Laser deinsectization device based on machine vision and deinsectization method thereof

Abstract

The invention belongs to the technical field of laser deinsectization, and particularly relates to a laser deinsectization device based on machine vision and a deinsectization method thereof.The laser deinsectization device comprises a base, an image acquisition and recognition camera, a controller, an upper cover, a laser generator, an air outlet pipe and a suction fan, and the deinsectization method comprises the following specific steps that S1, the laser deinsectization device is placed at the position where deinsectization is needed; s2, the controller controls a suction fan to work, and the suction fan sucks air in the upper cover body outwards through an air collecting cover and an air outlet pipe; and S3, the laser generator generates laser, the laser is output to the filter screen plate, momentary heating is conducted to kill the winged insects, and then the controller controls the laser generator to be closed. After identification and judgment, the laser generator is started, long-time starting of the laser generator is avoided, the service life of the laser generator is guaranteed, meanwhile, the filter screen plate is prevented from being in a high-temperature state for a long time, and a certain protection effect on the filter screen plate is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of laser insecticide, and in particular to a laser insecticide device and an insecticide method based on machine vision. Background Art

[0002] Agricultural pest control is a key link in ensuring crop yield and quality, and appropriate methods must be selected according to different crops, pest types and environmental conditions.

[0003] Existing insecticide methods and defects:

[0004] Chemical pesticides: Spraying pesticides like organophosphates and pyrethroids paralyzes or poisons the insect's nervous system, disrupting its physiological functions. Disadvantages: Pesticide residues are harmful to human health; long-term use can lead to increased pesticide resistance in pests; they harm beneficial insects like bees, disrupting the ecological balance; and they pollute soil, water, and air.

[0005] Physical trapping: Exploiting the insects' tendency to move toward light, blacklights or frequency-oscillating insecticidal lamps are used to attract the insects, causing them to be electrocuted or fall into a collection device. Alternatively, sticky traps are used to kill the insects with a sticky substance. Disadvantages: The trapping range is limited, making it difficult to cover large areas; beneficial insects are easily caught by mistake; frequent cleaning and replacement of the sticky traps is required, resulting in high labor costs; and the traps are less effective against pests with strong flying abilities and wide ranges.

[0006] Biological control methods include releasing natural enemies of pests (e.g., Trichogramma wasps to control corn borers), infecting pests with entomopathogens (e.g., Beauveria bassiana), and using insect growth regulators to disrupt normal pest growth and development. Disadvantages: The reproduction and release of natural enemies are subject to environmental constraints, making their effectiveness inconsistent; microbial agents are slow to act, making it difficult to control large-scale pests in a timely manner; and the introduction of foreign natural enemies can cause new ecological problems.

[0007] The above are the most commonly used insect control methods and their defects. With the development of laser technology, there are also some technologies that introduce lasers into insect control. However, it is difficult to match lasers with insects, which makes it difficult to kill insects in time. If the laser is turned on for a long time to kill insects, the corresponding position will be focused and heated, causing the corresponding position to be in a high temperature state for a long time and damaged. Summary of the Invention

[0008] The purpose of the present invention is to provide a laser insect extermination device and an insect extermination method based on machine vision to solve the problems raised in the above background technology.

[0009] To achieve the above objectives, the present invention provides the following technical solution: a laser insecticide device based on machine vision, comprising:

[0010] Base, image acquisition and recognition camera, controller, upper cover, laser generator, air outlet pipe and suction fan;

[0011] The base is hollow and has an opening at the top, the image acquisition and recognition camera and the controller are installed in the base, and the upper cover is installed at the upper end of the base through a support rod;

[0012] The upper cover is hollow and has an opening at the bottom. The laser generator is mounted on one side of the upper cover. An air collecting hood is provided on the side of the upper cover corresponding to the laser generator. A laser output port corresponding to the laser generator is provided on the upper cover. A filter screen is provided inside the air collecting hood. The air outlet pipe is connected to the outer end of the air outlet pipe. The suction fan is mounted on the pipe of the air outlet pipe. The laser output by the laser generator corresponds to the position of the filter screen.

