted into executable spraying instructions and transmitted, resulting in a spraying operation record report. Analysis of historical crop hazard reports and spraying operation record reports generates a pesticide residue prediction report. The standardization of the planting process is assessed, and corresponding improvement suggestions are provided, resulting in a planting standard compliance assessment report. This allows the system to achieve significantly more precise information acquisition and processing than traditional devices through the collaborative operation of the multispectral data acquisition module, data extraction and preprocessing module, and crop hazard identification and extraction module. Consequently, the device can precisely distinguish different hazard levels and types of Hypericum perforatum. This invention provides solid data support for subsequent decision-making by the device. Furthermore, through the establishment of a fixed-point spraying decision optimization module and a spraying instruction generation and transmission module, the device can dynamically optimize calculations based on multiple factors and transform them into precise, executable instructions. This minimizes pesticide usage while ensuring effective control, reducing planting costs and environmental pressure, and effectively mitigating the risk of pesticide residues caused by excessive pesticide use. Finally, through a pesticide residue prediction and traceability module and a planting standard compliance assessment module, planting management and product safety are deeply linked. This allows the device to not only predict pesticide residue risks at harvest but also automatically compare with standard planting specifications to ensure standardized planting, achieving full life-cycle management from crop planting to harvest. Overall, this invention has significant advantages such as high reliability of basic data, strong intelligent dynamic and precise control capabilities, and excellent crop life-cycle management effects. Attached Figure Description

[0076] Figure 1 This is a schematic diagram of a field-targeted spraying device for Hypericum perforatum based on multispectral recognition according to the present invention.

[0077] Figure 2 This is a schematic diagram of a field-targeted spraying method for Hypericum perforatum based on multispectral recognition according to the present invention. Detailed Implementation

[0078] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0079] The terminology used in the embodiments of this invention is for the purpose of describing particular embodiments only and is not intended to limit the invention. The singular forms “a,” “the,” and “the” used in the embodiments of this invention are also intended to include the plural forms, and “multiple” generally includes at least two unless the context clearly indicates otherwise.

[0080] Depending on the context, the words “if” or “suppose” as used here can be interpreted as “when” or “in response to determination” or “in response to detection.” Similarly, depending on the context, the phrases “if determination” or “if detection (of the stated condition or event)” can be interpreted as “when determination” or “in response to determination” or “when detection (of the stated condition or event)” or “in response to detection (of the stated condition or event).”

[0081] Furthermore, the timing of the steps in the following method embodiments is merely an example and not a strict limitation.

[0082] In practice, the server-side equipment deployed for the multispectral recognition-based targeted spraying device for Hypericum perforatum may consist of one or more devices. This multispectral recognition-based targeted spraying device for Hypericum perforatum can be implemented as a business instance, a virtual machine, or hardware device. For example, it can be implemented as a business instance deployed on one or more devices in a cloud node. Simply put, it can be understood as software deployed on a cloud node, providing multispectral recognition-based targeted spraying services to various user terminals. Alternatively, it can also be implemented as a virtual machine deployed on one or more devices in a cloud node, with application software installed to manage each user terminal. Alternatively, the multispectral recognition-based field-targeted spraying device for Hypericum perforatum can also be implemented as a server consisting of numerous identical or different types of hardware devices, with one or more hardware devices set up to provide the multispectral recognition-based field-targeted spraying device for Hypericum perforatum to each user terminal.

[0083] In terms of implementation, the multispectral recognition-based field-targeted spraying device for Hypericum perforatum and the user terminal are mutually compatible. Specifically, if the multispectral recognition-based field-targeted spraying device for Hypericum perforatum is implemented as an application installed on a cloud service platform, the user terminal acts as a client establishing a communication connection with that application; or if the multispectral recognition-based field-targeted spraying device for Hypericum perforatum is implemented as a website, the user terminal acts as a webpage; or if the multispectral recognition-based field-targeted spraying device for Hypericum perforatum is implemented as a cloud service platform, the user terminal acts as a mini-program within an instant messaging application.

[0084] like Figure 1 The figure shown is a system architecture diagram of a field-targeted spraying device for Hypericum perforatum based on multispectral recognition provided in an embodiment of the present invention.

