Quick-release intelligent migratory fly pest monitoring device based on edge calculation

Abstract

The embodiment of the invention provides a quick-release intelligent migratory fly pest monitoring device based on edge calculation, and the device comprises a quick-release interface which is used for achieving the quick disassembly with pest condition forecasting equipment; the heating module comprises an outer heating device and an inner heating device; the edge calculation module is used for generating a monitoring result of the migrated pests and uploading the monitoring result to a management platform, and the monitoring result comprises the types and the number of the migrated pests. According to the device, rapid butt joint with insect situation forecasting equipment is achieved through a modular rapid disassembly structure, the edge calculation module can recognize the types and the number of migrated insects and upload detection results to a management platform, the device breaks through a traditional mode of depending on large computing power of a cloud side, intelligent recognition is completed through low-power-consumption computing power of the edge side, and the efficiency is improved. And the hardware cost and the server pressure are obviously reduced.

Description

Technical Field

[0001] The present invention relates to the field of smart agriculture, and in particular to a quick-detachable intelligent migratory pest monitoring device based on edge computing. Background Art

[0002] Migratory pests refer to insects with the habit of long-distance group migration (such as fall armyworm, brown planthopper, etc.). Their migratory behavior is a survival strategy to adapt to resource distribution and climate change. This type of pest often travels hundreds to thousands of kilometers with the monsoon, and is characterized by suddenness, explosiveness, and cross-regional diffusion. Traditional high-altitude detection lamps use built-in resistance wire heating solutions, which have problems such as local overheating leading to carbonization of insect bodies and loss of morphological characteristics, seriously affecting the accuracy of subsequent classification. Existing image recognition devices mostly use fixed-focus cameras, which are difficult to ensure imaging clarity for insect bodies with large size differences. In addition, most monitoring equipment relies on 4G to transmit raw images, with a daily data volume of 1GB. In areas with weak farmland network coverage, data blockage is easy to occur, insect analysis is delayed, and the best prevention and control window is easy to be missed; and as the images uploaded to the server continue to accumulate, centralized image recognition solutions are prone to cause a large load on server storage and computing power.

[0003] Therefore, it is of great significance to provide a migratory pest monitoring device to overcome the above problems and provide decision support for the precise prevention and control of migratory pests. Summary of the invention

[0004] The present application provides a quick-detachable intelligent migratory pest monitoring device based on edge computing. The device can quickly achieve connection with insect monitoring equipment through a quick-detach module. The edge computing module can identify the types of migratory pests and count them, and upload the monitoring results to a management platform. The small computing power of the edge device is used to realize the intelligent recognition function that can only be done by the existing technology on the large computing power end, saving computing power and hardware costs.

[0005] In a first aspect, a quick-detachable intelligent migratory pest monitoring device based on edge computing is provided, the device comprising: A quick-release interface, which is used to achieve quick disassembly from the insect monitoring equipment; A heating module, the heating module is used to prevent the insect body from sticking; An edge computing module, wherein the edge computing module is used to generate monitoring results of the migratory pests and upload the monitoring results to a management platform, wherein the monitoring results include the types and quantities of the migratory pests.

[0006] It should be understood that by setting up the quick-release interface, heating module and edge computing module, the quick disassembly and intelligent monitoring functions of the insect monitoring equipment are realized. The quick-release interface directly solves the time-consuming problem of disassembly and assembly of traditional equipment, significantly improving deployment efficiency; the heating module prevents insect adhesion and maintains the integrity of the morphology through the dual design of external heating device and internal heating device, providing reliable samples for subsequent identification; the edge computing module completes the identification of pest types and quantities locally, and uploads the monitoring results directly to the management platform without relying on cloud processing, which greatly reduces the network transmission pressure and server computing power burden, while reducing hardware costs, meeting the core requirements of low power consumption and high efficiency of edge computing.

[0007] In combination with the first aspect, in some implementations of the first aspect, the quick release interface includes: An intelligent electromagnet array, which is composed of a plurality of independently controlled electromagnet units, each of which has a built-in pressure sensor for detecting the contact pressure with the inner wall of the insect monitoring device. The electromagnet array dynamically activates the electromagnet units in the corresponding area according to the shape of the inner wall. After power is turned on, it adaptively fits the monitoring devices with different inner diameters and shapes through magnetic attraction. The electromagnet array is connected to an intelligent power management module and automatically switches to a low-power maintenance mode after the adsorption is completed, thereby reducing energy consumption. The waterproof sealing layer adopts a double-layer structure, the inner layer is an elastic silicone wrapped electromagnet array, and the outer layer is a self-repairing polyurethane coating, which can automatically restore the sealing after being deformed by pressure.