[0013] The controller is electrically connected to the laser generator and the image acquisition and recognition camera.

[0014] Preferably, an outer protective cover is provided on one side wall of the upper cover body, and the outer protective cover covers the outside of the laser generator. Ventilation holes are opened on the side wall of the outer protective cover, and a lower support frame is installed on the bottom of the base.

[0015] Preferably, a side wall of the outer protective cover facing the upper cover body is provided with an adhesive strip, and the outer protective cover is bonded to the upper cover body via the adhesive strip.

[0016] Preferably, a transparent protective cover is detachably connected to the upper opening of the base, and the image acquisition and recognition camera is located on the lower side of the transparent protective cover.

[0017] Preferably, the filter screen plate is made of aluminum alloy, and a heat conducting plate is cross-arranged in a horizontal and vertical pattern on one side of the filter screen plate facing the interior of the gas collecting hood.

[0018] Preferably, it also includes a back-blowing pipe, both ends of which are connected to the pipeline of the air outlet pipe, and both ends of the back-blowing pipe are located on both sides of the suction fan. The air outlet pipe is connected to the back-blowing air inlet pipe, and electric-controlled valves are installed on the pipeline of the back-blowing air inlet pipe and the pipeline of the air outlet pipe. Two electric-controlled valves are set on the pipeline of the air outlet pipe, and the electric-controlled valve on the air outlet pipe is located on the side of the air collecting hood away from the back-blowing pipe, and the controller is electrically connected to the electric-controlled valve.

[0019] Preferably, a drive motor is installed inside the upper cover body, a turntable is connected to the output shaft of the drive motor, and a reflector is embedded in a circular array at the edge of the turntable. The laser path emitted by the laser generator passes through the reflector and is reflected by the reflector to the filter plate. The drive motor is electrically connected to the electric control valve.

[0020] Preferably, a protective screen is installed inside the upper cover, and the driving motor, turntable and reflector are all located inside the protective screen. The protective screen is provided with an opening for the laser to pass through.

[0021] A method for disinfestation based on machine vision, which uses a laser disinfestation device based on machine vision, and the specific steps of the method are as follows:

[0022] S1: Place the laser insecticide device at the location where insects need to be eliminated. The controller is connected to an external power supply and can also power the electrical appliances in the laser insecticide device.

[0023] S2: The controller controls the suction fan to work. The suction fan draws the air in the upper hood outward through the air collecting hood and the air outlet pipe. The air collecting hood is in a negative pressure state. The external air is replenished into the air collecting hood from the bottom of the air collecting hood, and at the same time, it drives the flying insects in the air to gather in the air collecting hood.

[0024] In the air collecting hood, the air gathers towards the filter plate and is filtered through the filter plate, so that the flying insects in the air are filtered out;

[0025] The image acquisition and recognition camera has been trained through deep learning. It is aimed at the filter plate and identifies in real time whether there are flying insects on the filter plate.

[0026] S3: When there are flying insects on the filter plate, the controller controls the laser generator to turn on, the laser generator generates laser, and the laser is output to the filter plate, instantly heating the flying insects, and then the controller controls the laser generator to turn off.

[0027] Compared with the prior art, the present invention has the following beneficial effects:

[0028] Under the cooperation between visual recognition and laser insecticide in this solution, the image acquisition and recognition camera observes in real time whether there are flying insects on the filter plate. When there are flying insects on the filter plate, the image acquisition and recognition camera feeds back the recognition results to the controller, and the controller controls the laser generator to start insecticide.

[0029] Through the setting of the suction fan, the suction fan draws the peripheral air into the upper cover, so that the flying insects in the air are concentrated on the filter plate inside the upper cover, which is convenient for gathering the flying insects on the filter plate and facilitating laser insecticide.

