Intelligent physical and chemical trapping and controlling system for cowpea leafhopper based on deep learning

Abstract

The embodiment of the application provides a kind of based on deep learning's cowpea thrips intelligent physical and chemical luring control system, the system includes double-pole collaborative architecture, the double-pole collaborative architecture includes: main rod, the main rod integrates reel type double-sided color luring plate device and thrips aggregation substance slow-release attractant, the reel type double-sided color luring plate device includes specific wavelength optimization coating;Auxiliary rod, the auxiliary rod is configured high-precision camera, electric lifting adjusting mechanism, environmental perception unit and edge computing module, the edge computing module carries lightweight insect target detection model, the camera supports macro photography, the environmental perception unit includes temperature and humidity sensor, illumination sensor and barometric pressure sensor.The system adopts double-pole collaborative architecture, main rod integrates color luring plate and food luring agent unit responsible for luring control, auxiliary rod real-time acquisition processing insect data, double-pole collaborative effectively improves prevention and control efficiency, helps green prevention and control.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent monitoring of agricultural pests, and in particular to an intelligent physico-chemical trapping and controlling system for cowpea thrips based on deep learning. BACKGROUND

[0002] In facility agriculture and large-scale cowpea planting, accurate monitoring and green prevention and control of small pests such as thrips are key links to ensure crop yield and quality. Traditional pest monitoring relies on manual observation and single trapping methods, which cannot meet the needs of modern agriculture for efficient, real-time and accurate management. The intelligent physico-chemical trapping and controlling system can realize automatic monitoring and early warning of pests by integrating physical trapping, chemical attraction and intelligent recognition technology, which is of great significance to reduce pesticide abuse, reduce labor costs and improve the safety of agricultural products.

[0003] However, the existing pest monitoring technology has multiple bottlenecks: first, the trapping method is single, the effective action radius of traditional color plates or food attractants is less than 5 meters, and it cannot dynamically respond to the differences in diurnal thrips tropism behavior, resulting in limited monitoring coverage; second, data collection and processing are lagging, manual replacement of trapping devices is time-consuming, and it is easy to miss the critical period of pest outbreak, and it relies on manual counting or traditional computer vision technology, which leads to high counting error rate due to the small size of thrips and thrips adhesion; third, the equipment is not intelligent and adaptable, and the existing systems generally lack environmental parameter sensing and autonomous adjustment capabilities, have short energy endurance and high maintenance costs, and are difficult to adapt to the complex field environment and crop monitoring needs throughout the growth cycle.

[0004] The above problems lead to delayed pest warning and poor targetedness of prevention and control measures, and there is an urgent need to develop a new trapping and controlling system with multi-modal cooperation, intelligent recognition and long-term autonomy. SUMMARY

[0005] The present application provides an intelligent physico-chemical trapping and controlling system for cowpea thrips based on deep learning, which adopts a double-bar cooperative architecture. The main bar integrates a reel-type double-sided color trap plate device and a thrips aggregation substance slow-release attractant, and undertakes the function of trapping and controlling, realizing the directional trapping and tropism regulation of pests. The auxiliary bar carries an intelligent monitoring module to collect and process images and environmental data in real time. The double-bar cooperative operation, combined with a long-term low-consumption energy system, significantly improves the prevention and control efficiency of cowpea thrips and the economic efficiency of field management, providing a reliable basis for subsequent pesticide control and helping green prevention and control of cowpea thrips.

[0006] In a first aspect, an intelligent physico-chemical trapping and controlling system for cowpea thrips based on deep learning is provided, which includes a double-bar cooperative architecture, the double-bar cooperative architecture comprising:

[0007] a main bar, the main bar integrating a reel-type double-sided color trap plate device and a thrips aggregation substance slow-release attractant, the reel-type double-sided color trap plate device comprising a specific wavelength optimized coating;

[0008] The auxiliary rod is configured with a high-precision camera, an electric lifting adjustment mechanism, an environment sensing unit, and an edge computing module, the edge computing module is loaded with a lightweight insect target detection model, the high-precision camera supports micro-distance shooting, and the environment sensing unit includes a temperature and humidity sensor, an illumination sensor, and an air pressure sensor.