[0085] The multispectral recognition-based targeted spraying device for Hypericum perforatum described in this invention can be installed on a cloud server. In terms of implementation, it can be used as one or more service devices, or as an application installed on the cloud (e.g., a mobile service operator's server, server cluster, etc.), or it can be developed into a website. Depending on the functions implemented, the multispectral recognition-based targeted spraying device for Hypericum perforatum may include a multispectral data acquisition module, a data extraction and preprocessing module, a crop hazard identification and extraction module, a targeted spraying decision optimization module, a spraying instruction generation and transmission module, a pesticide residue prediction and traceability module, and a planting standard compliance assessment module. The modules described in this invention can also be called units, referring to a series of computer program segments that can be executed by an electronic device's processor and perform a fixed function, stored in the electronic device's memory.

[0086] In this embodiment of the invention, in the multispectral recognition-based field-targeted spraying device for Hypericum perforatum, each of the above-mentioned modules can be implemented independently and called upon other modules. Here, "called upon" can be understood as a module connecting to multiple modules of another type and providing corresponding services to those connected modules. For example, the sharing and evaluation module can call the same information acquisition module to obtain information collected by that module. Based on the above characteristics, in the multispectral recognition-based field-targeted spraying device for Hypericum perforatum provided in this embodiment of the invention, without modifying the program code, the applicable scope of the architecture of the multispectral recognition-based field-targeted spraying device for Hypericum perforatum can be adjusted by adding modules and directly calling them, achieving cluster-style horizontal expansion to quickly and flexibly expand the multispectral recognition-based field-targeted spraying device for Hypericum perforatum. In practical applications, the above-mentioned modules can be set in the same device or different devices, or they can be set in a virtual device, such as a service instance in a cloud server.

[0087] Example 1

[0088] Please see Figure 1 As shown in this embodiment, a multispectral recognition-based field targeted spraying device for Hypericum perforatum includes:

[0089] The multispectral data acquisition module is used to acquire multispectral sensor data based on a sensor network to obtain a multispectral dataset;

[0090] Furthermore, the steps for acquiring multispectral sensor data based on sensor networks include:

[0091] S1.1: Based on the preset acquisition time, generate a multispectral data acquisition command and output the multispectral acquisition command to the sensor network for data acquisition to obtain the original spectral dataset;

[0092] It should be explained that the preset acquisition time refers to a fixed data acquisition time point, such as the preset acquisition time being 9:00 AM every day; multispectral data acquisition instructions include, but are not limited to, the target area acquisition range, device acquisition parameters, and data acquisition resolution parameters, etc.

[0093] S1.2: Receive the raw spectral dataset transmitted back from the sensor network based on the standard API interface, and perform integrity verification on the raw spectral dataset. Integrity verification is to calculate the checksum value of the raw spectral dataset and compare the calculated checksum value with the transmitted checksum value. If the comparison result is consistent, proceed to step S1.3. If the comparison result is inconsistent, request to retransmit the raw spectral dataset.

[0094] S1.3: Based on the spatiotemporal reference, the data items from different times and different sensor sources in the original spectral dataset are unified to obtain the spectral aligned dataset;

[0095] It should be explained that unification refers to converting the acquisition time and location coordinates of all data items in the original spectral dataset to a preset coordinate system and preset time standard.

[0096] S1.4: Integrate the spectral alignment dataset with the aligned timestamps to obtain a multispectral dataset;

[0097] S1.5: Output the multispectral dataset to the data extraction and preprocessing module;

[0098] The data extraction and preprocessing module is used to extract and preprocess features from the multispectral dataset to obtain a standardized dataset.

[0099] Further steps for feature extraction and preprocessing of the multispectral dataset include:

[0100] S2.1: Perform surface reflectance correction on the multispectral dataset to obtain the corrected dataset. The specific calculation formula for the correction is as follows:

[0101]

[0102] Surface reflectance A is obtained a , where A b For the radiance collected by the sensor network in the multispectral dataset, A c For path radiation, A d θ is the solar irradiance, and θ is the solar zenith angle.