[0030] After identification and judgment, turn on the laser generator to avoid the laser generator being turned on for a long time, ensure the service life of the laser generator, and at the same time prevent the filter plate from being in a high temperature state for a long time, which plays a certain protective role on the filter plate. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 It is a schematic diagram of the structure of the present invention;

[0032] Figure 2 This is a schematic structural diagram of the base and the transparent protective cover plate of the present invention;

[0033] Figure 3 This is a schematic structural diagram of the laser output port, laser generator, and outer protective cover of the present invention;

[0034] Figure 4 This is a schematic structural diagram of the gas collecting hood, filter screen plate, and heat conducting sheet of the present invention;

[0035] Figure 5 Schematic diagram of the internal structure of the upper cover body of the present invention.

[0036] In the figure: 1. Base; 2. Controller; 3. Image acquisition and recognition camera; 4. Transparent protective cover; 5. Lower support frame; 6. Upper cover; 7. Laser output port; 8. Laser generator; 9. Outer protective cover; 10. Adhesive strip; 11. Drive motor; 12. Turntable; 13. Reflector; 14. Protective mesh; 15. Opening; 16. Gas collecting hood; 17. Filter mesh; 18. Heat conducting plate; 19. Air outlet duct; 20. Suction fan; 21. Back-blowing pipe; 22. Electric control valve; 23. Back-blowing inlet pipe. DETAILED DESCRIPTION

[0037] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0038] In the description of the present invention, it should be understood that the terms "upper", "lower", "front", "back", "left", "right", "top", "bottom", "inside", "outside", etc., indicating directions or positional relationships, are based on the directions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific direction, be constructed and operated in a specific direction, and therefore should not be understood as limiting the present invention.

[0039] Example 1:

[0040] See also Figure 1-5 The present invention provides a technical solution: a laser insecticide device based on machine vision, comprising: a base 1, an image acquisition and recognition camera 3, a controller 2, an upper cover 6, a laser generator 8, an air outlet pipe 19 and a suction fan 20;

[0041] Among them, the base 1 is hollow and has an open top, the image acquisition and recognition camera 3 and the controller 2 are installed in the base 1, and the upper cover body 6 is installed on the upper end of the base 1 through a support rod; the upper cover body 6 is hollow and has an open bottom, the laser generator 8 is installed on one side of the upper cover body 6, and an air collecting hood 16 is provided on the side of the upper cover body 6 corresponding to the laser generator 8. A laser output port 7 corresponding to the laser generator 8 is provided on the upper cover body 6, and a filter plate 17 is provided in the air collecting hood 16. The air outlet pipe 19 is connected to the outer end of the air collecting hood 16, and the suction fan 20 is installed on the pipeline of the air outlet pipe 19. The laser output by the laser generator 8 corresponds to the position of the filter plate 17; the controller 2 is electrically connected to the laser generator 8 and the image acquisition and recognition camera 3.

[0042] Analysis of the above content: The implementation of the above solution is shown in Example 9. Here, the image acquisition and recognition camera 3 and the controller 2 need to undergo deep learning training. The deep learning training is based on existing technology and is briefly introduced here:

[0043] 1. Data Collection and Preprocessing

[0044] Multi-scene image acquisition

[0045] In agricultural scenes such as farmlands and orchards, flying insect images are collected through high-definition cameras, drones, fixed monitoring equipment, etc., covering the flying insect morphologies under different lighting, weather (such as sunny, cloudy, and rainy days), and crop types (such as rice, wheat, and fruit trees).

[0046] The collection objects include pests (such as aphids, moths, and beetles) and beneficial insects (such as bees and ladybugs) to ensure data diversity.

[0047] Data annotation and enhancement

[0048] Use tools such as LabelMe and YOLO format to manually label the flying insects in the image, marking information such as category (such as cotton bollworm, whitefly), location (bounding box), etc.

[0049] Expanding the dataset through data augmentation techniques:

[0050] Geometric transformation: rotation, flipping, scaling, simulating the shapes of flying insects from different perspectives.

