An intelligent mosquito catching device

Abstract

The application discloses an intelligent mosquito catching device and relates to the technical field of mosquito killing. The device comprises a mosquito attracting chamber, a worm passing channel, a releasing channel, a mosquito catching chamber, a channel switching mechanism, a camera and a controller. The controller is used to identify whether there are insects at the entrance of the worm passing channel based on a target recognition algorithm after receiving the entrance monitoring image. If yes, the insect image is intercepted, and whether the insect is a mosquito is identified based on a convolutional neural network. If it is a mosquito, the channel switching mechanism is controlled to open the outlet of the worm passing channel and close the entrance of the releasing channel. Otherwise, the channel switching mechanism is controlled to close the outlet of the worm passing channel and connect the entrance of the releasing channel with the worm passing channel. In this way, mosquitoes can be prevented from escaping and entering the mosquito catching chamber from the mosquito attracting chamber, other insects can be prevented from entering the mosquito catching chamber and escaping, and the purpose of catching and killing only mosquitoes can be achieved, so that the biological diversity is protected.

Description

Technical Field

[0001] This invention belongs to the field of mosquito control technology, specifically relating to an intelligent mosquito trapping device. Background Technology

[0002] Mosquitoes are one of the most familiar pests to humans. Besides directly biting and sucking blood, causing pain, itching, and disrupting sleep, mosquitoes are more seriously vectors for various infectious diseases, such as malaria, lymphatic filariasis, Japanese encephalitis, dengue fever, and dengue hemorrhagic fever. Therefore, effectively killing mosquitoes has a significant effect on curbing infectious diseases.

[0003] Currently, the commonly used mosquito control methods utilize mosquitoes' attraction to light and certain chemicals to lure and kill them. For example, mosquitoes are attracted to light by using ultraviolet light and then killed by electricity. However, this method also kills beneficial insects, which can have an adverse impact on biodiversity. Summary of the Invention

[0004] The purpose of this invention is to provide an intelligent mosquito-catching device to solve the problem that existing mosquito-catching methods indiscriminately kill mosquitoes and beneficial insects, thus adversely affecting biodiversity.

[0005] To achieve the above objectives, the present invention adopts the following technical solution:

[0006] In a first aspect, an intelligent mosquito-catching device is provided, comprising a mosquito-attracting chamber, an insect-passing channel, a release channel, a mosquito-catching chamber, a channel switching mechanism, a camera, and a controller, wherein the inlet of the insect-passing channel is connected to the mosquito-attracting chamber, the outlet of the insect-passing channel is connected to the mosquito-catching chamber, the inlet of the release channel is connected to the insect-passing channel through the channel switching mechanism, and the outlet of the release channel is connected to the outside.

[0007] The channel switching mechanism is communicatively connected to the controller and is used, under the control of the controller, to close the outlet of the insect passage and connect the inlet of the release channel to the insect passage, or to open the outlet of the insect passage and close the inlet of the release channel.

[0008] The camera is communicatively connected to the controller and is used to acquire images of the entrance of the insect passage in real time and transmit the acquired entrance monitoring images to the controller in real time.

[0009] The controller is used to perform intelligent mosquito trapping according to the following steps:

[0010] Upon receiving the imported monitoring image, the imported monitoring image is imported in real time into an insect recognition model that has been pre-trained based on a target recognition algorithm, and the insect recognition result is output.

[0011] If the insect identification result contains an insect marker box, then the insect image is captured in real time from the imported monitoring image based on the insect marker box;

[0012] The insect images are imported in real time into a mosquito recognition model pre-trained based on a first convolutional neural network, and the mosquito recognition results are output.

[0013] If the mosquito identification result indicates a mosquito, the channel switching mechanism is controlled to open the exit of the insect passage and close the entrance of the release passage; otherwise, the channel switching mechanism is controlled to close the exit of the insect passage and connect the entrance of the release passage to the insect passage.

[0014] Based on the above-mentioned invention, a novel mosquito-catching scheme is provided that uses a target recognition algorithm and a convolutional neural network to selectively kill mosquitoes and beneficial insects. The scheme includes a mosquito-attracting chamber, an insect passage, a release passage, a mosquito-catching chamber, a passage switching mechanism, a camera, and a controller. Upon receiving an inlet monitoring image, the controller first uses a target recognition algorithm to identify whether there are insects at the entrance of the insect passage. If so, it captures an image of the insect and uses a convolutional neural network to identify whether the insect is a mosquito. If it is a mosquito, it controls the passage switching mechanism to open the exit of the insect passage and close the entrance of the release passage; otherwise, it controls the passage switching mechanism to close the exit of the insect passage and connect the entrance of the release passage to the insect passage. This not only prevents mosquitoes from escaping and entering the mosquito-catching chamber from the mosquito-attracting chamber, but also prevents other insects from entering the mosquito-catching chamber and escaping. This achieves the goal of capturing and killing only mosquitoes, which is beneficial for protecting biodiversity and is easy to apply and promote.

[0015] In one possible design, the mosquito-attracting chamber is surrounded by a cap, bird netting, and transverse partitions arranged from top to bottom, and contains mosquito-attracting components. The upper surface of the transverse partitions has an inlet for the insect passage.

[0016] In one possible design, the mosquito attractant uses an exhaust port for attracting mosquito gas and is arranged on the wall of a support tube, wherein the support tube is located inside the mosquito attracting chamber, one end of the support tube is fixedly connected to the cap body, and the other end of the support tube is vertically fixedly connected to the upper surface of the transverse partition in order to support the cap body.

