Crop disease and insect pest detection method

Abstract

The invention relates to the technical field of crop disease and insect pest detection, in particular to a disease and insect pest monitoring method based on combined detection of an insect pest trap and a microscope. A multi-mode trapping mechanism and an automatic sample collection process are adopted, and the efficiency and precision of disease and pest detection can be remarkably improved. The multi-mode microscope is combined with a fluorescence module and a polarization module to comprehensively analyze insect body structural features, identify recessive pathogen infection and judge a development stage, and is matched with a convolutional neural network (CNN) model to automatically identify insect species, so that the analysis efficiency and accuracy are remarkably improved. A risk index model is arranged, the insect pest risk can be quantified, an insect pest thermodynamic diagram is generated in combination with a farmland GIS map, and early warning is given out 1-3 days in advance. A risk assessment result directly drives generation of a prevention and treatment prescription map, parameters of biological prevention and treatment, chemical prevention and treatment and physical prevention and treatment are accurately matched, and prevention and treatment resource allocation is optimized.

Description

Technical Field

[0001] The present invention relates to the technical field of crop pest and disease detection, and particularly to a pest and disease monitoring method based on the combined detection of a pest trap and a microscope. Background Art

[0002] Crop pests and diseases are important problems in agricultural production. Traditional pest and disease detection methods include manual observation, sticky board monitoring, pest attractants, etc. However, the existing technologies have the following problems: 1. Low detection efficiency: The coverage of manual inspections is limited, and it is difficult to detect early pests in a timely manner. 2. Insufficient accuracy: Traditional traps have limited ability to identify pest species and development stages and cannot provide detailed data. 3. Data lag: Samples need to be sent to the laboratory for analysis after collection, resulting in a lag in prevention and control decisions. 4. Lack of dynamic monitoring: Existing methods cannot monitor the spread path and risk changes of pests in real time. 5. Strong subjectivity in prevention and control decisions: Traditional prevention and control measures mostly rely on empirical judgments and lack quantitative basis.

[0003] With the development of modern agricultural technologies, there is an urgent need for an efficient, accurate, and quantifiable pest and disease detection method to detect and control pests and diseases in a timely manner and reduce agricultural production losses. Summary of the Invention

[0004] The purpose of the present invention is to provide a pest and disease monitoring method based on the combined detection of a pest trap and a microscope, which can achieve efficient and accurate detection of pests and diseases by capturing pests with the trap and conducting detailed analysis in combination with the microscope.

[0005] To solve the above technical problems, the technical solutions provided by the present invention are as follows:

[0006] A crop pest and disease detection method includes the following steps:

[0007] S1. Arrange multi-chamber pest traps in a checkerboard pattern in the farmland. The trap is internally provided with sex pheromones, food attractants, and physical barrier areas;

[0008] S2. After the trap collects samples, fix them with an ethanol slow-release sponge and attach double identifications of RFID and two-dimensional code;

[0009] S3. Place the samples under a multi-modal microscope for analysis, and automatically identify the pest species and development stages in combination with an AI model;

[0010] S4. Integrate the pest data with the farmland GIS map, construct a risk index model, and generate a prevention and control prescription map.

[0011] As an improvement, the trap is connected to the Internet of Things using a GPS positioning module to achieve remote monitoring and data transmission.

[0012] As an improvement, the microscope is equipped with a fluorescence and polarization module, and the dynamic adjustment range of the magnification is 100 - 600.

[0013] As an improvement, the parameters for calculation by the risk index model include: insect population density, development synchronization rate, natural enemy control rate, and crop tolerance threshold.

[0014] As an improvement, the control prescription map includes precise parameter matching of biological control, chemical control, and physical control.

[0015] As an improvement, the top of the multi - chamber pest trap is a light - trapping area, the middle part is a sex pheromone area, and the bottom part is a physical barrier area.

[0016] As an improvement, the sex pheromone is used to release specific pheromones for target pests to trap adults; the food attractant, including fragrance agents, flavoring agents, or plant extracts, is used to attract larvae and nymphs.

[0017] As an improvement, the multi - modal microscope is equipped with a fluorescence module and a polarization module. The fluorescence module is used to detect the chitin fluorescence labeling on the insect body epidermis to identify latent pathogen infections; the polarization module is used to analyze the internal crystal structure of the eggs to determine the development stage.

[0018] As an improvement, the AI model integrates a convolutional neural network (CNN) model and pre - trains a morphological feature library of farmland pests.

[0019] As an improvement, the calculation method of the risk index of the risk index model is:

[0020]

[0021] Where N is the insect population density, S is the development synchronization rate, C is the natural enemy control rate, and T is the crop tolerance threshold.