[0051] Color transformation: adjust brightness, contrast, and saturation to adapt to different lighting conditions.

[0052] Noise addition: simulate image blur and noise interference to improve model robustness.

[0053] Synthetic Data: Generative Adversarial Networks (GANs) are used to synthesize images of rare or tiny flying insects.

[0054] 2. Model Construction and Training

[0055] 1. Classic deep learning model

[0056] Convolutional Neural Networks (CNNs)

[0057] Structure: The model is constructed through multiple layers of convolutional layers (extracting spatial features), pooling layers (dimensionality reduction), and fully connected layers (classification decisions).

[0058] Typical model:

[0059] ResNet (Residual Network): solves the gradient vanishing problem in deep network training and is suitable for complex feature extraction.

[0060] YOLO (You Only Look Once): A single-stage target detection model that balances speed and accuracy and is suitable for real-time field monitoring.

[0061] Faster R-CNN: A two-stage object detection model that first generates candidate regions (RPNs) and then classifies them. It has high accuracy but slow speed.

[0062] Transformer model

[0063] The self-attention mechanism is introduced to capture long-range dependencies in flying insect images (such as the relationship between wing texture and body structure). Representative models such as DeiT (Detection Transformer) are suitable for classifying small targets or flying insects with similar morphology (such as different subspecies of aphids).

[0064] 2. Model Optimization and Training Strategy

[0065] Loss function design

[0066] Classification task: Use cross-entropy loss to distinguish between flying insect categories.

[0067] Detection task: Use loss functions such as SmoothL1 (regression bounding box coordinates) + cross entropy (classification confidence).

[0068] Training techniques

[0069] Transfer learning: Use models pre-trained on public datasets such as ImageNet (such as ResNet50) to fine-tune the parameters of the last few layers to adapt to the flying insect recognition task, reducing the demand for training data.

[0070] Few-Shot Learning: For rare pests, it quickly adapts to new categories through meta-learning and can identify them with only a small number of labeled samples.

[0071] Online Learning: After deployment, new images are continuously collected and the model is dynamically updated to adapt to changes in field pest species.

[0072] 3. Feature Extraction and Classification Decision

[0073] 1. Multi-level feature extraction

[0074] Bottom-level features: Extract basic information about the flying insects, such as edges, textures, and colors (such as wing spots and body stripes) through shallow convolutional layers.

[0075] Middle-layer features: The middle-layer network extracts local structural features, such as head shape, antenna length, wing spread angle, etc.

[0076] High-level features: Deep networks integrate global information, learn semantic features (such as "beetles with elytra" and "aphids that are small and clustered"), and distinguish between morphologically similar pests (such as moths and butterflies).

[0077] 2. Object Detection and Classification Process

[0078] Single-stage detection (such as YOLO)

[0079] The input image is divided into grids, and each grid predicts multiple bounding boxes and their class probabilities.

[0080] Non-maximum suppression (NMS) is used to filter overlapping boxes and retain the detection results with the highest confidence.

[0081] Two-stage detection (such as FasterR-CNN)

[0082] The Region Proposal Network (RPN) generates candidate regions that may contain flying insects.

[0083] Perform feature extraction and classification on the candidate areas and output the final detection results.

[0084] Example 2:

[0085] See also Figure 1-5 The present invention provides a technical solution based on Example 1: an outer protective cover 9 is provided on one side wall of the upper cover body 6, and the outer protective cover 9 covers the outside of the laser generator 8. Ventilation holes are opened on the side wall of the outer protective cover 9, and a lower support frame 5 is installed at the bottom of the base 1.

[0086] Analysis of the above content: The outer protective cover 9 surrounds the laser generator 8 from the outside, and can protect the laser generator 8 without affecting ventilation. The lower support frame 5 can lift the whole thing higher than the surrounding crops (or other objects) to expand the area of ​​insecticide.

[0087] Example 3:

[0088] See also Figure 1-5 The present invention provides a technical solution based on the second embodiment: an adhesive strip 10 is provided on one side wall of the outer protective cover 9 facing the upper cover body 6 , and the outer protective cover 9 is bonded to the upper cover body 6 through the adhesive strip 10 .

