Mosquito and fly monitoring method and system based on food processing scene

Abstract

The invention relates to a mosquito and fly monitoring method and system based on a food processing scene, and belongs to the field of mosquito and fly monitoring. The method comprises the following steps: dividing a food processing scene into a non-clean area, a buffer area, a quasi-clean area and a clean area; constructing a mosquito and fly image recognition model; presetting a mosquito and fly monitoring period, acquiring a mosquito and fly monitoring image according to the mosquito and fly monitoring period, inputting the mosquito and fly monitoring image into a mosquito and fly image recognition model and outputting a mosquito and fly image recognition result, the mosquito and fly image recognition result including a mosquito and fly result and a non-mosquito and fly result, and counting the mosquito and fly results to obtain a mosquito and fly recognition number; threshold values are set according to regions, the mosquito and fly quantity level is judged, and mosquito and fly prevention measures are adopted for regions with many mosquitoes and flies according to local conditions. According to the method, regional treatment of mosquito and fly pollution of the food processing factory is achieved, the treatment means is scientific and easy to implement, and no threat is caused to food safety.

Description

Technical Field

[0001] The present invention belongs to the technical field of mosquito and fly monitoring, and in particular relates to a mosquito and fly monitoring method and system based on a food processing scenario. Background Art

[0002] Mosquitoes and flies are the spreaders of many diseases. They can carry various germs and viruses, such as Shigella, Escherichia coli, and Salmonella, and contaminate food, thus causing human illness. In addition, mosquitoes are also the vectors of many viruses and infectious diseases, such as malaria, dengue fever, filariasis, and yellow fever. In food processing plants, the activities of mosquitoes and flies not only cause food contamination, but also directly affect the quality and taste of food. They may lay eggs on food, causing food to spoil. At the same time, the excrement and vomitus of mosquitoes and flies will also contaminate food, causing it to lose its original flavor. The harm of mosquitoes and flies to food can also lead to economic losses. Due to food contamination, food may deteriorate and cannot be sold, resulting in economic losses for merchants. The current mosquito and fly control methods mostly use chemical means, which will cause certain harm to food and cause food quality to decline. There are no targeted mosquito and fly control measures for various areas of food processing plants, which is easy to cause waste or insufficient mosquito and fly control resources. Summary of the invention

[0003] In order to solve the above problems existing in the prior art, the present invention provides a mosquito and fly monitoring method and system based on a food processing scenario.

[0004] The purpose of the present invention can be achieved through the following technical solutions: A mosquito and fly monitoring method based on a food processing scenario, the implementation of the mosquito and fly monitoring method comprising the following steps: Divide the food processing scene into non-clean area, buffer area, quasi-clean area, and clean area; Build a mosquito and fly image recognition model; In the food processing scene, mosquito and fly killing lamps and sticky boards are arranged, a mosquito and fly monitoring cycle is preset, a mosquito and fly monitoring image of the sticky board is obtained according to the mosquito and fly monitoring cycle, the mosquito and fly monitoring image is input into the mosquito and fly image recognition model and a mosquito and fly image recognition result is output, the mosquito and fly image recognition result includes a mosquito and fly result and a non-mosquito and fly result, and the mosquito and fly results are counted to obtain a number of mosquito and fly identifications, and the number of mosquito and fly identifications includes a number of mosquito and fly identifications in a non-clean area, a number of mosquito and fly identifications in a buffer area, a number of mosquito and fly identifications in a quasi-clean area, and a number of mosquito and fly identifications in a clean area; A mosquito and fly threshold in a non-clean area is preset, and when the number of mosquitoes and flies identified in the non-clean area is greater than or equal to the mosquito and fly threshold in the non-clean area, an instruction to add a mosquito and fly killing lamp and an instruction to add a sticky board are output; A mosquito and fly threshold in the buffer zone is preset, and when the number of mosquito and flies identified in the buffer zone is greater than or equal to the mosquito and fly threshold in the buffer zone, an inert powder aerosol spraying instruction is output; A mosquito and fly threshold in the quasi-clean area is preset, and when the number of mosquito and flies identified in the quasi-clean area is greater than or equal to the mosquito and fly threshold in the quasi-clean area, a sonic wave expulsion instruction is output to expel mosquito and flies in the quasi-clean area; A mosquito and fly threshold in the clean area is preset, and when the number of mosquitoes and flies identified in the clean area is greater than or equal to the mosquito and fly threshold in the clean area, the sound wave expulsion instruction and the high-pressure airflow expulsion instruction are output to expel mosquitoes and flies in the clean area.

[0005] Preferably, the construction of the mosquito and fly image recognition model includes: Construct a standard image set of mosquitoes and flies; The mosquito and fly image recognition model is constructed through the GCHCNet network model.