[0008] It should be understood that the intelligent electromagnet array dynamically adapts to the inner wall of the measuring equipment with different inner diameters and shapes through independently controlled electromagnet units and pressure sensors, and combines the low-power maintenance mode of the intelligent power management module to ensure the stability and compatibility of the magnetic connection and reduce energy consumption. The double-layer waterproof sealing layer uses elastic silicone and self-healing polyurethane coating, which can still restore the sealing after being deformed by pressure, effectively preventing rainwater from penetrating, and ensuring the long-term reliable operation of the device in the complex environment of farmland. These technical features further enhance the rapid adaptation ability and environmental adaptability of the quick-release interface, providing hardware guarantee for the core functions.

[0009] In combination with the first aspect, in some implementations of the first aspect, the edge computing module includes: A fill light system, the fill light system comprising a fill light, the fill light system is used to reduce ghosting of insect images and improve insect recognition accuracy; A high-precision imaging unit, the high-precision imaging unit is used to collect images of the migratory pests trapped, and the high-precision imaging unit works synchronously with the fill light system; A controller having a built-in two-stage edge detection model and a communication unit, wherein the two-stage edge detection model is used to identify the categories of the migratory pests and count them, and the communication unit is used to upload the monitoring results to the management platform, and the communication unit is also used to receive instructions from the management platform.

[0010] In combination with the first aspect, in certain implementations of the first aspect, the two-stage edge detection model workflow includes: in the first stage, potential pest areas are quickly screened through low-resolution images, and in the second stage, candidate areas are classified and counted with high precision, significantly reducing the amount of calculation.

[0011] It should be understood that the fill-light system and the insect monitoring equipment's lure lights work at staggered peaks, and through the synergistic effect of multiple light sources, spectral interference that leads to a decrease in lure efficiency is avoided, while providing stable lighting conditions for the high-precision imaging unit. The high-precision imaging unit and the fill-light system synchronously collect pest images, and combined with the controller's two-stage edge detection model, first quickly screen potential pest areas in low-resolution images, and then perform high-precision classification and counting of candidate areas. This staged processing method significantly reduces the amount of calculation, allowing the edge computing module to still achieve millisecond-level recognition at low power consumption, reducing dependence on cloud servers, while improving recognition accuracy.

[0012] In combination with the first aspect, in some implementations of the first aspect, the heating module includes a dual-mode heating mode: a flexible heating pad is wrapped around the outer wall of the heating module, and an infrared heating tube is built in the falling path of the insect body to directional radiate the insect body, and the migratory pests are heated to an inactivated state and the wings of the migratory pests are directed toward the camera through variable temperature control and heating time adjustment, so as to prevent the reduction of recognition accuracy due to adhesion and feature loss of the insect body, and at the same time avoid high temperature damage to the insect body morphology; A temperature control unit is used to adaptively control the heating power of the flexible heating pad and the built-in infrared heating tube according to environmental changes.

[0013] It should be understood that the heating module is designed by combining a flexible heating pad with a built-in infrared heating tube. The external heating pad evenly controls the temperature to prevent the insect body from sticking, and the infrared heating tube directional radiates the insect body's falling path to avoid local high temperature causing carbonization of the insect body. The temperature control unit dynamically adjusts the heating power according to changes in ambient temperature and humidity. For example, it starts heating under high humidity or low temperature conditions and enters energy-saving mode at other times. This adaptive heating strategy not only effectively maintains the morphological integrity of the insect body, but also reduces energy consumption, ensuring the stable operation of the device in extreme environments.

[0014] In combination with the first aspect, in certain implementations of the first aspect, the device further includes an insect collection box, which is detachably connected to the edge computing module, and the insect collection box is used to collect the migratory pests.

[0015] In combination with the first aspect, in some implementations of the first aspect, a first opening is provided between the heating module and the edge computing module, and a second opening is provided between the edge computing module and the insect collection box.

[0016] In combination with the first aspect, in some implementations of the first aspect, the edge computing module also includes a control unit, and the control unit is used to control the opening and closing of the first opening and the second opening.

[0017] In combination with the first aspect, in some implementations of the first aspect, the device also includes a dual-mode power supply system, and the dual-mode power supply system supports 220V AC power and battery working modes.