[0089] Analysis of the above content: Adhesive strips 10 are used here to bond and fix the outer protective cover 9. Adhesive strips 10 that meet the bonding requirements are selected so that the outer protective cover 9 is stably bonded, and the outer protective cover 9 can be made of lightweight plastic material.

[0090] Example 4:

[0091] See also Figure 1-5 The present invention provides a technical solution based on the first embodiment: a transparent protective cover plate 4 is detachably connected to the upper opening of the base 1 , and the image acquisition and recognition camera 3 is located on the lower side of the transparent protective cover plate 4 .

[0092] Analysis of the above content: The transparent protective cover 4 is provided to prevent falling flying insects or dust from damaging the image acquisition and recognition camera 3 and the controller 2.

[0093] Embodiment 5:

[0094] See also Figure 1-5 The present invention provides a technical solution based on the first embodiment: the filter screen plate 17 is made of aluminum alloy, and a heat conducting plate 18 is cross-arranged in a horizontal and vertical pattern on one side of the filter screen plate 17 facing the inside of the gas collecting cover 16 .

[0095] Analysis of the above content: The heat conducting plate 18 is set in the shape of a fin. When one point of the filter plate 17 is heated, the heat is quickly transferred to other positions through the heat conducting plate 18. When the laser emission position deviates, insects can also be killed.

[0096] Example 6:

[0097] See also Figure 1-5, the present invention provides a technical solution based on Example 1: it also includes a back-blowing pipe 21, both ends of the back-blowing pipe 21 are connected to the pipeline of the air outlet pipe 19, the two ends of the back-blowing pipe 21 are located on both sides of the suction fan 20, the air outlet pipe 19 is connected to the back-blowing air inlet pipe 23, and the pipeline of the back-blowing air inlet pipe 23 and the pipeline of the air outlet pipe 19 are both installed with electric-controlled valves 22. Two electric-controlled valves 22 are set on the pipeline of the air outlet pipe 19, and the electric-controlled valves 22 on the air outlet pipe 19 are located on the side of the air collecting hood 16 away from the back-blowing pipe 21, and the controller 2 is electrically connected to the electric-controlled valve 22.

[0098] Analysis of the above content: The connection method of the electric control valve 22 is as follows: Figure 5 As shown, when the filter plate 17 is clogged due to dust or accumulated flying insects, the dust or accumulated flying insects on the filter plate 17 can be blown off regularly by backblowing. Specifically, under normal conditions, the suction fan 20 sucks air from the lower side of the upper cover body 6 to the air collecting hood 16 and the air outlet pipe 19. At this time, the electric control valve 22 on the air outlet pipe 19 is opened, and the electric control valve 22 on the backblowing air inlet pipe 23 is closed; during backblowing, the electric control valve 22 on the left side of the air outlet pipe 19 is opened, the electric control valve 22 on the right side is opened, and the electric control valve 22 on the backblowing air inlet pipe 23 is opened. The suction fan 20 sucks in the intake air through the backblowing air inlet pipe 23, and the incoming gas enters the left side of the air outlet pipe 19 through the backblowing pipe 21 and is blown off the filter plate 17, and the dust, insects, etc. on the left side wall of the filter plate 17 are blown off.

[0099] Embodiment seven:

[0100] See also Figure 1-5 The present invention provides a technical solution based on the first embodiment: a driving motor 11 is installed inside the upper cover body 6, and a turntable 12 is connected to the output shaft of the driving motor 11. The edge of the turntable 12 is embedded with a reflector 13 in a circular array. The laser path emitted by the laser generator 8 passes through the reflector 13 and is reflected by the reflector 13 to the filter plate 17. The driving motor 11 is electrically connected to the electric control valve 22.