[0022] The advantages of the present invention are as follows:

[0023] 1. The present invention adopts a multi - modal trapping mechanism and an automated sample collection process, which can significantly improve the efficiency and accuracy of pest and disease detection. The trap of the present invention adopts a hierarchical design, combines GPS positioning and Internet of Things connection, can realize remote monitoring and data back - transmission, and reduce the frequency of manual inspections; after the sample is collected, it is fixed by an ethanol - slow - release sponge and uses double identification of two - dimensional code and RFID to ensure data traceability and precise matching.

[0024] 2. The multi - modal microscope of the present invention combines fluorescence and polarization modules to comprehensively analyze the structural characteristics of the insect body, identify latent pathogen infections and determine the development stage, and cooperate with a convolutional neural network (CNN) model to automatically identify the insect species, significantly improving the analysis efficiency and accuracy.

[0025] 3. The present invention sets up a risk index model, which can quantify the pest risk, generate a pest heat map in combination with the farmland GIS map, and issue an early warning 1 - 3 days in advance. The risk assessment results directly drive the generation of the control prescription map, precisely matching the parameters of biological control, chemical control, and physical control, and optimizing the allocation of control resources. Description of the Drawings

[0026] Figure 1 It is a flowchart of the present invention. Detailed Embodiments

[0027] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Usually, the components of the embodiments of the present invention described and illustrated herein can be arranged and designed in various different configurations.

[0028] The present invention discloses a method for detecting crop pests and diseases, including the following steps:

[0029] S1. Arrange multi-chamber pest traps in a checkerboard layout in the farmland. The trap is internally provided with sex pheromones, food attractants, and physical barrier areas.

[0030] The present invention optimizes the layout of traps and the like.

[0031] Among them, the trap adopts a multi-chamber structure and is internally provided with a layered trapping area. The top of the trap is a light trapping area, the middle is a sex pheromone area, and the bottom is a physical barrier area.

[0032] The outer shell material of the trap is weather-resistant ABS plastic, and the surface is coated with an anti-ultraviolet coating to extend the service life to ≥2 years.

[0033] The bottom of the trap is provided with a GPS positioning module to transmit position data in real time, facilitating the grid management of the farmland.

[0034] The trap is internally provided with an attractant. The core attractant adopts a synergistic formula of sex pheromone plus volatile food attractant. Among them, the sex pheromone: releases specific pheromones for target pests (such as aphids and locusts, etc.) to trap adults. The food attractant: adds maltitol and plant volatiles (such as β-pinene) to attract larvae and nymphs. The release rate of the attractant is regulated by the microporous membrane controlled release technology, and the effective period is ≥30 days.

[0035] The layout strategy of the trap in the farmland is:

[0036] According to the checkerboard layout, 2 traps are arranged per mu, with a spacing of ≤30 meters to ensure uniform signal coverage. At the same time, additional protective belt traps are installed at the edge of the farmland, with a density increase of 50% to intercept migratory pests.

[0037] S2. After the traps collect samples, they are fixed with ethanol-releasing sponges and attached with dual identification of RFID and QR codes.

[0038] A timing collection device is set inside the trap, which automatically turns on at 0:00 every Wednesday to avoid damage to the samples caused by high temperatures during the day. After collection, the samples are fixed with ethanol-releasing sponges to maintain morphological integrity and prevent decay.

[0039] The collected samples are managed with QR code labeling according to the trap number, collection time, environmental temperature and humidity. At the same time, an RFID chip is introduced and embedded in the sample container to achieve full-process data traceability.

[0040] The sample collection process is automated, reducing human operation errors, ensuring data consistency, and at the same time, through the double insurance of QR code + RFID, ensuring the accurate matching of samples with the farmland location and time.

[0041] S3. Place the samples under a multi-modal microscope for analysis, and combine with an AI model to automatically identify the insect species and development stage.

[0042] The multi-modal microscope is equipped with a fluorescence module and a polarization module to identify different insect body structure characteristics. The fluorescence module is used to detect the chitin fluorescence labeling on the insect body surface to identify latent pathogen infections; the polarization module is used to analyze the internal crystal structure of the eggs to judge the development stage.

[0043] At the same time, the magnification is automatically switched to 100 - 600 times according to the size of the insect body to ensure that the microstructures are clearly visible.

[0044] The magnified images are automatically identified through an AI model. The AI model integrates a convolutional neural network (CNN) model and pre-trains a morphological feature library of farmland pests. It can automatically identify insect species, automatically classify insect species, and label the development stage (for example, eggs, larvae, pupae, adults).

[0045] At the same time, the damaged area of the insect body epidermis is calculated through an image analysis algorithm to evaluate the control effect of natural enemies.

[0046] S4. Integrate the pest data with the farmland GIS map to build a risk index model and generate a control prescription map.