[0018] In combination with the first aspect, in some implementations of the first aspect, the dual-mode power supply system has an adaptive power consumption management function. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 A schematic structural diagram of a quick-detachable intelligent migratory pest monitoring device based on edge computing provided in an embodiment of the present application.

[0020] Figure 2 A schematic diagram of a quick-release interface structure provided in an embodiment of the present application.

[0021] Figure 3 A schematic diagram of the structure of an edge computing module provided in an embodiment of the present application. DETAILED DESCRIPTION

[0022] The terms used in the following embodiments are only for the purpose of describing specific embodiments, and are not intended to be used as limitations on the present application. As used in the specification and the appended claims of the present application, the singular expressions "one", "a kind of", "said", "above", "the" and "this" are intended to also include expressions such as "one or more", unless there is a clear contrary indication in the context. It should also be understood that in the following embodiments of the present application, "at least one", "one or more" refer to one, two or more. The term "and / or" is used to describe the association relationship of associated objects, indicating that three relationships may exist; for example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone, where A and B can be singular or plural. The character " / " generally indicates that the objects associated before and after are in an "or" relationship.

[0023] References to "one embodiment" or "some embodiments" etc. described in this specification mean that a particular feature, structure or characteristic described in conjunction with the embodiment is included in one or more embodiments of the present application. Thus, the phrases "in one embodiment", "in some embodiments", "in some other embodiments", "in some other embodiments", etc. that appear at different places in this specification do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in other ways. The terms "including", "comprising", "having" and their variations all mean "including but not limited to", unless otherwise specifically emphasized in other ways.

[0024] Migratory pests adapt to resource distribution and climate change by migrating in groups over long distances. Their cross-regional spread with the help of monsoons often causes sudden insect disasters. Traditional high-altitude monitoring lamps use a resistance wire heating design, which is prone to carbonization of insect bodies due to local high temperatures, resulting in the destruction of morphological characteristics and reduced classification accuracy. Fixed-focus cameras are difficult to take into account the clear imaging of insect bodies with significant size differences, affecting the accuracy of intelligent recognition. The daily data volume of raw images transmitted over 4G networks reaches 1GB, which is prone to transmission delays in areas with weak farmland network coverage, causing insect situation assessment to lag behind the optimal prevention and control window. In addition, the method of centrally uploading massive image data sets to servers for processing continues to increase storage and computing loads, restricting the operating efficiency of the monitoring system.

[0025] The embodiment of the present application provides a quick-detachable intelligent migratory pest monitoring device based on edge computing, which can achieve rapid docking with insect monitoring equipment through a modular quick-detachable structure. Its edge computing module is equipped with classification recognition and counting algorithms, which can analyze the types and numbers of migratory pests in real time and upload the detection results to the management platform. The device breaks through the traditional model of relying on large computing power on the cloud, and uses low-power computing power on the edge side to complete intelligent identification, significantly reducing hardware costs and server pressure.

[0026] The technical solution of the embodiments of the present application will be described below with reference to the accompanying drawings.

[0027] Figure 1 A schematic diagram of the structure of a quick-detachable intelligent migratory pest monitoring device based on edge computing provided in an embodiment of the present application. In some examples, the device includes: A quick-release interface, which is used to achieve quick disassembly from the insect monitoring equipment; A heating module, the heating module is used to prevent the insect body from sticking; An edge computing module, wherein the edge computing module is used to generate monitoring results of the migratory pests and upload the monitoring results to a management platform, wherein the monitoring results include the types and quantities of the migratory pests.

[0028] Figure 2A schematic diagram of a quick-release interface structure provided in an embodiment of the present application, in some examples, the quick-release interface includes: An intelligent electromagnet array, which is composed of a plurality of independently controlled electromagnet units, each of which has a built-in pressure sensor for detecting the contact pressure with the inner wall of the insect monitoring device. The electromagnet array dynamically activates the electromagnet units in the corresponding area according to the shape of the inner wall. After power is turned on, it adaptively fits the monitoring devices with different inner diameters and shapes through magnetic attraction. The electromagnet array is connected to an intelligent power management module and automatically switches to a low-power maintenance mode after the adsorption is completed, thereby reducing energy consumption. The waterproof sealing layer adopts a double-layer structure, the inner layer is an elastic silicone wrapped electromagnet array, and the outer layer is a self-repairing polyurethane coating, which can automatically restore the sealing after being deformed by pressure.