[0006] Preferably, the mosquito and fly image recognition model is constructed by using the GCHCNet network model, including: The mosquito and fly monitoring image is scaled to a size of 224×224×3 and then input into the mosquito and fly image recognition model, and a 112×112×16 feature map is obtained through a 3×3 ordinary convolution layer; the 112×112×16 feature map passes through a convolution block attention mechanism and sequentially passes through G-bneck1, G-bneck2, G-bneck3, G-bneck4, and G-bneck5, and finally a 7×7×960 feature map is output through a 1×1 ordinary convolution layer; the 7×7×960 feature map is capsule-encoded into 8 groups of 32×3×3 capsules and dynamic routing is performed between layers, and finally the mosquito and fly image recognition result is output.

[0007] Preferably, the outputting of the sound wave expulsion instruction to expel mosquitoes and flies from the quasi-clean area comprises: The sonic mosquito repellent device arranged in the quasi-clean area receives the sonic mosquito repellent instruction and collects the real-time audio signal of the quasi-clean area; Framing and windowing the real-time audio signal of the quasi-clean area, and performing the windowing by using a Hanning window function; Outputting a mosquito and fly audio recognition result according to the real-time audio signal of the quasi-cleaning area, the mosquito and fly audio recognition result includes a mosquito and fly audio result and a non-mosquito and fly audio result; When the mosquito and fly audio result is output, the mosquito and fly expulsion in the quasi-clean area is performed according to the azimuth angle of the mosquito and fly sound source.

[0008] Preferably, the outputting of the mosquito and fly audio recognition result according to the real-time audio signal of the quasi-cleaning area comprises: Collecting a standard mosquito-fly audio signal and obtaining standard mosquito-fly acoustic features of the standard mosquito-fly audio signal, wherein the standard mosquito-fly acoustic features include a standard short-time average zero-crossing rate, a standard short-time average energy, a standard short-time average amplitude, and a standard Mel-frequency cepstrum coefficient; Acquiring the quasi-clean area acoustic features of the real-time audio signal of the quasi-clean area, wherein the quasi-clean area acoustic features include the short-time average zero-crossing rate of the quasi-clean area, the short-time average energy of the quasi-clean area, the short-time average amplitude of the quasi-clean area, and the Mel-frequency cepstrum coefficient of the quasi-clean area; The mosquito and fly audio recognition result is obtained by comparing the standard mosquito and fly acoustic characteristics with the quasi-clean area acoustic characteristics.

[0009] Preferably, the method of expelling mosquitoes and flies from the quasi-clean area by using the azimuth angle of the mosquito and fly sound source includes: Randomly select two of the acoustic mosquito repellent devices, and obtain the azimuth of the mosquito and fly sound source through the acoustic mosquito repellent devices; According to the azimuth angle of the mosquito and fly sound source, mosquito and fly repelling sound waves are emitted in the direction of the mosquito and fly sound source to expel mosquitoes and flies from the quasi-cleaning area.

[0010] Preferably, the high-pressure airflow expulsion instruction to expel mosquitoes and flies from the clean area includes: The terminal receives the high-pressure airflow expulsion instruction and turns on the high-pressure airflow device arranged at the exit of the clean area. The high-pressure airflow device absorbs the gas in the clean area and forms a high-pressure airflow to discharge the gas in the clean area out of the clean area.

[0011] A mosquito and fly monitoring system based on a food processing scenario, used to implement the mosquito and fly monitoring method described above, comprising a partitioning module, a mosquito and fly identification module, and a mosquito and fly expulsion module; The partition module is used to divide the food processing scene into a non-clean area, a buffer area, a quasi-clean area, and a clean area, and the buffer area includes a first buffer area and a second buffer area; The mosquito and fly identification module is used to construct a mosquito and fly image recognition model; mosquito and fly killing lamps and sticky boards are arranged in the food processing scene, a mosquito and fly monitoring cycle is preset, a mosquito and fly monitoring image of the sticky board is obtained according to the mosquito and fly monitoring cycle, the mosquito and fly monitoring image is input into the mosquito and fly image recognition model and a mosquito and fly image recognition result is output, the mosquito and fly image recognition result includes a mosquito and fly result and a non-mosquito and fly result, and the mosquito and fly results are counted to obtain a number of mosquito and fly identifications, and the number of mosquito and fly identifications includes a number of mosquito and fly identifications in a non-clean area, a number of mosquito and fly identifications in a buffer area, a number of mosquito and fly identifications in a quasi-clean area, and a number of mosquito and fly identifications in a clean area; The mosquito and fly expulsion module is used to preset a mosquito and fly threshold for a non-clean area. When the number of mosquito and flies identified in the non-clean area is greater than or equal to the mosquito and fly threshold for the non-clean area, an instruction to add a mosquito and fly killing lamp and an instruction to add a sticky board are output; a mosquito and fly threshold for a buffer area is preset. When the number of mosquito and flies identified in the buffer area is greater than or equal to the mosquito and fly threshold for the buffer area, an instruction to spray an inert powder aerosol is output; a mosquito and fly threshold for a quasi-clean area is preset. When the number of mosquito and flies identified in the quasi-clean area is greater than or equal to the mosquito and fly threshold for the quasi-clean area, a sound wave expulsion instruction is output to expel mosquito and flies from the quasi-clean area; a mosquito and fly threshold for a clean area is preset. When the number of mosquito and flies identified in the clean area is greater than or equal to the mosquito and fly threshold for the clean area, the sound wave expulsion instruction and the high-pressure airflow expulsion instruction are output to expel mosquito and flies from the clean area.