[0103] S2.2: Based on the calibration dataset, the spectral characteristics of vegetation are calculated to obtain a feature dataset, which includes vegetation features, red edge features, and soil-modified vegetation features. The specific formula set for calculation is as follows:

[0104]

[0105] Vegetation characteristics B were obtained respectively. a Red-edge feature B b and soil-regulated vegetation characteristics B c Among them, B d B represents the surface reflectance of vegetation in the near-infrared band. e B represents the surface reflectance of vegetation in the red light band. f κ represents the surface reflectance of vegetation in the red-edge band, and κ is the soil adjustment coefficient.

[0106] S2.3: Calculate the feature dataset using the Z-score standardization formula to obtain the standardized dataset;

[0107] S2.4: Output the standardized dataset to the crop hazard identification and extraction module;

[0108] The crop hazard identification and extraction module is used to analyze hazard factors based on a standardized dataset and obtain a crop hazard report;

[0109] Furthermore, the steps for analyzing hazard factors based on standardized datasets include:

[0110] S3.1: Retrieve a pre-trained hazard classification model from the database, input the standardized dataset into the hazard classification model, and output the hazard probability dataset;

[0111] It should be explained that the hazard probability dataset includes the hazard probability of all data items in the standardized dataset, and the hazard probability includes the probability of health, the probability of mild hazard, and the probability of severe hazard.

[0112] S3.2: Retrieve the hazard assessment decision rule table from the database and compare it with the hazard probability dataset to obtain a hazard level report;

[0113] It should be explained that the hazard assessment decision rule table is manually set and input into the system. For example, when the hazard probability is the same as the health probability, the data item is marked as a health level; when the hazard probability is a mild hazard probability and the mild hazard probability value is greater than or equal to 0.6, the data item is marked as a mild hazard level; when the hazard probability is a severe hazard probability and the severe hazard probability value is greater than or equal to 0.5, the data item is marked as a severe hazard level. The hazard level report is a specific value obtained by assigning a value based on the hazard level. For example, when the hazard level is healthy, the hazard level report is 1.

[0114] S3.3: Package the hazard level report, the coordinate data of all data items in the hazard level report, and the comparison timestamp to obtain the crop hazard report;

[0115] S3.4: Output crop damage reports to the targeted spraying decision optimization module;

[0116] The fixed-point spraying decision optimization module is used to analyze crop damage reports and obtain crop spraying strategy reports.

[0117] Further steps in analyzing crop damage reports include:

[0118] S4.1: Based on the DBSCAN density spatial clustering algorithm, the connected hazard areas in the crop hazard report are identified, and the connected hazard areas with an area less than or equal to the preset area are filtered out to obtain the hazard area.

[0119] It should be explained that the preset area is manually set and entered into the system, for example, the preset area is 2 square meters;

[0120] S4.2: The priority assessment value is calculated based on the crop damage report and the area of ​​the damaged area. The specific calculation method is to obtain the priority assessment value by weighted summation of the damage level report, the area of ​​the damaged area and the distance from the healthy area.

[0121] It should be explained that the distance from the healthy area refers to the straight-line distance from the edge of the hazardous area to the edge of the nearest healthy area larger than the preset area;

[0122] S4.3: Calculate the spraying amount based on the area of ​​the affected area to obtain the spraying amount data. The specific formula for calculation is as follows:

[0123] C a =C b ×(1+γ1×C c +γ2×C d )×C e ;

[0124] Obtain spraying amount data C a , where Cb For hazard level reporting, γ1 and γ2 are weighting coefficients, C c C represents the distance from the healthy area. d C represents canopy coverage. e The area of ​​the affected region;

[0125] It should be explained that canopy coverage refers to the proportion of the area of ​​the Hypericum perforatum canopy projected vertically to the ground to the total planted land area, where the canopy includes, but is not limited to, the leaves and stems of Hypericum perforatum.

[0126] S4.4: Package the boundary coordinates of the affected area, priority assessment values, and pesticide application data to obtain a crop spraying strategy report;

[0127] S4.5: Output the crop spraying strategy report to the spraying instruction generation and transmission module;

[0128] The spraying instruction generation and transmission module is used to convert the crop spraying strategy report into executable spraying instructions and transmit them to obtain a spraying operation record report.