[0029] Figure 3 A schematic diagram of the structure of an edge computing module provided in an embodiment of the present application, in some examples, the edge computing module includes: A fill light system, the fill light system comprising a fill light, the fill light system is used to reduce ghosting of insect images and improve insect recognition accuracy; A high-precision imaging unit, the high-precision imaging unit is used to collect images of the migratory pests trapped, and the high-precision imaging unit works synchronously with the fill light system; A controller having a built-in two-stage edge detection model and a communication unit, wherein the two-stage edge detection model is used to identify the categories of the migratory pests and count them, and the communication unit is used to upload the monitoring results to the management platform, and the communication unit is also used to receive instructions from the management platform.

[0030] In one possible implementation, the high-precision imaging unit uses a 12-megapixel high-definition camera, combined with an edge computing board equipped with a lightweight recognition model, to achieve millisecond-level recognition of insect species and quantities.

[0031] In one possible implementation, when the attracting light is turned off, the controller triggers the fill light system to start a light source of a specific wavelength to provide interference-free lighting for the imaging area; the high-precision imaging unit immediately takes continuous shots after the fill light is turned on to capture dynamic images of the insect body; the image data is transmitted to the edge computing module in real time, and the background difference method is used to remove static interference and extract effective targets; after completing the image acquisition, the fill light system automatically turns off and the attracting light restarts to continue to attract pests. This process ensures imaging clarity through strict timing control and spectral isolation, providing high-quality input data for edge computing.

[0032] In some examples, the two-stage edge detection model workflow includes: a first stage of quickly screening potential pest areas through low-resolution images, and a second stage of high-precision classification and counting of candidate areas, significantly reducing the amount of calculation.

[0033] Optionally, the two-stage edge detection model can also automatically adjust the model recognition type and detection threshold according to the pest type and pest density in different regional environments to reduce the false detection rate. Motion blur correction is performed on continuous frame images, and low-quality recognition results are filtered through confidence thresholds.

[0034] In one possible implementation, the two-stage edge detection model includes an improved YOLO model. The improvement measures include: using a lightweight backbone network, channel pruning, quantized deployment, etc. The improved YOLO model is adapted to the edge computing module and can achieve real-time processing capabilities of more than 15 frames per second.

[0035] In some examples, the heating module includes a dual-mode heating mode: a flexible heating pad is wrapped around the outer wall of the heating module, and an infrared heating tube is built in the falling path of the insect body to directional radiate the insect body, and the migratory pests are heated to an inactive state and the wings of the migratory pests are directed toward the camera through variable temperature control and heating time adjustment, so as to prevent the reduction of recognition accuracy due to adhesion and feature loss of the insect body, and at the same time avoid high temperature damage to the insect body morphology; A temperature control unit is used to adaptively control the heating power of the flexible heating pad and the built-in infrared heating tube according to environmental changes.

[0036] In some examples, the device also includes an insect collection box, which is detachably connected to the edge computing module and is used to collect the migratory pests.

[0037] In some examples, a first opening is provided between the heating module and the edge computing module, and a second opening is provided between the edge computing module and the insect collection box.

[0038] In some examples, the edge computing module also includes a control unit for controlling the opening and closing of the first opening and the second opening.

[0039] In some examples, the device also includes a dual-mode power supply system, which supports 220V AC power and battery operating modes.

[0040] In some examples, the dual-mode power supply system has an adaptive power consumption management function.

[0041] In one possible implementation, through a hardware sleep mechanism (CPU frequency is reduced to 10% during non-detection periods) and dynamic power management (camera and communication modules are turned on and off as needed), the standby power consumption of the entire device can be ≤1W and the peak power consumption can be ≤5W.