[0017] In one possible design, a storage chamber is also included, wherein a carbon dioxide cylinder and a gas mixing chamber are housed within the storage chamber;

[0018] The inlet of the mixing chamber is connected to the carbon dioxide cylinder, and the outlet of the mixing chamber is connected to the support pipe.

[0019] In one possible design, the camera is hung upside down inside the top of the cap, with the lens aimed at the entrance of the insect passage.

[0020] In one possible design, a ventilation window is provided on the first side of the mosquito trapping chamber, and a mosquito filter is installed on the ventilation window. An exhaust fan for exhausting indoor air is also installed outside the ventilation window, wherein the controlled end of the exhaust fan is communicatively connected to the controller.

[0021] The controller is also used to control the start / stop and speed adjustment of the exhaust fan.

[0022] In one possible design, the second side of the mosquito trap is fitted with a drawer and / or a maintenance door for opening the interior, wherein the drawer is located at the bottom of the mosquito trap.

[0023] In one possible design, the channel switching mechanism includes a stepper motor and a rotating flap, wherein the controlled end of the stepper motor is communicatively connected to the controller, and the rotating flap shaft is connected to the output shaft of the stepper motor;

[0024] The rotating flap is used to close the outlet of the insect passage and connect the inlet of the release passage to the insect passage when the output shaft is rotated to the first position, and to open the outlet of the insect passage and close the inlet of the release passage when the output shaft is rotated to the second position.

[0025] In one possible design, the controller is further configured to, when the mosquito identification result indicates a mosquito, upload the insect image and the time and / or location of the insect image acquisition to a cloud server in real time, so that the cloud server can import the insect image into a mosquito classification model pre-trained based on a second convolutional neural network, output a mosquito classification result, and perform statistical analysis based on the mosquito classification result and the acquisition time and / or the acquisition location to obtain a mosquito statistical analysis report.

[0026] In one possible design, the camera is a high-speed camera, and the controller is a Raspberry Pi microcomputer.

[0027] The beneficial effects of the above scheme are:

[0028] (1) This invention creatively provides a novel mosquito trapping scheme based on target recognition algorithm and convolutional neural network for the differentiated killing of mosquitoes and beneficial insects. The scheme includes a mosquito trapping chamber, an insect passage, a release passage, a mosquito trapping chamber, a passage switching mechanism, a camera, and a controller. The controller is used to, after receiving the inlet monitoring image, first identify whether there are insects at the inlet of the insect passage based on the target recognition algorithm. If so, it captures the insect image and identifies whether the insect is a mosquito based on the convolutional neural network. If it is a mosquito, it controls the passage switching mechanism to open the outlet of the insect passage and close the inlet of the release passage. Otherwise, it controls the passage switching mechanism to close the outlet of the insect passage and connect the inlet of the release passage to the insect passage. This not only prevents mosquitoes from escaping and allows them to enter the mosquito trapping chamber from the mosquito trapping chamber, but also prevents other insects from entering the mosquito trapping chamber and allowing them to escape. This achieves the purpose of capturing and killing only mosquitoes, which is beneficial to protecting biodiversity.

[0029] (2) Mosquito images, along with the time and / or location of image acquisition, can be uploaded to a cloud server to obtain mosquito classification results based on a convolutional neural network. Statistical analysis can then be performed using the time and / or location of image acquisition to generate a mosquito statistical analysis report, which will help in developing appropriate mosquito control strategies and facilitate practical application and promotion. Attached Figure Description

[0030] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0031] Figure 1 This is a three-dimensional structural diagram of the intelligent mosquito-catching device provided in the embodiments of this application.

[0032] Figure 2 This is a three-dimensional structural diagram of the channel switching mechanism in the intelligent mosquito-catching device provided in the embodiments of this application, wherein, Figure 2 (a) shows a three-dimensional structural diagram of the channel switching mechanism when the insect passage exit is closed. Figure 2 (b) shows a three-dimensional structural diagram of the channel switching mechanism when the insect passage exit is opened.

[0033] Figure 3 This is a front view structural diagram of the intelligent mosquito-catching device provided in the embodiments of this application.

[0034] Figure 4 This is a right-side view of the intelligent mosquito-catching device provided in an embodiment of this application.

[0035] Figure 5 This is a top view of the intelligent mosquito-catching device provided in the embodiments of this application.

[0036] Figure 6 This is a three-dimensional structural diagram of a carbon dioxide gas enrichment device in an intelligent mosquito-catching device provided in an embodiment of this application.

[0037] Figure 7 This is a front view structural diagram of a carbon dioxide gas enrichment device in an intelligent mosquito-catching device provided in an embodiment of this application.

[0038] Figure 8 This is a left-side view of the carbon dioxide gas enrichment device in the intelligent mosquito-catching device provided in the embodiments of this application.

[0039] Figure 9 This is a top view of the carbon dioxide gas enrichment device in the intelligent mosquito-catching device provided in the embodiments of this application.

[0040] In the attached diagrams: 11-Cap body; 12-Bird netting; 13-Horizontal partition; 14-Support pipe; 15-Support column; 2-Insect passage; 21-Inlet; 3-Release passage; 32-Outlet; 4-Mosquito trapping chamber; 42-Mosquito filter; 43-Exhaust fan; 44-Drawer; 45-Maintenance door; 510-Output shaft; 52-Rotating flap; 6-Camera; 7-Controller; 8-Storage chamber; 81-Carbon dioxide cylinder; 82-Mix Air chamber; 83-Air hole; 100-Upper bottle body; 200-Lower bottle body; 300-Connecting pipe; 310-Air outlet; 400-Bottle body tilting mechanism; 410-Support; 411-Base plate; 412-Inverted "V" shaped support rod; 420-Rotating shaft; 430-Bearing; 431-First bearing; 432-Second bearing; 440-Motor; 500-Air supply pump; 600-High temperature filter; 700-Control module. Detailed Implementation

[0041] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the present invention will be briefly introduced below in conjunction with the accompanying drawings and descriptions of the embodiments or the prior art. Obviously, the following description of the structure of the accompanying drawings is only some embodiments of the present invention. For those skilled in the art, other embodiments can be obtained based on these embodiments without creative effort. It should be noted that the description of these embodiments is for the purpose of helping to understand the present invention, but does not constitute a limitation of the present invention.