[0012] An electronic device comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the above-mentioned mosquito and fly monitoring method is implemented when the processor executes the program.

[0013] A storage medium containing computer executable instructions, wherein the computer executable instructions are used to execute the above-mentioned mosquito and fly monitoring method when executed by a computer processor.

[0014] The beneficial effects of the present invention are: (1) By testing the number of mosquitoes and flies in food processing plants in different regions, we can carry out regional control and achieve scientific, reasonable and safe mosquito and fly control in food processing scenes; (2) The number of mosquitoes and flies within a certain period can be determined through image recognition. The determination method is simple and intelligent recognition can be flexibly used to achieve fast and accurate mosquito and fly identification. (3) By crawling and enhancing the mosquito and fly standard image set, the obtained mosquito and fly standard image set is more realistic and conducive to subsequent image recognition; (4) Constructing a mosquito and fly image recognition model through the GCHCNet network model to improve the feature expression ability and generalization ability of the mosquito and fly image recognition model; (5) The sonic mosquito repellent device identifies the sound of mosquitoes and flies, determines whether there are mosquitoes and flies in the current scene, and locates the direction of mosquitoes and flies by the azimuth of the mosquito and fly sound source, thereby emitting mosquito and fly repellent sound waves to repel mosquitoes, thereby achieving accurate positioning of mosquitoes and flies. It is not necessary to emit mosquito and fly repellent sound waves in the entire area, which saves resources and is not likely to make workers feel uncomfortable. At the same time, physical sound waves will not contaminate food; (6) Mosquitoes and flies are expelled from the clean area through high-pressure airflow expulsion instructions, combined with sound wave expulsion, to effectively expel mosquitoes and flies from the clean area. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to facilitate understanding by those skilled in the art, the present invention is further described below with reference to the accompanying drawings.

[0016] Figure 1 The present invention is a flowchart of the steps of a mosquito and fly monitoring method based on a food processing scenario. DETAILED DESCRIPTION

[0017] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the specific implementation methods, structures, features and effects of the present invention are described in detail below in conjunction with the accompanying drawings and preferred embodiments.

[0018] The working principle and use process of the present invention: See also Figure 1 , a mosquito and fly monitoring method based on food processing scenarios, comprising: S1: Divide the food processing scene into a non-clean area, a buffer area, a quasi-clean area, and a clean area. The buffer area includes a first buffer area and a second buffer area. The first buffer area is the area between the non-clean area and the quasi-clean area. The second buffer area is the area between the quasi-clean area and the clean area. The non-clean area is an area that does not directly contact food. The quasi-clean area is an area such as a packaging area. The clean area is an area that directly contacts food. S2: Build a mosquito and fly image recognition model; S3: mosquito and fly killing lamps and sticky boards are arranged in the food processing scene, a mosquito and fly monitoring cycle is preset, a mosquito and fly monitoring image of the sticky board is obtained according to the mosquito and fly monitoring cycle, the mosquito and fly monitoring image is input into the mosquito and fly image recognition model and a mosquito and fly image recognition result is output, the mosquito and fly image recognition result includes a mosquito and fly result and a non-mosquito and fly result, and the mosquito and fly results are counted to obtain the number of mosquito and fly identifications, and the number of mosquito and fly identifications includes the number of mosquito and fly identifications in the non-clean area, the number of mosquito and fly identifications in the buffer area, the number of mosquito and fly identifications in the quasi-clean area, and the number of mosquito and fly identifications in the clean area; S4: Preset a mosquito and fly threshold in the unclean area, and when the number of mosquitoes and flies identified in the unclean area is greater than or equal to the mosquito and fly threshold in the unclean area, output an instruction to add a mosquito and fly killing lamp and an instruction to add a sticky board; S5: Preset a mosquito and fly threshold in the buffer zone, and when the number of mosquitoes and flies identified in the buffer zone is greater than or equal to the mosquito and fly threshold in the buffer zone, output an inert powder aerosol spraying instruction to spray food-grade inert powder aerosol into the buffer zone; S6: Preset a mosquito and fly threshold in the quasi-clean area, and when the number of mosquitoes and flies identified in the quasi-clean area is greater than or equal to the mosquito and fly threshold in the quasi-clean area, output a sonic wave expulsion instruction to expel mosquitoes and flies in the quasi-clean area; S7: Preset a mosquito and fly threshold in the clean area, and when the number of mosquitoes and flies identified in the clean area is greater than or equal to the mosquito and fly threshold in the clean area, output the sound wave expulsion instruction and the high-pressure airflow expulsion instruction to expel mosquitoes and flies in the clean area.