[0129] Furthermore, the steps of converting crop spraying strategy reports into executable spraying instructions and transmitting them include:

[0130] S5.1: Based on the crop spraying strategy report, generate the first spraying instruction set, which includes spraying operation area boundary instructions and spraying quantity instructions;

[0131] S5.2: Calculate the optimal spraying path based on the crop spraying strategy report to obtain the optimal spraying operation path;

[0132] S5.3: Combine the boundary instructions for the spraying operation area, the spraying quantity instructions, and the optimal spraying operation path to obtain a comprehensive spraying instruction set;

[0133] S5.4: Output the integrated spraying command set to the intelligent controller, monitor the intelligent controller's reception results, and obtain a command reception report, which includes reception success and reception failure.

[0134] S5.5: Package the integrated spraying instruction set and instruction reception report to obtain the spraying operation record report;

[0135] S5.6: Store the spraying operation record report in the database and output it to the pesticide residue prediction and traceability module;

[0136] The pesticide residue prediction and tracing module is used to analyze historical crop damage reports and spraying operation record reports to obtain pesticide residue prediction reports.

[0137] Further steps in analyzing historical crop damage reports and pesticide application records include:

[0138] S6.1: Based on the meteorological release API interface, collect meteorological data of the planting area to obtain a meteorological dataset, which includes real-time temperature data and real-time rainfall data;

[0139] S6.2: Based on the pre-trained pesticide residue prediction model retrieved from the database, historical crop damage reports, historical spraying operation records, and historical meteorological datasets are input into the pesticide residue prediction model, and the predicted pesticide residue data is output. The specific formula for prediction is as follows:

[0140] D a，t =D b ×e -k×t ;

[0141] Obtain the predicted pesticide residue data D at time point t a，t , where D b is the initial pesticide residue, and k is the degradation rate coefficient;

[0142] It should be explained that the initial pesticide residue level was calculated based on historical crop damage reports and historical spraying operation records; the degradation rate coefficient was calculated based on real-time temperature data and real-time rainfall data collected from meteorological datasets.

[0143] S6.3: Based on the predicted pesticide residue data, calculate the predicted pesticide residue data at the harvest time point and divide it by the maximum allowable residue level to obtain the pesticide residue risk value;

[0144] S6.4: Based on the residue risk threshold range (Q1, Q2), when the pesticide residue risk value is less than Q1, a low-risk label is generated; when the pesticide residue risk value is greater than or equal to Q1 and less than Q2, a medium-risk label is generated; when the pesticide residue risk value is greater than or equal to Q2, a high-risk label is generated.

[0145] Based on risk labels, generate a risk level distribution map;

[0146] S6.5: Integrate the predicted pesticide residue data into the risk level distribution map to obtain a pesticide residue prediction report;

[0147] S6.6: Output the pesticide residue prediction report to the planting standard compliance assessment module;

[0148] The planting standard compliance assessment module is used to assess the standardization of the planting process, provide corresponding improvement suggestions, and generate a planting standard compliance assessment report.

[0149] Furthermore, the steps provided for assessing the standardization of the planting process and offering corresponding improvement suggestions include:

[0150] S7.1: Calculate the ratio of low-risk pesticide residues to total risk pesticide residues based on pesticide residue prediction reports, and multiply by 100% to obtain crop safety.

[0151] The number of times pesticides were sprayed within a specified area is retrieved from the database and compared with the standard number of times pesticides were sprayed. When the number of times pesticides were sprayed was less than or equal to the standard number of times pesticides were sprayed, it was recorded as compliant spraying.

[0152] It should be explained that the standard number of sprays is manually set and entered into the system, for example, the standard number of sprays is three.

[0153] Calculate the ratio of compliant spraying times to the total number of spraying times, and multiply by 100% to obtain the pesticide use safety level;

[0154] It should be explained that the total number of spraying operations refers to the total number of compliant or non-compliant spraying operations.

[0155] S7.2: A comprehensive compliance score is obtained by comprehensively calculating the crop safety and pesticide use safety. The specific formula for the calculation is as follows:

[0156]

[0157] The overall compliance score was D. c Where N is the number of samples, ω n E represents the weight coefficient of the nth sample. n The safety score of the nth sample;

[0158] S7.3: Based on crop safety, pesticide use safety, and comprehensive compliance score, generate a planting standard assessment report. The planting standard assessment report includes a crop safety compliance report, a pesticide use safety compliance report, a comprehensive compliance score compliance report, a risk warning report, and a traceability query report.