[0009] It should be understood that by constructing a double-rod cooperative architecture, the main rod integrates an intelligent trapping device, and a specific wavelength coating cooperates with a food attractant to achieve dual trapping; the auxiliary rod detects the number and density of pests in real time through a high-precision camera, environment sensing, and edge computing, and adjusts adaptively to form a “monitoring-trapping” closed loop, thereby improving the precision and efficiency of the chinch bug control and reducing the cost of manual intervention.

[0010] It should be understood that the temperature and humidity sensor and the air pressure sensor integrated in the environment sensing unit provide basic data for a pest population dynamic prediction model by collecting field environment parameters (temperature, relative humidity, and atmospheric pressure) in real time, and realize quantitative analysis and prediction of the occurrence trend of the pest situation based on the influence of temperature and humidity changes on the development period and reproduction rate of the chinch bug; the illumination sensor is used to monitor the ambient light intensity and spectral distribution in real time, and when the detected illumination intensity is lower than the imaging threshold of the camera, the light supplement mechanism is automatically triggered to assist the high-precision camera in obtaining high-contrast pest situation pictures at different times, thereby avoiding pest body missed detection or identification errors caused by insufficient light and improving the detection accuracy and reliability of the image analysis module.

[0011] In combination with the first aspect, in some implementations of the first aspect, the surface coating of the reel-type double-sided color trap plate device comprises a photochromic material, which can dynamically adjust the reflectivity and polarization degree according to the ambient light intensity to match the color preference differences of the chinch bug.

[0012] It should be understood that photochromic materials are a kind of functional materials that can reversibly change color or optical properties (such as reflectivity and polarization degree) under light stimulation, and can dynamically adjust their optical parameters according to the ambient light intensity. The diurnal color preference difference of the chinch bug refers to the tendency of the chinch bug to light wavelengths that changes with the alternation of day and night, for example, the chinch bug prefers a specific high reflectivity spectrum during the day, and is sensitive to polarized light at night. Traditional fixed color plates cannot take into account this dynamic behavior.

[0013] It should be understood that the photochromic material enables the color trap plate to match the diurnal tendency difference of the chinch bug in real time: it enhances the reflectivity of specific wavelengths to attract active period chinch bugs during the day, and adjusts the polarization degree to match the light-avoiding or weak-light tendency of the chinch bug at night, thereby breaking through the limitations of traditional single-spectrum color plates and effectively improving the trapping efficiency.

[0014] In some implementations of the first aspect, the edge computing module is loaded with a lightweight insect target detection model, an input of the lightweight insect target detection model being an insect image captured by a visible light camera, and an output of the lightweight insect target detection model being a number of pests and a density level.

[0015] It should be understood that the lightweight insect target detection model is an optimized small-sized algorithm that can run locally on the edge computing module of the sub-pole, directly process the insect image captured by the visible light camera, and quickly output the number of pests and the density level without relying on a cloud server.

[0016] It should be understood that through localized real-time detection, data upload delay is avoided, and field pest monitoring response time is reduced. The lightweight design of the model reduces the hardware computing power requirement, adapts to low-power consumption scenarios, and significantly reduces the daily average energy consumption. The efficiency is significantly improved compared with manual visual inspection, and the accuracy of density grading is greatly improved, providing data support for precise prevention and control.

[0017] In some implementations of the first aspect, the double-sided color trap plate device adopts a double-layer PET substrate, and the double-sided color trap plate device is periodically replaced by a stepping motor drive to ensure monitoring accuracy.

[0018] It should be understood that the PET substrate is a thin film material made of polyethylene terephthalate (PET), which has high strength, weather resistance, chemical corrosion resistance, and good flexibility, and is often used for functional boards that need to be used outdoors for a long time. In this scheme, the double-layer PET substrate provides a lightweight and durable carrier for the color trap plate, which can withstand changes in temperature, humidity, light, and other environmental changes in the field, and can be smoothly rolled out through the reel design, avoiding the problem of easy breakage and curling of traditional paper or single-layer materials.