[0042] It should be understood that although the terms "first" and "second", etc., may be used herein to describe various objects, these objects should not be limited by these terms. These terms are only used to distinguish one object from another. For example, the first object may be referred to as the second object, and similarly, the second object may be referred to as the first object, without departing from the scope of the exemplary embodiments of the invention.

[0043] It should be understood that the term "and / or" that may appear in this document is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can mean: A exists alone, B exists alone, or A and B exist simultaneously. Another example is A, B and / or C, which can mean that any one of A, B, and C or any combination thereof exists. The term " / and" that may appear in this document describes another relationship between related objects, indicating that two relationships can exist. For example, A / and B can mean: A exists alone or A and B exist simultaneously. In addition, the character " / " that may appear in this document generally indicates that the related objects before and after it are in an "or" relationship.

[0044] Example 1

[0045] like Figures 1-5 As shown, the intelligent mosquito-catching device provided in this embodiment includes, but is not limited to, a mosquito-attracting chamber, an insect-passing channel 2, a release channel 3, a mosquito-catching chamber 4, a channel switching mechanism, a camera 6, and a controller 7. The inlet 21 of the insect-passing channel 2 connects to the mosquito-attracting chamber, and the outlet of the insect-passing channel 2 connects to the mosquito-catching chamber 4. The inlet of the release channel 3 connects to the insect-passing channel 2 via the channel switching mechanism, and the outlet 32 ​​of the release channel 3 connects to the outside. Figure 1 As shown, the mosquito-attracting chamber is used to lure mosquitoes from outdoors into the room by utilizing their attraction to light and certain chemical substances. Specifically, the mosquito-attracting chamber may be, but is not limited to, surrounded by a cap 11, a bird net 12, and a horizontal partition 13 arranged from top to bottom, and contains mosquito-attracting components. The upper surface of the horizontal partition 13 has an inlet 21 for the insect passage 2. The bird net 12 is used to prevent animals larger than mosquitoes, such as birds, from entering the mosquito-attracting chamber (i.e., only animals the size of mosquitoes are allowed to enter). The mosquito-attracting components lure mosquitoes to the chamber. The insect passage 2 connects the mosquito-attracting chamber and the mosquito-catching chamber 4 when its outlet is open, allowing mosquitoes to enter the mosquito-catching chamber 4. The release passage 3 connects the mosquito-attracting chamber and the outside when the outlet of the insect passage 2 is closed, allowing non-mosquito insects to escape from the mosquito-attracting chamber, such as... Figure 2As shown, the insect passage 2 and the release passage 3, for example, employ an L-shaped passage structure with the passage switching mechanism located at the central corner. The mosquito trapping chamber 4 is used to imprison mosquitoes that enter the chamber until they die.

[0046] The channel switching mechanism is communicatively connected to the controller 7, and is used, under the control of the controller 7, to close the outlet of the insect passage 2 and connect the inlet of the release channel 3 to the insect passage 2, or to open the outlet of the insect passage 2 and close the inlet of the release channel 3. Figure 2 As shown, specifically, the channel switching mechanism includes, but is not limited to, a stepper motor and a rotating flap 52. The controlled end of the stepper motor is communicatively connected to the controller 7, and the rotating flap 52 is axially connected to the output shaft 510 of the stepper motor. The rotating flap 52 is used to close the outlet of the insect passage 2 and connect the inlet of the release channel 3 to the insect passage 2 when the output shaft 510 rotates to a first position; and to open the outlet of the insect passage 2 and close the inlet of the release channel 3 when the output shaft 510 rotates to a second position. The stepper motor is used to rotate stepwise under the control of the controller 7; the first position is as follows... Figure 2 As shown in (a), the second position is as follows Figure 2 As shown in (b), when the stepper motor is in the second position, if the stepper motor is controlled to rotate 180 degrees clockwise in a single step, it can be rotated to the first position, thereby achieving the purpose of closing the outlet of the insect passage 2 and connecting the inlet of the release channel 3 to the insect passage 2. When the stepper motor is in the first position, if the stepper motor is controlled to rotate 180 degrees counterclockwise in a single step, it can be rotated to the second position, thereby achieving the purpose of opening the outlet of the insect passage 2 and closing the inlet of the release channel 3.

[0047] The camera 6 is communicatively connected to the controller 7, and is used to acquire images of the inlet 21 of the insect passage 2 in real time, and transmit the acquired inlet monitoring images to the controller 7 in real time. Figures 3-4 As shown, specifically, the camera 6 is hung upside down inside the top of the cap 11, with its lens aimed at the inlet 21 of the insect passage 2. Furthermore, to obtain high-frequency monitoring images of the inlet, the camera 6 is preferably a high-speed camera.

[0048] The controller 7, specifically but not limited to a Raspberry Pi microcomputer (an existing microcomputer based on Linux and the size of a credit card) is arranged inside the top of the cap 11 and is used, but not limited to, to perform intelligent mosquito-catching work according to the following steps S11 to S14.

[0049] S11. Upon receiving the imported monitoring image, the imported monitoring image is imported in real time into an insect recognition model pre-trained based on a target recognition algorithm, and the insect recognition result is output.