[0042] Optionally, the device provided in the embodiment of the present application is installed in the bottom cavity of the standard high-altitude detection lamp, and a possible workflow thereof includes: when the ambient humidity exceeds 75%, the infrared heating tube starts the intermittent working mode to dry the insect body; after the insect body to be detected enters the imaging area, the infrared photoelectric sensor triggers the camera of the high-precision imaging unit to shoot continuously at a rate of 3 frames / second, and automatically adjusts the heating time and heating temperature according to the insect species in the captured image to ensure that the insect enters the inactivation state for easy shooting. The edge computing board calls a lightweight two-stage target detection model based on 80,000 sample training, filters out static interference and extracts dynamic targets through the background difference method, and uses a two-stage detection mechanism (first stage positioning of the pest candidate frame, second stage fine-grained classification) to complete the recognition, and uploads to the management platform in real time through 4G / LoRa dual-mode transmission and pushes the farmer terminal synchronously, realizing the closed loop of insect monitoring and early warning. The compatibility test of the device shows that it is compatible with the mainstream high-altitude lamp models on the market, and there is no need to recalibrate when replacing the equipment. The disassembly and assembly time is ≤3 minutes, and it can be quickly upgraded to an intelligent high-altitude lamp, which greatly improves the recognition automation rate, accuracy and real-time performance compared with the traditional manual classification and counting working method. At the same time, the two-stage model reduces data transmission by 90% compared to traditional cloud solutions, and the edge recognition delay is ≤200ms. In addition, the device has better energy sustainability, with an average daily energy consumption of ≤0.5kWh under solar power supply, which is suitable for areas without grid coverage.

[0043] The above are only preferred embodiments of the present invention, and the protection scope of the present invention is not limited to the above embodiments. Any equivalent modifications or changes made by ordinary technicians in this field based on the contents disclosed by the present invention should be included in the protection scope recorded in the claims.

Claims

1. A quick-detachable intelligent migratory pest monitoring device based on edge computing, characterized in that: The device comprises: A quick-release interface, which is used to achieve quick disassembly from the insect monitoring equipment; A heating module, the heating module is used to prevent the insect body from sticking; An edge computing module, wherein the edge computing module is used to generate monitoring results of the migratory pests and upload the monitoring results to a management platform, wherein the monitoring results include the types and quantities of the migratory pests.

2. The device according to claim 1, characterized in that The quick release interface comprises: An intelligent electromagnet array, which is composed of a plurality of independently controlled electromagnet units, each of which has a built-in pressure sensor for detecting the contact pressure with the inner wall of the insect monitoring device. The electromagnet array dynamically activates the electromagnet units in the corresponding area according to the shape of the inner wall. After power is turned on, it adaptively fits the monitoring devices with different inner diameters and shapes through magnetic attraction. The electromagnet array is connected to an intelligent power management module and automatically switches to a low-power maintenance mode after the adsorption is completed, thereby reducing energy consumption. The waterproof sealing layer adopts a double-layer structure, the inner layer is an elastic silicone wrapped electromagnet array, and the outer layer is a self-repairing polyurethane coating, which can automatically restore the sealing after being deformed by pressure.

3. The device according to claim 1, characterized in that The edge computing module includes: A fill light system, the fill light system comprising a fill light, the fill light system is used to reduce ghosting of insect images and improve insect recognition accuracy; A high-precision imaging unit, the high-precision imaging unit is used to collect images of the migratory pests trapped, and the high-precision imaging unit works synchronously with the fill light system; A controller having a built-in two-stage edge detection model and a communication unit, wherein the two-stage edge detection model is used to identify the categories of the migratory pests and count them, and the communication unit is used to upload the monitoring results to the management platform, and the communication unit is also used to receive instructions from the management platform.

4. The device according to claim 3, characterized in that The two-stage edge detection model workflow includes: the first stage quickly screens potential pest areas through low-resolution images, and the second stage performs high-precision classification and counting of candidate areas, significantly reducing the amount of calculation.

5. The device according to claim 1, characterized in that The heating module includes a dual-mode heating mode: a flexible heating pad is wrapped around the outer wall of the heating module, and an infrared heating tube is built in the falling path of the insect body to directional radiate the insect body, and the migrating pests are heated to an inactive state and the wings of the migrating pests are directed toward the camera through variable temperature control and heating time adjustment, so as to prevent the reduction of recognition accuracy due to adhesion and feature loss of the insect body, and at the same time avoid high temperature damage to the insect body morphology; A temperature control unit is used to adaptively control the heating power of the flexible heating pad and the built-in infrared heating tube according to environmental changes.

6. The device according to claim 1, characterized in that The device also includes an insect collection box, which is detachably connected to the edge computing module and is used to collect the migratory pests.

7. The device according to claim 6, characterized in that A first opening is provided between the heating module and the edge computing module, and a second opening is provided between the edge computing module and the insect collecting box.

8. The device according to claim 7, characterized in that The edge computing module also includes a control unit, which is used to control the opening and closing of the first opening and the second opening.

9. The device according to claim 1, characterized in that The device also includes a dual-mode power supply system, which supports 220V AC power and battery working modes.

10. The device according to claim 9, characterized in that The dual-mode power supply system has an adaptive power consumption management function.

Images