[0019] It should be understood that through the combination of the double-layer PET substrate and the stepping motor drive, the new sticky surface is automatically expanded, and the old sticky surface is cleaned of debris by the edge vibrator, so that the effective sticking area of the color trap plate is always maintained at a high level. The replacement frequency is significantly reduced compared with traditional manual replacement, and no manual intervention is required during the replacement cycle. At the same time, the low friction coefficient and tear resistance of the PET substrate ensure the stable operation of the reel system, significantly reducing material waste and maintenance costs, balancing the continuity of trapping and control and the reliability of the equipment, and adapting to the long-term use requirements of large-scale farmland.

[0020] In some implementations of the first aspect, the thrips attractant slow-release agent includes, for example, acetyl acetate, acetyl geranate, and (S)-2-methyl butyric acid and (R)-lavender acetic acid, (R)-lavender acetic acid, (R)-3-methyl-3-butyric acid, and (R)-lavender acetic acid), and the diffusion radius of the thrips attractant slow-release agent is up to 15 m, and the effective period is ≥60 days.

[0021] It should be understood that the diffusion unit realizes multi-dimensional attraction to the plant bugs through the composite system design, significantly expands the coverage range, effectively traps the distance much higher than single component, has long continuous release time, reduces the frequent replacement or manual operation of supplementary application. The slow-release technology avoids rapid loss of components, ensures stable and long-lasting trapping effect, and forms a three-dimensional trapping mode of "odor guidance-visual capture" combined with the color trap plate, greatly improves the trapping efficiency of adult plant bugs. At the same time, the use of chemical pesticides is reduced, the disturbance to the field ecology is reduced, and the environmental friendliness is considered, providing a long-term, low-maintenance green prevention and control scheme for cowpea planting.

[0022] In combination with the first aspect, in some implementations of the first aspect, the electric lifting adjusting mechanism adopts a worm gear transmission, and the electric lifting adjusting mechanism is equipped with a Hall sensor to realize height closed-loop control, for adapting to the change of inflorescence height during the flowering period of cowpea.

[0023] It should be understood that the electric lifting adjusting mechanism adopts a worm gear transmission, and the electric lifting adjusting mechanism is equipped with a Hall sensor to realize height closed-loop control, for adapting to the change of inflorescence height during the flowering period of cowpea.

[0024] It should be understood that the inflorescence development patterns of different cowpea varieties are different, some varieties show top-to-base flowering (inflorescences open from the upper part of the plant to the lower part in turn), and some varieties show base-to-top flowering (inflorescences open from the lower part of the plant to the upper part in turn). The system provided by the present application can dynamically adjust the equipment height according to the flowering sequence characteristics of different varieties, so that the monitoring module is matched with the inflorescence development area in real time, thereby improving the trapping efficiency of the activity track of pests and the monitoring coverage. In addition, the precise calibration of the Hall sensor controls the height adjustment error within millimeters, ensuring the consistency of the monitoring angle at different growth stages and the effective coverage of the trapping device. This design reduces the labor cost of frequent manual adjustment of the equipment, ensures the continuous and stable operation of the system during the whole growth period of the crops, and significantly improves the reliability of the monitoring data and the pertinence of the trapping strategy.

[0025] In combination with the first aspect, in some implementations of the first aspect, the system further comprises an energy system, the energy system adopts a dual-mode power supply of solar panels and lithium batteries to ensure that the system continuously works under different weather conditions.

[0026] It should be understood that through the complementary design of multiple energy sources, the system does not need to rely on external power supply, completely gets rid of the wiring restriction in the field, and is suitable for remote or large-scale planting areas. The dual-mode power supply mechanism ensures stable operation in different scenarios such as sunny days, cloudy days, and night, avoids interruption of monitoring or trapping functions caused by weather changes. This scheme takes into account the green environmental protection and reliability, provides a solid energy guarantee for long-term unattended operation of the system, and significantly enhances the adaptability of the equipment in complex field environments.

[0027] In combination with the first aspect, in some implementations of the first aspect, the reel type double-sided color lure device is provided with a tension sensor built-in, which is used to monitor the remaining length of the color lure in real time, and sends a replacement warning when the remaining length is less than a first threshold.