[0050] In step S11, the object detection algorithm is an existing artificial intelligence recognition algorithm used to identify objects in an image and mark their locations. Specifically, it can be, but is not limited to, Faster R-CNN (Faster Regions with Convolutional Neural Networks features, proposed by Kaiming He et al. in 2015, which won multiple first places in the 2015 ILSVRC and COCO competitions), SSD (Single Shot MultiBox Detector, proposed by Wei Liu at ECCV 2016, and currently one of the most popular detection frameworks), or YOLO (You Only Look...) object detection algorithm. Once, currently at version V5, is widely used in the industry. Its basic principle is: first, the input image is divided into a 7×7 grid; two bounding boxes are predicted for each grid; then, low-probability target windows are removed based on a threshold; finally, redundant windows are removed using bounding box merging to obtain the detection result. Therefore, insect image recognition can be performed in real-time based on this target detection algorithm to obtain the insect recognition result. Specifically, the target recognition algorithm can be, but is not limited to, the YOLO v4 target detection algorithm. The specific model structure of the YOLO v4 target detection algorithm consists of three parts: a backbone network, a neck network, and a head network. The backbone network can use a CSPDarknet53 (CSP stands for Cross Stage Partial) network for feature extraction. The neck network consists of an SPP (Spatial Pyramid Pooling block) and a PANet (Path Aggregation Network). The former increases the receptive field and isolates the most important features, while the latter ensures that semantic features are received from higher-level layers and fine-grained features are received from lower-level layers of the lateral backbone network. The head network performs detection based on anchor boxes and detects three different feature maps of sizes: 13×13, 26×26, and 52×52, respectively, for detecting targets from largest to smallest (here, larger feature maps contain more information; therefore, the 52×52 feature map is used to detect small targets, and vice versa). The aforementioned insect recognition model can be trained using conventional sample training methods so that, after inputting a test image, it can output the recognition results of whether an image contains insects and their predicted confidence values.Furthermore, the insect recognition model can be pre-trained on a server with abundant computing resources before being deployed to the controller 7, which has relatively limited computing resources.

[0051] In step S11, considering that the imported monitoring image is acquired at a high frequency, it is preferable to select the frame with the highest clarity (the specific clarity algorithm is an existing algorithm) from multiple consecutive frames of the imported monitoring image and import it into the insect recognition model. This eliminates the need to perform insect recognition for every frame of the acquired image, thus saving computational resources. Furthermore, after cropping the imported monitoring image by defining the region of interest (i.e., the area where the import 21 is located; since the import 21 and the camera 6 are fixed, the region of interest can be pre-defined), the resulting region of interest image can be imported into the insect recognition model for insect recognition.

[0052] S12. If the insect identification result contains an insect marker box, then the insect image is captured in real time from the imported monitoring image based on the insect marker box.

[0053] In step S12, if the insect identification result includes the insect marker box, it indicates that an insect is present at the inlet 21; otherwise, it is not. Therefore, further processing is needed to identify whether the insect is a mosquito based on the insect image. Furthermore, if the insect identification result includes multiple insect marker boxes, the corresponding insect image can be cropped for each insect marker box (generally, when the size of the inlet 21 is appropriately designed, the insect identification result will contain at most one insect marker box).

[0054] S13. Import the insect image into the mosquito recognition model pre-trained based on the first convolutional neural network in real time, and output the mosquito recognition result.

[0055] In step S13, Convolutional Neural Networks (CNNs) are a type of feedforward neural network that includes convolutional computation and has a deep structure. They are one of the representative algorithms of deep learning and are commonly used for classifying and recognizing input images. Therefore, the mosquito recognition model can be trained based on a certain number of positive sample images (i.e., mosquito sample images) and negative sample images (i.e., other insect sample images) using a conventional calibration and verification modeling method (the specific process includes model calibration and verification, i.e., first comparing the model simulation results with the measured data, and then adjusting the model parameters based on the comparison results to make the simulation results match the actual results). Furthermore, the mosquito recognition model can also be pre-trained on a server with abundant computing resources before being deployed to the controller 7, which has relatively limited computing resources.

[0056] S14. If the mosquito identification result indicates a mosquito, control the channel switching mechanism to open the outlet of the insect passage 2 and close the inlet of the release channel 3; otherwise, control the channel switching mechanism to close the outlet of the insect passage 2 and connect the inlet of the release channel 3 to the insect passage 2.

[0057] In step S14, when the mosquito identification result indicates a mosquito, the channel switching mechanism opens the outlet of the insect passage 2 and closes the inlet of the release channel 3, thus preventing the mosquito from escaping and allowing it to enter the mosquito trapping chamber 4 from the mosquito-attracting chamber. Conversely, when the mosquito identification result indicates an insect other than a mosquito, the outlet of the insect passage 2 is closed and the inlet of the release channel 3 is connected to the insect passage 2, preventing the other insect from entering the mosquito trapping chamber 4 and allowing it to escape. This achieves the goal of capturing and killing only mosquitoes, which is beneficial for protecting biodiversity and facilitates practical application and promotion. Furthermore, to ensure that the channel switching mechanism can be controlled in a timely manner to close the outlet of the insect passage 2 and connect the inlet of the release channel 3 to the insect passage 2, the length of the insect passage 2 needs to be positively correlated with the execution time required for the aforementioned steps S11 to S14; that is, the longer the execution time required for the aforementioned steps S11 to S14, the longer the length of the insect passage 2 needs to be.