[0047] S4-1. Spatiotemporal data modeling

[0048] Integrate the trap location, the number of pest species, and the development stage with the farmland GIS map to generate a pest heat map; combine meteorological data (wind direction, humidity) to predict the pest diffusion path and issue a warning 1 - 3 days in advance.

[0049] S4-2. Establish a risk quantification model for risk early warning

[0050] The risk quantification model is constructed based on the pest density, the development synchronization rate, and the crop growth period to build a risk index for risk assessment and support the pest control decision-making.

[0051] The calculation method of the risk index of the risk index model is as follows:

[0052]

[0053] Where N is the pest density, S is the development synchronization rate, C is the natural enemy control rate, and T is the crop tolerance threshold.

[0054] When RI > the threshold value, a control prescription map is automatically generated. For example, the following operations can be recommended:

[0055] Biological control: The number of released parasitic wasps = 1.5 times the number of eggs.

[0056] Chemical control: The chemical concentration is precisely matched with the spraying area.

[0057] Physical control: The light trapping power is dynamically adjusted according to the pest population density.

[0058] The present invention has multiple advantages: In terms of pest trapping, a multi-modal trapping mechanism is adopted, which can improve the capture rate of target pests and reduce the mis-capture probability of non-target insects; at the same time, remote monitoring is achieved through GPS positioning and Internet of Things connection, which can reduce the frequency of manual inspections.

[0059] In terms of pest sample collection, the sample collection is automated, which can reduce the error of manual operation and improve the accuracy of data. At the same time, a two-insurance system of QR code + RFID is set up to ensure the precise matching of the sample with the farmland location and time, and improve the traceability of data.

[0060] In terms of analysis, through multi-modal microscopic imaging, the breadth of analysis parameters can be improved, the accuracy of analysis can be enhanced, and at the same time, the efficiency of image analysis can be increased through AI assistance, and the subjective deviation of manual observation can be reduced.

[0061] In terms of generating pest control decisions, this solution can establish a spatio-temporal model and issue an early warning in advance, with good foresight. At the same time, a risk index is set to quantify the basis for decision-making and replace the traditional empirical judgment.

[0062] The above are only the preferred specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution and inventive concept of the present invention, making equivalent substitutions or changes should be covered within the protection scope of the present invention.

Claims

1. A method for detecting crop pests and diseases, characterized in that, The following steps are involved: S1. Arrange multi-chambered insect pest traps in a checkerboard pattern across the farmland. Each trap contains sex pheromones, food attractants, and physical barriers. S2. After the sample is collected by the trap, it is fixed with an ethanol slow-release sponge and dually labeled with an RFID and QR code. S3. Analyze the sample under a multimodal microscope and automatically identify the insect species and developmental stage using an AI model. S4. Integrate pest data with farmland GIS maps to construct a risk index model and generate a prevention and control prescription map.

2. The method for detecting crop pests and diseases according to claim 1, wherein The trap adopts a GPS positioning module to connect with the Internet of Things to achieve remote monitoring and data feedback.

3. The method for detecting crop pests and diseases according to claim 1, characterized in that, The microscope is equipped with fluorescence and polarization modules, and the dynamic adjustment range of magnification is 100-600.

4. The method for detecting crop pests and diseases according to claim 1, wherein, The parameters calculated by the risk index model include: insect population density, development synchronization rate, natural enemy control rate and crop tolerance threshold.

5. The method for detecting crop pests and diseases according to claim 1, wherein The control prescription map includes precise parameter matching for biological control, chemical control and physical control.

6. The method for detecting crop pests and diseases according to claim 1, wherein, The top of the multi-chamber insect pest trap is a light attraction area, the middle is a sex pheromone area, and the bottom is a physical barrier area.

7. The method for detecting crop pests and diseases according to claim 1, wherein The sex pheromone is used to release specific pheromones to target pests and trap adult insects; the food attractant, including fragrance, flavoring or plant extract, is used to attract larvae and nymphs.

8. The method for detecting crop pests and diseases according to claim 1, wherein The multimodal microscope is equipped with a fluorescence module and a polarization module. The fluorescence module is used to detect the fluorescent chitin marker on the insect skin and identify latent pathogen infection; the polarization module is used to analyze the internal crystal structure of the insect eggs and determine the development stage.

9. The method for detecting crop pests and diseases according to claim 1, characterized in that, The AI model integrates a convolutional neural network (CNN) model and pre-trains a morphological feature library of farmland pests.

10. The method for detecting crop pests and diseases according to claim 4, characterized in that, The calculation method of the risk index of the risk index model is: Among them, N is the insect population density, S is the development synchronization rate, C is the natural enemy control rate, and T is the crop tolerance threshold.

Images