[0028] It should be understood that the tension sensor can accurately convert the remaining length of the color lure by sensing the change of the tensile force of the reel type color lure in real time, avoiding the lag and subjectivity of traditional manual inspection. When the remaining length is less than the first threshold, the system automatically sends a replacement warning to prompt the management personnel to maintain in time. The intelligent monitoring of the use state of the color lure is realized, the failure of the lure control due to the lack of replacement in time is avoided, and the continuity of the prevention and control effect is ensured; through early warning, passive maintenance is changed to preventive maintenance, the frequency of manual regular inspection is reduced, and the field management cost is reduced. The tension monitoring and early warning mechanism are combined with the reel type automatic deployment design to ensure that the system is always in an efficient working state within the annual replacement cycle, avoid blind spots, and improve the reliability and management efficiency of the chinch bug control. BRIEF DESCRIPTION OF DRAWINGS

[0029] Figure 1 is a structure schematic diagram of a cowpea chinch bug intelligent physical and chemical lure and control system based on deep learning provided by an embodiment of the present application.

[0030] Figure 2 is a main rod and auxiliary rod structure diagram of a cowpea chinch bug intelligent physical and chemical lure and control system based on deep learning provided by an embodiment of the present application.

[0031] Figure 3 is a main rod and auxiliary rod part top view of a cowpea chinch bug intelligent physical and chemical lure and control system based on deep learning provided by an embodiment of the present application. DETAILED DESCRIPTION

[0032] The terms used in the following embodiments are only for the purpose of describing specific embodiments and are not intended to be limiting of the present application. As used in the specification and the appended claims of the present application, the singular forms "a," "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will also be understood that "at least one" or "one or more" means one, two, three, or more. The term "and / or" is used to describe the association between the associated objects, which means that there can be three relationships; for example, A and / or B can mean that A exists alone, A and B exist together, and B exists alone, where A and B can be singular or plural. The character " / " generally represents an "or" relationship between the associated objects.

[0033] Reference in this specification to "one embodiment" or "some embodiments" means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the application. The appearances of the phrase "in one embodiment" or "in some embodiments" in various places in the specification are not necessarily all referring to the same embodiment, although it can. The terms "including," "comprising," "having" and variations thereof are meant to encompass the items listed thereafter and equivalents thereof as well as additional items.

[0034] In facility agriculture and large-scale cowpea planting, monitoring and prevention of small pests such as thrips is crucial. Traditional methods rely on manual observation and single trapping, which cannot meet the needs of efficient and accurate management. Existing technologies have multiple bottlenecks: single trapping method, small action radius, unable to respond to pest diurnal tropism, limited coverage; data collection and processing lag, long manual replacement cycle, high misjudgment rate, easy to miss the critical period of pest control; lack of device intelligence, short endurance, high maintenance cost, difficult to adapt to field environment and crop cycle monitoring. These lead to delayed early warning and poor targeted prevention and control, and there is an urgent need for a new type of trapping and control system with multi-modal collaboration, intelligent identification and long-term autonomy.

[0035] The embodiment of the application provides a cowpea thrips intelligent physical and chemical trapping and control system based on deep learning, which innovatively adopts a double-pole collaborative architecture, integrates an intelligent monitoring module and a dynamic trapping and control device, and can realize real-time pest number density detection, growth cycle dynamic analysis and prediction. Through environmental self-adaptive adjustment technology, combined with long-acting low-consumption energy system and ecological friendly trapping and control means, the efficiency of cowpea thrips prevention and control and the economy of field management are significantly improved.

[0036] The technical solutions provided by the embodiments of the application will be described below with reference to the drawings.

[0037] Figure 1 A cowpea thrips intelligent physical and chemical trapping and control system based on deep learning provided by the embodiments of the application is shown in the structural schematic diagram.

[0038] Reference Figure 2 and Figure 3 In some examples, the system includes a double-pole collaborative architecture, which includes:

[0039] A main pole integrates a reel-type double-sided color trapping plate device 1 and a thrips aggregation substance slow-release attractant 2, and the reel-type double-sided color trapping plate device 1 contains a specific wavelength optimized coating.

[0040] The auxiliary rod is configured with a high-precision camera 3, an electric lifting adjusting mechanism 4, an environment sensing unit 5, and an edge computing module 6, the edge computing module 6 is loaded with a lightweight insect target detection model, the high-precision camera 3 supports micro-distance shooting, and the environment sensing unit 5 includes a temperature and humidity sensor, an illumination sensor, and a barometric pressure sensor.