[0159] It should be explained that the crop safety compliance report, pesticide use safety compliance report, comprehensive compliance score compliance report, and risk warning report are improvement suggestions generated based on the comparison of crop safety, pesticide use safety, and comprehensive compliance score with the corresponding compliance thresholds.

[0160] S7.4: Package the crop safety level, pesticide use safety level, comprehensive compliance score and planting standard assessment report to obtain the planting standard compliance assessment report;

[0161] In this embodiment, the beneficial effects are achieved by acquiring multispectral sensor data based on a sensor network to obtain a multispectral dataset. Feature extraction and preprocessing of the multispectral dataset yield a standardized dataset. Analysis of hazard factors based on the standardized dataset generates a crop hazard report. Further analysis of the crop hazard report yields a crop spraying strategy report. This report is then converted into executable spraying instructions and transmitted, resulting in a spraying operation record report. Analysis of historical crop hazard reports and spraying operation record reports generates a pesticide residue prediction report. The standardization of the planting process is assessed, and corresponding improvement suggestions are provided, resulting in a planting standard compliance assessment report. This allows the system to achieve significantly more precise information acquisition and processing than traditional devices through the collaborative operation of the multispectral data acquisition module, data extraction and preprocessing module, and crop hazard identification and extraction module. Consequently, the device can precisely distinguish different hazards associated with Hypericum perforatum. The invention provides solid data support for subsequent decision-making by classifying and classifying pesticide levels and types. Furthermore, through the establishment of a fixed-point spraying decision optimization module and a spraying instruction generation and transmission module, the invention enables the device to dynamically optimize calculations based on multiple factors and transform them into precise, executable instructions. This minimizes pesticide usage while ensuring effective control, reducing planting costs and environmental pressure, and effectively mitigating the risk of pesticide residues caused by excessive pesticide use. Finally, through a pesticide residue prediction and traceability module and a planting standard compliance assessment module, planting management and product safety are deeply linked. This allows the device to not only predict pesticide residue risks at harvest but also automatically compare with standard planting specifications to ensure standardized planting, achieving full life-cycle management from crop planting to harvest. Overall, this invention has significant advantages such as high reliability of basic data, strong intelligent dynamic and precise control capabilities, and excellent crop life-cycle management effects.

[0162] Example 2

[0163] Please see Figure 2 As shown, for parts not described in detail in this embodiment, please refer to the description in Embodiment 1. A field-targeted spraying method for Hypericum perforatum based on multispectral recognition is provided, the method comprising:

[0164] S1: Collect multispectral sensor data based on sensor network to obtain multispectral dataset;

[0165] S2: Perform feature extraction and preprocessing on the multispectral dataset to obtain a standardized dataset;

[0166] S3: Analyze the hazard factors based on standardized datasets to obtain crop hazard reports;

[0167] S4: Analyze crop damage reports to obtain crop spraying strategy reports;

[0168] S5: Convert the crop spraying strategy report into executable spraying instructions and transmit them to obtain a spraying operation record report;

[0169] S6: Analyze historical crop damage reports and pesticide spraying operation records to obtain a pesticide residue prediction report;

[0170] S7: Assess the standardization of the planting process and provide corresponding improvement suggestions to obtain a planting standard compliance assessment report.

[0171] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the present invention.

Claims

1. A hypericum perforatum field targeted spraying device based on multispectral recognition, characterized in that, The device comprises a crop damage identification extraction module, a fixed-point pesticide spraying decision optimization module, a pesticide spraying instruction generation and transmission module, a pesticide residue prediction and traceability module, and a planting standard compliance evaluation module, wherein: The crop damage identification extraction module is configured to analyze hazard factors based on a standardized data set to obtain a crop damage report; The fixed-point pesticide spraying decision optimization module is configured to analyze the crop damage report to obtain a crop pesticide spraying strategy report; The pesticide spraying instruction generation and transmission module is configured to convert the crop pesticide spraying strategy report into executable pesticide spraying instructions and transmit the same to obtain a pesticide spraying operation record report; The pesticide residue prediction and traceability module is configured to analyze historical crop damage reports and pesticide spraying operation record reports to obtain a pesticide residue prediction report; The planting standard compliance evaluation module is configured to evaluate the standard degree of the planting process and provide corresponding improvement suggestions to obtain a planting standard compliance evaluation report.