[0101] Analyzing the above content: the reflector 13 can change the direction of the emitted laser, ultimately causing the laser to fall on the filter plate 17. The laser is reflected by different reflectors 13, causing the laser to fall on different positions of the filter plate 17. Dynamic adjustment is made based on the specific area where the flying insects are concentrated.

[0102] Embodiment 8:

[0103] See also Figure 1-5The present invention provides a technical solution based on Example 7: a protective mesh plate 14 is installed inside the upper cover body 6, and the driving motor 11, turntable 12, and reflector 13 are all located inside the protective mesh plate 14. The protective mesh plate 14 is provided with an opening 15 for the laser to pass through.

[0104] Analysis of the above content: The protective mesh plate 14 protects the inner driving motor 11, turntable 12, and reflector 13 from the outside to prevent flying insects from flying into the driving motor 11, turntable 12, and reflector 13 and affecting the driving motor 11, turntable 12, and reflector 13. The opening 15 has a range sufficient to allow laser emission at all angles.

[0105] Embodiment 9:

[0106] See also Figure 1-5 Based on the seventh embodiment, the present invention provides a technical solution: a method for disinfestation based on machine vision, which uses a laser disinfestation device based on machine vision. The specific steps of the method are as follows:

[0107] S1: Place the laser insecticide device at the location where insects need to be eliminated. Controller 2 is connected to an external power supply and can also power the electrical appliances in the laser insecticide device.

[0108] S2: The controller 2 controls the suction fan 20 to work. The suction fan 20 draws the air in the upper cover 6 outward through the air collecting hood 16 and the air outlet pipe 19. The air collecting hood 16 is in a negative pressure state. The external air is replenished into the air collecting hood 16 from the lower part of the air collecting hood 16, and at the same time, the flying insects in the air are attracted to gather in the air collecting hood 16.

[0109] In the air collecting hood 16, the air gathers toward the filter screen 17 and is filtered by the filter screen 17, so that the flying insects in the air are filtered out;

[0110] The image acquisition and recognition camera 3 is trained through deep learning. The image acquisition and recognition camera 3 is aimed at the filter plate 17 and recognizes in real time whether there are flying insects on the filter plate 17;

[0111] S3: When there are flying insects on the filter plate 17, the controller 2 controls the laser generator 8 to turn on, and the laser generator 8 generates laser light, which is output to the filter plate 17, instantly heating the flying insects, and then the controller 2 controls the laser generator 8 to turn off.

[0112] The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention. For those skilled in the art, it is obvious that the present invention is not limited to the details of the above exemplary embodiments, and the present invention can be implemented in other specific forms without departing from the spirit or basic characteristics of the present invention; therefore, no matter from which point of view, the embodiments should be regarded as exemplary and non-restrictive. The scope of the present invention is limited by the appended claims rather than the above description. Therefore, it is intended that all changes that fall within the meaning and scope of the equivalent elements of the claims are included in the present invention, and any figure signs in the claims should not be regarded as limiting the claims involved.

[0113] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A laser insecticide device based on machine vision, characterized in that: include: A base (1), an image acquisition and recognition camera (3), a controller (2), an upper cover (6), a laser generator (8), an air outlet pipe (19) and a suction fan (20); The base (1) is hollow and has an open top, the image acquisition and recognition camera (3) and the controller (2) are installed in the base (1), and the upper cover (6) is installed on the upper end of the base (1) via a support rod; The upper cover (6) is hollow and has an open lower portion. The laser generator (8) is mounted on one side of the upper cover (6). An air collecting cover (16) is provided on one side of the upper cover (6) corresponding to the laser generator (8). A laser output port (7) corresponding to the laser generator (8) is provided on the upper cover (6). A filter screen (17) is provided in the air collecting cover (16). The air outlet pipe (19) is connected to the outer end of the air collecting cover (16). The suction fan (20) is mounted on the pipeline of the air outlet pipe (19). The laser output from the laser generator (8) corresponds to the position of the filter screen (17). The controller (2) is electrically connected to the laser generator (8) and the image acquisition and recognition camera (3).