[0041] In some examples, the surface coating of the reel type double-sided color lure plate device 1 contains a photochromic material, which can dynamically adjust the reflectivity and polarization degree according to the intensity of the ambient light to match the differences in the color preference behavior of the thrips.

[0042] In a possible implementation, the photochromic material adopts a spiropyran-azobenzene composite photochromic material, which is uniformly coated on the surface of the double-layer PET substrate through a nanoscale dispersion process. Under strong light conditions, the material activates the long-wavelength high-reflection mode through molecular structure isomerization, enhancing the reflectivity of the thrips sensitive spectrum; under low light intensity, it switches to a polarization degree adjustment mode to generate polarized light at a specific angle, matching the thrips' night light-seeking behavior. The coating bottom layer integrates a titanium dioxide nano anti-fouling layer, enhancing the hydrophobicity of the adhesive surface and reducing the impact of dust and dew on the adhesion performance.

[0043] In some examples, the edge computing module 6 is loaded with a lightweight insect target detection model, the input of which is the insect image collected by the visible light camera, and the output is the number and density level of the pests.

[0044] In some examples, the reel type double-sided color lure plate device 1 adopts a double-layer PET substrate, which is replaced regularly by a stepping motor drive to ensure monitoring accuracy.

[0045] In a possible implementation, the color lure plate adopts a double-layer PET substrate as the carrier, which is made of polyethylene terephthalate and has high strength, weather resistance, and good flexibility, and can withstand changes in temperature, humidity, and light in the field and remain stable in shape. The color lure plate is integrated into the trapping and controlling device through a reel type structure and is driven by a stepping motor to automatically expand the adhesive surface. It can expand 10-15 cm of new adhesive surface, with a cumulative expansion length of ≥3.6 m, so that the effective trapping area of the adhesive plate is continuously updated without the need for frequent manual replacement. The low friction coefficient and tear resistance of the double-layer PET substrate ensure smooth operation of the reel system, reduce the risk of jamming or damage, and adapt to long-term outdoor use requirements. The mechanical automatic surface replacement design significantly reduces the frequency of manual maintenance, ensuring the continuous effectiveness of the trapping and controlling device.

[0046] In some examples, the thrips aggregating pheromone slow-release attractant 2 includes compounds such as acetyl acetate, acetyl geranate, S-2-methyl butyric acid, and R-2-methyl butyric acid, R-3-methyl-3-butenic acid, R-2-methyl butyric acid, and the diffusion radius of the thrips aggregating pheromone slow-release attractant 2 is up to 15 m, and the effective period is ≥60 days.

[0047] In some examples, the electric lifting adjusting mechanism 4 adopts a worm gear transmission, and is equipped with a Hall sensor to realize closed-loop control of height, so as to adapt to the change of the flower cluster height of the cowpea in the flowering period.

[0048] In a possible implementation, the electric lifting adjusting mechanism 4 can adopt a worm gear transmission assembly made of an aluminum alloy material, the worm gear is driven by a low-speed high-torque motor, the worm wheel is engaged with the bottom gear of the sub-rod lifting section, the transmission ratio is designed to be 1:50 to realize smooth and slow lifting, the magnetic ring type Hall sensor is integrated at the end of the worm shaft, the magnetic sensitive switch is pasted equidistantly on the inner wall of the fixed rod, and the millimeter level accurate feedback of the lifting height is realized by detecting the pulse signal when the magnetic ring rotates. In terms of control logic, the system pre-stores the flower cluster height range in different growth periods of the cowpea, when the environmental perception unit 5 detects the change of the flower cluster height caused by the growth of the plant (calculated by recognizing the flower cluster position through the camera), the edge computing module 6 sends instructions to the motor drive board, the worm gear drives the sub-rod lifting section to move up and down along the linear guide rail, the Hall sensor collects pulse numbers in real time and converts the actual height, and when the target height is reached, the worm gear and the worm wheel lock the position by virtue of the self-locking characteristic, so that the adaptive adjustment of the monitoring angle in the whole growth period is realized without manual intervention.