2. The device according to claim 1, characterized in that, The system further comprises a multispectral data acquisition module and a data extraction and preprocessing module, wherein: The multispectral data acquisition module is configured to acquire multispectral sensor data based on a sensor network to obtain a multispectral data set; The data extraction and preprocessing module is configured to extract features and preprocess the multispectral data set to obtain a standardized data set.

3. The hypericum patulum target spraying device based on multispectral recognition according to claim 2, characterized in that, The step of acquiring multispectral sensor data based on a sensor network comprises: S1.1: Based on a preset acquisition time, generate a multispectral data acquisition instruction, and output the multispectral acquisition instruction to the sensor network for data acquisition to obtain an original spectral data set; S1.2: Receive the original spectral data set returned by the sensor network based on a standard API interface, and perform integrity checking on the original spectral data set. The integrity checking is to calculate a checksum value of the original spectral data set, and the calculated checksum value is compared with the transmitted checksum value. When the comparison result is consistent, proceed to step S1.

3. When the comparison result is inconsistent, request to retransmit the original spectral data set; S1.3: Based on a space-time reference, unify the data items of different times and different sensors in the original spectral data set to obtain a spectral alignment data set; S1.4: Integrate the spectral alignment data set with an alignment timestamp to obtain a multispectral data set; S1.5: Output the multispectral data set to the data extraction and preprocessing module.

4. The hypericum field targeting spraying device based on multispectral recognition according to claim 2, characterized in that, The step of extracting features and preprocessing the multispectral data set comprises: S2.1: Perform ground reflectance correction on the multispectral data set to obtain a corrected data set; S2.2: Based on the corrected data set, calculate the spectral features of vegetation to obtain a feature data set, which includes vegetation features, red edge features, and soil-adjusted vegetation features; S2.3: Calculate the feature data set using a Z-score standardization formula to obtain a standardized data set; S2.4: Output the standardized data set to the crop damage identification extraction module.

5. The hypericum field targeting spraying device based on multispectral recognition according to claim 1, characterized in that, The step of analyzing hazard factors based on the standardized data set comprises: S3.1: Obtain a pre-trained hazard classification model based on a database, input the standardized data set into the hazard classification model, and output a hazard probability data set; S3.2: Obtain a hazard determination decision rule table based on a database, and compare it with the hazard probability data set to obtain a hazard level report; S3.3: Package the hazard level report, the coordinates of all data items in the hazard level report, and the comparison timestamp to obtain a crop hazard report; S3.4: Output the crop hazard report to the fixed-point pesticide spraying decision optimization module.

6. The hypericum field targeting spraying device based on multispectral recognition according to claim 1, characterized in that, The steps of analyzing based on the crop hazard report include: S4.1: Identify connected hazard areas in the crop hazard report based on the DBSCAN density space clustering algorithm, and filter the identified connected hazard areas with an area less than or equal to a preset area to obtain a hazard area; S4.2: Calculate the priority evaluation value based on the crop hazard report and the hazard area, and the specific way of calculation is to obtain the priority evaluation value by weighted summation of the hazard level report, the hazard area, and the distance from the healthy area; S4.3: Calculate the pesticide spraying amount based on the hazard area to obtain the pesticide spraying amount data; S4.4: Package the boundary coordinate data of the hazard area, the priority evaluation value, and the pesticide spraying amount data to obtain a crop pesticide spraying strategy report; S4.5: Output the crop pesticide spraying strategy report to the pesticide spraying instruction generation and transmission module.