[0058] Preferably, the mosquito-attracting component uses a mosquito-attracting gas exhaust port and is arranged on the wall of the support pipe 14. The support pipe 14 is located inside the mosquito-attracting chamber. One end of the support pipe 14 is fixedly connected to the cap 11, and the other end is vertically fixedly connected to the upper surface of the transverse partition 13 to support the cap 11. Figures 4-5 As shown, in addition to the support tube 14, there are three support columns 15 for supporting the cap 11 to ensure stability. The mosquito-attracting gas can be, but is not limited to, a conventional mixture of carbon dioxide gas, air, and mosquito bait odor. Specifically, it also includes a storage chamber 8, which contains a carbon dioxide cylinder 81 and a mixing chamber 82; the inlet of the mixing chamber 82 is connected to the carbon dioxide cylinder 81, and the outlet of the mixing chamber 82 is connected to the support tube 14. This allows for a continuous supply of the mosquito-attracting gas to lure mosquitoes.

[0059] Preferably, a ventilation window is provided on the first side of the mosquito trapping chamber 4, and a mosquito filter 42 is installed on the ventilation window. An exhaust fan 43 for exhausting indoor air is also installed outside the ventilation window. The controlled end of the exhaust fan 43 is communicatively connected to the controller 7. The controller 7 is also used to control the start / stop and speed adjustment of the exhaust fan 43 (for example, increasing the speed of the exhaust fan 43 when the mosquito identification result indicates a mosquito, and decreasing the speed of the exhaust fan 43 when the mosquito identification result indicates a non-mosquito). Figures 3-4 As shown, a ventilation window is specifically provided on the rear-view side of the mosquito-catching chamber 4. The mosquito filter 42 is used to prevent indoor mosquitoes from escaping. The exhaust fan 43 is used to generate sufficient indoor negative pressure after activation, so as to draw mosquitoes passing through the inlet 21 from the mosquito-attracting chamber into the mosquito-catching chamber 4, and prevent indoor mosquitoes from escaping from the mosquito-catching chamber to the mosquito-attracting chamber through the insect passage 3. Specifically, the ventilation window is provided on the vertical partition located between the mosquito-catching chamber 4 and the storage chamber 8, and the exhaust fan 43 is located inside the storage chamber 8. Several air holes 83 are also provided on the side wall of the storage chamber 8 to ensure smooth airflow while keeping the entire intelligent mosquito-catching device neat and aesthetically pleasing. The exhaust fan 43 operates continuously after being turned on. To further ensure smooth airflow, the outlet of the insect passage 2 can be set to a normally open position, and the inlet of the release channel 3 can be set to a normally closed position. This allows the channel switching mechanism to close the outlet of the insect passage 2 and open the inlet of the release channel 3 only when the mosquito identification result indicates a non-mosquito. Furthermore, to ensure timely control of the channel switching mechanism to close the outlet of the insect passage 2 and open the inlet of the release channel 3, the length of the insect passage 2 must be positively correlated with the rotational speed of the exhaust fan 43; that is, the faster the exhaust fan 43 rotates, the longer the insect passage 2 needs to be.

[0060] Preferably, a drawer 44 and / or a maintenance door 45 for opening the interior are installed on the second side of the mosquito trapping chamber 4, wherein the drawer 44 is located at the bottom of the mosquito trapping chamber 4. Figures 4-5 As shown, specifically, the drawer 44 and the maintenance door 45 are installed on the main side of the mosquito trapping chamber 4. The drawer 44 is used to hold the killed indoor mosquitoes and can be pulled out for cleaning at any time. The maintenance door 45 can be opened at any time to repair the mosquito filter 42 or other indoor components.

[0061] Preferably, the controller 7 is further configured to, when the mosquito identification result indicates a mosquito, upload the insect image and the acquisition time and / or acquisition location of the insect image to a cloud server in real time, so that the cloud server can import the insect image into a mosquito classification model pre-trained based on a second convolutional neural network, output a mosquito classification result, and perform statistical analysis based on the mosquito classification result and the acquisition time and / or acquisition location to obtain a mosquito statistical analysis report. The classification ability of the mosquito classification model may include, but is not limited to, mosquito species classification ability, mosquito growth stage classification ability, and / or mosquito male and female classification ability, etc. Therefore, the mosquito statistical analysis report specifically includes, but is not limited to, mosquito species statistical analysis results based on time latitude and / or geographical latitude, mosquito growth stage statistical analysis results, and mosquito male and female statistical analysis results, which can help to make appropriate mosquito control strategies. Because the mosquito classification model has a stronger classification ability than the mosquito recognition model, the second convolutional neural network is more complex than the first convolutional neural network (e.g., the second convolutional neural network uses an existing RCNN or ResNet network) and requires more computing resources. Therefore, it needs to be deployed on the cloud server to complete the mosquito classification task. Furthermore, the mosquito classification model can also be trained based on a certain number of sample images and sample labels using conventional calibration and verification modeling methods.

[0062] In summary, the intelligent mosquito-catching device provided in this embodiment has the following technical effects:

[0063] (1) This embodiment provides a new mosquito trapping scheme based on target recognition algorithm and convolutional neural network for differentiated killing of mosquitoes and beneficial insects. It includes a mosquito trapping chamber, an insect passage, a release passage, a mosquito trapping chamber, a passage switching mechanism, a camera, and a controller. The controller is used to identify whether there are insects at the entrance of the insect passage after receiving the inlet monitoring image. If so, the insect image is captured and the convolutional neural network is used to identify whether the insect is a mosquito. If it is a mosquito, the passage switching mechanism is controlled to open the exit of the insect passage and close the entrance of the release passage. Otherwise, the passage switching mechanism is controlled to close the exit of the insect passage and connect the entrance of the release passage to the insect passage. This not only prevents mosquitoes from escaping and allows them to enter the mosquito trapping chamber from the mosquito trapping chamber, but also prevents other insects from entering the mosquito trapping chamber and allowing them to escape. This achieves the purpose of capturing and killing only mosquitoes, which is beneficial to protecting biodiversity.