[0049] In some examples, the system further includes an energy system, which adopts a solar panel 7 and a power storage device 8 for dual-mode power supply, so as to ensure that the system continuously works under different weather conditions.

[0050] In some examples, the reel type double-sided color lure plate device 1 is provided with a tension sensor, which is used for monitoring the remaining length of the color lure plate in real time, and sends a replacement warning when the remaining length is less than a first threshold value.

[0051] In a possible implementation, one intelligent device is arranged at an interval of every 5 mu in the cowpea planting area, the camera of the intelligent device cooperates with the annular light supplement lamp to realize clear imaging at all times. The color lure plate reel adopts a modular architecture and is provided with a built-in tension sensor for monitoring the remaining length in real time, and when it is detected that the remaining length is less than 10%, the intelligent device automatically sends a replacement warning through a LoRaWAN wireless communication module, so as to ensure that the lure control device continuously and effectively works.

[0052] In a possible implementation, the system automatically starts the monitoring process at 06:00 every day, and the lifting mechanism automatically adjusts the equipment to the optimal monitoring position according to the preset cowpea growth period height curve; in operation, environmental data are collected in real time every 30 minutes, and image analysis is performed once every hour to evaluate the insect situation; when it is predicted through historical data and a real-time detection model that the insect population density will exceed the early warning threshold in the next 3 days, the unmanned aerial vehicle spraying system is immediately linked to implement targeted and precise pesticide application for prevention and control.

[0053] The above merely describes preferred embodiments of the present application, and the protection scope of the present application is not limited to the above-described embodiments. Any equivalent modifications or changes made by those skilled in the art according to the disclosed content of the present application shall be included in the protection scope recorded in the claims.

Claims

1. A deep learning-based intelligent physical and chemical trapping and controlling system for cowpea leafhoppers, characterized in that, The system comprises a double-pole cooperative framework, which comprises: a main pole, which integrates a reel-type double-sided color trap device (1) and a thrips aggregation pheromone slow-release attractant (2), the thrips aggregation pheromone slow-release attractant (2) comprising acetic acid farnesyl ester, acetic acid geranyl ester, opal meat base (S)-2-methyl butyric acid ester and (R)-lavender acetic acid ester, (R)-3-methyl-3-butenic acid lavender ester, (R)-2-methyl butyric acid-(R)-lavender ester, the diffusion radius of the thrips aggregation pheromone slow-release attractant (2) being up to 15 m, the effective period being ≥60 days; the reel-type double-sided color trap device (1) comprises a specific wavelength optimized coating, the surface coating of the reel-type double-sided color trap device (1) comprising a photochromic material, which can dynamically adjust reflectivity and polarization degree according to environmental light intensity, match the color preference behavior difference of thrips, the photochromic material being a spiropyran-azobenzene composite photochromic material, which is uniformly coated on the surface of a double-layer PET substrate through a nanoscale dispersion process; the reel-type double-sided color trap device (1) adopts a double-layer PET substrate, and is driven by a stepping motor to realize regular replacement, thereby ensuring monitoring accuracy; the reel-type double-sided color trap device (1) is provided with a tension sensor for real-time monitoring of the remaining length of the color trap device, and sends a replacement warning when the remaining length is less than a first threshold value; a secondary pole, which is provided with a high-precision camera (3), an electric lifting adjusting mechanism (4), an environment sensing unit (5) and an edge computing module (6), the edge computing module (6) carrying a lightweight insect target detection model, the high-precision camera (3) supporting macro photography, and the environment sensing unit (5) comprising a temperature and humidity sensor, an illumination sensor and a barometric pressure sensor; the electric lifting adjusting mechanism (4) adopts a worm and gear transmission, and is provided with a Hall sensor to realize height closed-loop control, thereby adapting to the change of the height of flower spikes in the flowering period of cowpea.

2. The system of claim 1, wherein, The edge computing module (6) carries a lightweight insect target detection model, the input of the lightweight insect target detection model being an insect image collected by a visible light camera, and the output being the number of pests and the density grade.

3. The system of claim 1, wherein, The system further comprises an energy system, which adopts a solar panel (7) and a power storage device (8) for dual-mode power supply, thereby ensuring continuous work of the system under different weather conditions.

Images