7. The hypericum field targeting spraying device based on multispectral recognition according to claim 1, characterized in that, The steps of converting the crop pesticide spraying strategy report into executable pesticide spraying instructions and transmitting include: S5.1: Based on the crop pesticide spraying strategy report, generate a first pesticide spraying instruction set, which includes pesticide spraying operation area boundary instructions and pesticide spraying amount instructions; S5.2: Calculate the optimal pesticide spraying path based on the crop pesticide spraying strategy report to obtain the optimal pesticide spraying operation path; S5.3: Package the pesticide spraying operation area boundary instructions, the pesticide spraying amount instructions, and the optimal pesticide spraying operation path to obtain a comprehensive pesticide spraying instruction set; S5.4: Output the comprehensive pesticide spraying instruction set to the intelligent controller and monitor the receiving result of the intelligent controller to obtain an instruction receiving report, which includes receiving success and receiving failure; S5.5: Package the comprehensive pesticide spraying instruction set and the instruction receiving report to obtain a pesticide spraying operation record report; S5.6: Store the pesticide spraying operation record report in the database and output it to the pesticide residue prediction and traceability module.

8. The hypericum field targeting spraying device based on multispectral recognition according to claim 1, characterized in that, The steps of analyzing based on the historical crop hazard report and the pesticide spraying operation record report include: S6.1: Based on the meteorological release API interface, collect the meteorological data of the planting area to obtain a meteorological data set, which includes real-time temperature data and real-time rainfall data; S6.2: Obtain a pre-trained pesticide residue prediction model based on a database, input the historical crop hazard report, the historical pesticide spraying operation record report, and the historical meteorological data set into the pesticide residue prediction model, and output the predicted pesticide residue data; S6.3: Based on the predicted pesticide residue data, calculate the predicted pesticide residue data at the harvesting time point divided by the maximum allowable residue to obtain a pesticide residue risk value; S6.4: Based on the residual risk threshold interval (Q1, Q2), when the pesticide residual risk value is less than Q1, a low-risk label is generated, when the pesticide residual risk value is greater than or equal to Q1 and less than Q2, a medium-risk label is generated, and when the pesticide residual risk value is greater than or equal to Q2, a high-risk label is generated; Based on the risk label, a risk level distribution map is generated; S6.5: Integrate the predicted pesticide residue data into the risk level distribution map to obtain a pesticide residue prediction report; S6.6: Output the pesticide residue prediction report to the planting standard compliance evaluation module.

9. The hypericum field targeting spraying device based on multispectral recognition according to claim 1, characterized in that, The steps for evaluating the standard degree of the planting process and providing corresponding improvement suggestions include: S7.1: Calculate the ratio of the number of pesticide residue low-risk to the total number of residual risks based on the pesticide residue prediction report, and multiply by 100% to obtain the crop safety degree; Based on the database, the number of pesticide spraying times in the specified area is retrieved and compared with the standard pesticide spraying times. When the number of pesticide spraying times is less than or equal to the standard pesticide spraying times, it is recorded as compliant pesticide spraying; Calculate the ratio of the number of compliant pesticide spraying times to the total number of pesticide spraying times and multiply by 100% to obtain the pesticide use safety degree; S7.2: Based on the crop safety degree and the pesticide use safety degree, a comprehensive compliance standard score is calculated; S7.3: Based on the crop safety degree, the pesticide use safety degree, and the comprehensive compliance standard score, a planting standard evaluation report is generated, which includes a crop safety degree standard compliance report, a pesticide use safety degree standard compliance report, a comprehensive compliance standard score standard compliance report, a risk prompt report, and a traceability query report; S7.4: Package the crop safety degree, the pesticide use safety degree, the comprehensive compliance standard score, and the planting standard evaluation report to obtain a planting standard compliance evaluation report.

10. A method for field targeted spraying of Hypericum perforatum based on multispectral recognition, implemented by a device for field targeted spraying of Hypericum perforatum based on multispectral recognition according to any one of claims 1-9, characterized in that, The steps include: S1: Collect multi-spectral sensor data based on a sensor network to obtain a multi-spectral data set; S2: Feature extraction and preprocessing are performed on the multi-spectral data set to obtain a standardized data set; S3: Based on the standardized data set, the hazard factors are analyzed to obtain a crop damage report; S4: Based on the crop damage report, an analysis is performed to obtain a crop pesticide spraying strategy report; S5: The crop pesticide spraying strategy report is converted into executable pesticide spraying instructions and transmitted to obtain a pesticide spraying operation record report; S6: Based on the historical crop damage report and the pesticide spraying operation record report, an analysis is performed to obtain a pesticide residue prediction report; S7: Evaluate the standard degree of the planting process and provide corresponding improvement suggestions to obtain a planting standard compliance evaluation report.

Images