[0064] (2) Mosquito images, along with the time and / or location of image acquisition, can be uploaded to a cloud server to obtain mosquito classification results based on a convolutional neural network. Statistical analysis can then be performed using the time and / or location of image acquisition to generate a mosquito statistical analysis report, which will help in developing appropriate mosquito control strategies and facilitate practical application and promotion.

[0065] Example 2

[0066] Based on the technical solution of Embodiment 1, this embodiment also provides a novel intelligent mosquito-catching solution that replaces the carbon dioxide cylinder 81 with a carbon dioxide gas enrichment device, namely as follows: Figures 6-9 As shown, the carbon dioxide gas enrichment device used to replace the carbon dioxide cylinder 81 includes, but is not limited to, an upper cylinder 100, a lower cylinder 200, a connecting pipe 300, a cylinder tilting mechanism 400, an air supply pump 500, and a high-temperature filter 600. The openings of the upper cylinder 100 and the lower cylinder 200 are vertically opposite each other and connected by the connecting pipe 300. The lower cylinder 200 contains a carbon dioxide adsorbent solution. The cylinder tilting mechanism 400 is used to tilt the upper cylinder 100, the lower cylinder 200, and the connecting pipe vertically. The bottle assembly structure consists of 300 components; the output port of the air supply pump 500 is connected to the bottom of the inner cavity of the lower bottle 200 so as to introduce outside air into the carbon dioxide adsorbent solution, thereby causing the carbon dioxide gas in the air to react with the carbon dioxide adsorbent in the carbon dioxide adsorbent solution to form water-insoluble carbonate crystals; the high-temperature filter 600 is arranged inside the connecting pipe 300 so as to intercept the carbonate crystals when the bottle assembly structure is flipped up and down, and decompose the carbonate crystals at high temperature to obtain carbon dioxide gas and carbon dioxide adsorbent that can fall to the bottom on its own.

[0067] like Figures 6-9 As shown, in the specific structure of the carbon dioxide gas enrichment device, the upper bottle 100 and the lower bottle 200 are a pair of bottles that can be interchanged by flipping them upside down, so as to hold the carbon dioxide adsorbent solution at different times; as Figures 6-7 As shown, specifically, the upper bottle 100 and / or the lower bottle 200 can be, but are not limited to, conical flasks. The connecting pipe 300 is used to connect the upper bottle 100 and the lower bottle 200 so that the carbon dioxide adsorbent solution can flow from one bottle to the other during the up-and-down inversion process. In order to be able to export the air that has been adsorbed with carbon dioxide gas and the carbon dioxide gas obtained from high-temperature decomposition, specifically, an air outlet 310 is provided on the side wall of the connecting pipe 300; and in order to further export the air that has been adsorbed with carbon dioxide gas and the carbon dioxide gas obtained from high-temperature decomposition to different flow directions at different times, the air outlet 310 can be connected to the air inlet port of a three-way valve so that when outside air is introduced into the carbon dioxide adsorbent solution, the air inlet port of the three-way valve and the first air outlet port of the three-way valve can be selectively connected to achieve the export of the air that has been adsorbed with carbon dioxide gas and the carbon dioxide gas obtained from high-temperature decomposition to different flow directions at different times, the air outlet 310 can be connected to the air inlet port of a three-way valve so that when outside air is introduced into the carbon dioxide adsorbent solution, the air inlet port of the three-way valve can be selectively connected to the first air outlet port of the three-way valve to achieve the export of the air that has been adsorbed with carbon dioxide gas and the carbon dioxide gas obtained from high-temperature decomposition to different flow directions at different times, the air outlet 310 can be connected to the air inlet port of a three-way valve so that the air inlet port of the three-way valve and the first air outlet port of the three-way valve can be selectively connected to the air inlet port of the three-way valve so as to achieve the export of the air that has been adsorbed with carbon dioxide gas and the carbon dioxide gas obtained from high-temperature decomposition to different flow directions at different times, the air outlet 310 can be connected to The purpose of directing carbon dioxide gas into the first flow direction is to, during the high-temperature decomposition of the carbonate crystals, connect the inlet port of the three-way valve to the second outlet port of the three-way valve to direct the carbon dioxide gas obtained from the high-temperature decomposition into the second flow direction (e.g., the flow direction to the mosquito killer, i.e., the second outlet port of the three-way valve connects to the inlet end of the mixing chamber 82). Simultaneously, to prevent the carbon dioxide adsorbent solution from leaking out through the outlet 310 during the up-and-down flipping process, a gas separation membrane is preferably arranged in the outlet 310. The air supply pump 500 can be, but is not limited to, using an existing air pump. The high temperature in the high-temperature filter 600 is a temperature relatively high compared to room temperature and needs to reach the decomposition temperature of the carbonate crystals (e.g., above 35 degrees Celsius). It generally needs to be obtained by heating after intercepting the carbonate crystals. That is, the high-temperature filter 600 can be, but is not limited to, woven from electric heating wires and non-electric heating wires (specifically, woven from electric heating wires or woven from both), so that the high temperature can be obtained through heating by the electric heating wires. Furthermore, the carbon dioxide adsorbent may specifically be, but is not limited to, guanidine or m-phenylenediamine.