2. The laser insecticide device based on machine vision according to claim 1, characterized in that: An outer protective cover (9) is provided on one side wall of the upper cover body (6), and the outer protective cover (9) covers the outside of the laser generator (8). Ventilation holes are provided on the side wall of the outer protective cover (9), and a lower support frame (5) is installed at the bottom of the base (1).

3. The laser insecticide device based on machine vision according to claim 2, characterized in that: A bonding strip (10) is provided on a side wall of the outer protective cover (9) facing the upper cover body (6), and the outer protective cover (9) is bonded to the upper cover body (6) via the bonding strip (10).

4. The laser insecticide device based on machine vision according to claim 1, characterized in that: A transparent protective cover plate (4) is detachably connected to the upper opening of the base (1), and the image acquisition and recognition camera (3) is located on the lower side of the transparent protective cover plate (4).

5. The laser insecticide device based on machine vision according to claim 1, characterized in that: The filter screen plate (17) is made of aluminum alloy, and a heat conducting plate (18) is arranged in a horizontal and vertical cross pattern on one side of the filter screen plate (17) facing the inside of the gas collecting cover (16).

6. The laser insecticide device based on machine vision according to claim 1, characterized in that: It also includes a back-blowing pipe (21), both ends of which are connected to the pipeline of the air outlet pipe (19), and both ends of the back-blowing pipe (21) are located on both sides of the suction fan (20), and the air outlet pipe (19) is connected to the back-blowing air inlet pipe (23), and the pipeline of the back-blowing air inlet pipe (23) and the pipeline of the air outlet pipe (19) are both installed with electric control valves (22), and two electric control valves (22) are set on the pipeline of the air outlet pipe (19), and the electric control valves (22) on the pipeline of the air outlet pipe (19) are located on the side of the air collecting hood (16) away from the back-blowing pipe (21), and the controller (2) is electrically connected to the electric control valve (22).

7. The laser insecticide device based on machine vision according to claim 1, characterized in that: A driving motor (11) is installed inside the upper cover (6), and a turntable (12) is connected to the output shaft of the driving motor (11). A reflector (13) is embedded in the edge of the turntable (12) in a circular array. The laser path emitted by the laser generator (8) passes through the reflector (13) and is reflected by the reflector (13) to the filter plate (17). The driving motor (11) is electrically connected to the electric control valve (22).

8. The laser insecticide device based on machine vision according to claim 7, characterized in that: A protective screen (14) is installed inside the upper cover (6); the drive motor (11), the turntable (12), and the reflector (13) are all located inside the protective screen (14); and an opening (15) for laser light to pass through is provided on the protective screen (14).

9. A method for disinfestation based on machine vision, characterized in that: The pest control method uses the machine vision-based laser pest control device according to any one of claims 1 to 8, and the specific steps of the pest control method are as follows: S1: The laser insecticide device is placed at a location where insects need to be eliminated, and the controller (2) is connected to an external power supply, which can also power the electrical appliances in the laser insecticide device; S2: The controller (2) controls the suction fan (20) to work. The suction fan (20) draws the air in the upper cover (6) outward through the air collecting cover (16) and the air outlet pipe (19). The air collecting cover (16) is in a negative pressure state. The external air is replenished into the air collecting cover (16) from the lower part of the air collecting cover (16), and at the same time, the flying insects in the air are driven to gather in the air collecting cover (16); In the air collecting hood (16), air gathers toward the filter screen (17) and is filtered through the filter screen (17), so that flying insects in the air are filtered out; The image acquisition and recognition camera (3) is trained through deep learning, and the image acquisition and recognition camera (3) is aimed at the filter plate (17) and recognizes in real time whether there are flying insects on the filter plate (17); S3: When there are flying insects on the filter screen (17), the controller (2) controls the laser generator (8) to turn on, and the laser generator (8) generates laser light, which is output to the filter screen (17), instantly heating the flying insects, and then the controller (2) controls the laser generator (8) to turn off.

Images