[0068] The working principle of the aforementioned carbon dioxide gas enrichment device may include, but is not limited to, the following: (1) introducing outside air into the carbon dioxide adsorbent solution through the air supply pump 500, so that the carbon dioxide gas in the air reacts with the carbon dioxide adsorbent in the carbon dioxide adsorbent solution to form water-insoluble carbonate crystals; (2) after a period of time, turning off the air supply pump 500, and flipping the bottle assembly structure up and down through the bottle body flipping mechanism 400, so that the upper bottle body 100 and the lower bottle body 200 are interchanged, and the carbon dioxide adsorbent solution flows into the new lower bottle body (i.e., the upper bottle body 100 before the change of identity) after passing through the high-temperature filter 600. At this time, due to the The presence of the high-temperature filter 600 will intercept the carbonate crystals; (3) the carbonate crystals are decomposed at high temperature by the high-temperature filter 600 to obtain carbon dioxide gas and carbon dioxide adsorbent that can fall to the bottom (i.e., it falls to the carbon dioxide adsorbent solution below and dissolves to obtain a carbon dioxide adsorbent solution with a higher concentration). Although there is air in the bottle at this time, since the carbon dioxide gas obtained from the decomposition is adsorbed from a large amount of air outside the bottle, the carbon dioxide gas obtained from the decomposition can still achieve the purpose of enriching carbon dioxide gas after mixing with the air inside the bottle (note that this is not the purpose of purifying carbon dioxide gas); (4) the air outside the bottle is introduced into the carbon dioxide adsorbent solution again by the air supply pump 500 to start the next cycle. Thus, the carbon dioxide gas enricher provided in this embodiment can greatly simplify the air separation system and reduce the system volume while achieving the purpose of enriching carbon dioxide gas, making it suitable for small devices such as mosquito killing devices that use the attraction of mosquitoes to carbon dioxide gas to kill mosquitoes, which is convenient for practical application and promotion.

[0069] Preferably, the bottle body tilting mechanism 400 includes, but is not limited to, a bracket 410, a rotating shaft 420, a bearing 430, and a motor 440, wherein the rotating shaft 420 is fixed and horizontally passes through the middle of the connecting pipe 300; the rotating shaft 420 is horizontally mounted on the top of the bracket 410 via the bearing 430, suspending the bottle body assembly structure; the output shaft of the motor 440 is drively connected to the rotating shaft 420. Figures 6-9 As shown, the aforementioned bottle-flipping mechanism allows for easy flipping of the bottle assembly structure. To further enhance stability during the flipping process, preferably, the support 410 includes, but is not limited to, a base plate 411 and inverted "V"-shaped support rods 412. The two ends of the inverted "V"-shaped support rods 412 are fixedly connected to the upper surface of the base plate 411, so that the inflection point at the middle of the inverted "V"-shaped support rods 412 forms the top of the support 410. The number of inverted "V"-shaped support rods 412 can be as follows: Figure 6 The diagram shows two bearings, but it can also show one or more. To further enhance stability during the flipping process, preferably, the bearing 430 includes, but is not limited to, a first bearing 431 and a second bearing 432; one end of the rotating shaft 420 is mounted on the top side of the bracket 410 via the first bearing 431, and the other end of the rotating shaft 420 is mounted on the other side of the top of the bracket 410 via the second bearing 432, so that the bottle assembly structure is suspended between the top side of the bracket 410 and the other side of the top of the bracket 410. The motor 440 can be as follows... Figure 6 It can be mounted on top of the bracket 410 as shown, or it can be mounted on the base plate 411. Furthermore, the motor 440 can be, but is not limited to, a stepper motor.

[0070] Preferably, there are two air supply pumps 500, and their output ports are respectively connected to the bottom of the inner cavity of the two bottles that are divided into lower bottles in the bottle assembly structure. Figures 6-9 As shown, based on the fact that the upper bottle 100 and the lower bottle 200 are a pair of bottles that can be interchanged by flipping them upside down, carbon dioxide gas can be continuously enriched, thereby improving the enrichment efficiency.

[0071] Preferably, the system also includes a control module 700, wherein the output of the control module 700 is communicatively connected to the controlled end of the bottle tilting mechanism 400 (specifically, the controlled end of the motor 440) and the controlled end of the air supply pump 500. This allows the tilting action of the bottle tilting mechanism 400 and the start / stop action of the air supply pump 500 to be controlled by the control module 700 (specifically implemented through internal conventional control circuitry), thereby facilitating intelligent operation. Furthermore, the output of the control module 700 can also be communicatively connected to the controlled end of the power supply module of the electric heating wire and the controlled end of the three-way valve, so that the heating start / stop action of the electric heating wire and the channel switching action of the three-way valve are also controlled by the control module 700, further facilitating intelligent operation.

[0072] In summary, based on the technical effects of Embodiment 1, this embodiment also has the following technical effects: (1) By replacing the carbon dioxide cylinder with a carbon dioxide gas enricher, not only can the purpose of carbon dioxide gas enrichment be achieved to continuously supply mosquito-attracting gas, but the air separation system can also be greatly simplified and the system volume reduced, which makes it easier to miniaturize the entire intelligent mosquito-catching device and further facilitate its practical application and promotion.

[0073] Finally, it should be noted that the above description is merely a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A smart mosquito-catching device, characterized in that, It includes a mosquito-attracting chamber, an insect passage (2), a release passage (3), a mosquito-catching chamber (4), a passage switching mechanism, a camera (6), a controller (7), and a storage chamber (8). The inlet (21) of the insect passage (2) is connected to the mosquito-attracting chamber, the outlet of the insect passage (2) is connected to the mosquito-catching chamber (4), the inlet of the release passage (3) is connected to the insect passage (2) through the passage switching mechanism, and the outlet (32) of the release passage (3) is connected to the outside. The channel switching mechanism is communicatively connected to the controller (7) and is used to, under the control of the controller (7), close the outlet of the insect passage (2) and connect the inlet of the release channel (3) to the insect passage (2), or open the outlet of the insect passage (2) and close the inlet of the release channel (3); The camera (6) is communicatively connected to the controller (7) for real-time image acquisition of the inlet (21) of the insect passage (2) and real-time transmission of the acquired inlet monitoring image to the controller (7). The controller (7) is used to perform intelligent mosquito trapping according to the following steps: Upon receiving the imported monitoring image, the imported monitoring image is imported in real time into an insect recognition model that has been pre-trained based on a target recognition algorithm, and the insect recognition result is output. If the insect identification result contains an insect marker box, then the insect image is captured in real time from the imported monitoring image based on the insect marker box; The insect images are imported in real time into a mosquito recognition model pre-trained based on a first convolutional neural network, and the mosquito recognition results are output. If the mosquito identification result indicates a mosquito, the channel switching mechanism is controlled to open the outlet of the insect passage (2) and close the inlet of the release passage (3); otherwise, the channel switching mechanism is controlled to close the outlet of the insect passage (2) and connect the inlet of the release passage (3) to the insect passage (2). The mosquito-attracting chamber is equipped with a mosquito-attracting device, wherein the mosquito-attracting device is an exhaust port for mosquito-attracting gas; The storage chamber (8) is equipped with a carbon dioxide gas enrichment device and a gas mixing chamber (82). The carbon dioxide gas enrichment device includes an upper bottle (100), a lower bottle (200), a connecting pipe (300), a bottle body tilting mechanism (400), an air supply pump (500), and a high-temperature filter (600). The openings of the upper bottle (100) and the lower bottle (200) are vertically opposite each other and connected by the connecting pipe (300). The lower bottle (200) contains a carbon dioxide adsorbent solution. The bottle body tilting mechanism (400) is used to tilt the upper bottle (100), the lower bottle (200), and the connecting pipe (300) vertically. The bottle assembly structure is composed of the following: the output port of the air supply pump (500) is connected to the bottom of the inner cavity of the lower bottle (200) so as to introduce outside air into the carbon dioxide adsorbent solution, thereby causing the carbon dioxide gas in the air to react with the carbon dioxide adsorbent in the carbon dioxide adsorbent solution to form water-insoluble carbonate crystals; the high temperature filter (600) is arranged inside the connecting pipe (300) so as to intercept the carbonate crystals when the bottle assembly structure is flipped up and down, and decompose the carbonate crystals at high temperature to obtain carbon dioxide gas and carbon dioxide adsorbent that can fall to the bottom on its own; The inlet of the gas mixing chamber (82) is connected to the carbon dioxide gas enricher, and the outlet of the gas mixing chamber (82) is connected to the exhaust port of the mosquito-attracting gas.

2. The intelligent mosquito-catching device as described in claim 1, characterized in that, The mosquito-attracting chamber is surrounded by a cap (11), a bird-proof net (12), and a transverse partition (13) arranged from top to bottom. The upper surface of the transverse partition (13) is provided with the inlet (21) of the insect passage (2).

3. The intelligent mosquito-catching device as described in claim 2, characterized in that, The exhaust port of the mosquito-attracting gas is arranged on the wall of the support pipe (14), wherein the support pipe (14) is located inside the mosquito-attracting chamber, one end of the support pipe (14) is fixedly connected to the cap body (11), and the other end of the support pipe (14) is vertically fixedly connected to the upper surface of the transverse partition (13) so as to support the cap body (11). The outlet of the mixing chamber (82) is connected to the support pipe (14).

4. The intelligent mosquito-catching device as described in claim 2, characterized in that, The camera (6) is hung upside down inside the top of the cap (11) and the lens is aimed at the inlet (21) of the insect passage (2).

5. The intelligent mosquito-catching device as described in claim 1, characterized in that, The mosquito trapping chamber (4) has a ventilation window on its first side and a mosquito filter (42) is installed on the ventilation window. An exhaust fan (43) for exhausting indoor air is also installed outside the ventilation window. The controlled end of the exhaust fan (43) is communicatively connected to the controller (7). The controller (7) is also used to control the start / stop and speed adjustment of the exhaust fan (43).

6. The intelligent mosquito-catching device as described in claim 1, characterized in that, The second side of the mosquito trap (4) is equipped with a drawer (44) and / or a maintenance door (45) for opening the interior, wherein the drawer (44) is located at the bottom of the mosquito trap (4).

7. The intelligent mosquito-catching device as described in claim 1, characterized in that, The channel switching mechanism includes a stepper motor and a rotating flap (52), wherein the controlled end of the stepper motor is communicatively connected to the controller (7), and the rotating flap (52) is shaft-connected to the output shaft (510) of the stepper motor. The rotating flap (52) is used to close the outlet of the insect passage (2) and connect the inlet of the release channel (3) to the insect passage (2) when the output shaft (510) is rotated to the first position, and to open the outlet of the insect passage (2) and close the inlet of the release channel (3) when the output shaft (510) is rotated to the second position.

8. The intelligent mosquito-catching device as described in claim 1, characterized in that, The controller (7) is also used to upload the insect image and the collection time and / or collection location of the insect image to the cloud server in real time when the mosquito identification result indicates that it is a mosquito, so that the cloud server can import the insect image into a mosquito classification model pre-trained based on a second convolutional neural network, output the mosquito classification result, and perform statistical analysis based on the mosquito classification result and the collection time and / or the collection location to obtain a mosquito statistical analysis report.

9. The intelligent mosquito-catching device as described in claim 1, characterized in that, The camera (6) is a high-speed camera, and the controller (7) is a Raspberry Pi microcomputer.

Images