[0019] In this embodiment, the mosquito and fly image recognition model is constructed, which can be implemented by the following steps: S201: construct a standard image set of mosquitoes and flies; S202: Construct the mosquito and fly image recognition model through the GCHCNet network model.

[0020] In this embodiment, constructing a mosquito and fly standard image set can be implemented by the following steps: S201-1: Select mosquito and fly related website resources as crawling target websites through search engines; S201-2: crawling pictures in batches on the crawling target website using a picture crawling plug-in to obtain an initial set of mosquito and fly pictures; S201-3: Perform image enhancement processing on the initial mosquito and fly picture set to obtain the mosquito and fly standard image set. The image enhancement processing includes but is not limited to random rotation, random flipping, contrast processing, brightness processing, and color dithering processing. The obtained mosquito and fly standard image set is more realistic and is conducive to subsequent image recognition.

[0021] In this embodiment, the mosquito and fly image recognition model is constructed by using the GCHCNet network model, which can be implemented by the following steps: A capsule network based on a dynamic routing algorithm is embedded in a phantom network; a 3×3 ordinary convolution layer is constructed; a convolution block attention mechanism is embedded in the phantom network to strengthen the connection of contextual semantic relationships; the ReLU activation function in the phantom module is replaced by a Hardswish activation function to obtain an improved Ghost bottleneck structure, the improved Ghost bottleneck structure includes G-bneck1, G-bneck2, G-bneck3, G-bneck4, and G-bneck5, and the improved Ghost bottleneck structure is composed of a Ghost bottleneck structure with a step size of 1 and a Ghost bottleneck structure with a step size of 2, which is used to improve the feature expression ability and generalization ability of the mosquito and fly image recognition model; a 1×1 ordinary convolution layer is constructed; the mosquito and fly standard image set is divided into a verification set and a test set, the mosquito and fly image recognition model is verified and tested, and the division ratio of the verification set and the test set is 7:3.

[0022] In this embodiment, the mosquito and fly monitoring image is input into the mosquito and fly image recognition model and the mosquito and fly image recognition result is output, which can be specifically implemented by the following steps: The mosquito and fly monitoring image is scaled to a size of 224×224×3 and then input into the mosquito and fly image recognition model, and a 112×112×16 feature map is obtained through a 3×3 ordinary convolution layer; the 112×112×16 feature map passes through a convolution block attention mechanism and sequentially passes through G-bneck1, G-bneck2, G-bneck3, G-bneck4, and G-bneck5, and finally a 7×7×960 feature map is output through a 1×1 ordinary convolution layer; the 7×7×960 feature map is capsule-encoded into 8 groups of 32×3×3 capsules and dynamic routing is performed between layers, and finally the mosquito and fly image recognition result is output.

[0023] In this embodiment, outputting a sonic wave expulsion instruction to expel mosquitoes and flies from the quasi-clean area can be specifically implemented through the following steps: S601: The sonic mosquito repellent device arranged in the quasi-clean area receives the sonic expulsion instruction and collects the real-time audio signal of the quasi-clean area, wherein the sonic mosquito repellent device includes a microphone and a sound generating device, etc.; S602: Frame and window the real-time audio signal of the quasi-clean area, and perform the windowing using a Hanning window function. The expression of the Hanning window function is: , where w(n) is the Hanning window function, n is the nth sample point in the Hanning window function, and N is the total length of the Hanning window function; S603: outputting a mosquito and fly audio recognition result according to the real-time audio signal of the quasi-cleaning area, wherein the mosquito and fly audio recognition result includes a mosquito and fly audio result and a non-mosquito and fly audio result; S604: When the mosquito and fly audio result is output, mosquito and fly are expelled from the quasi-clean area according to the mosquito and fly sound source azimuth.

[0024] In this embodiment, outputting the mosquito and fly audio recognition result according to the real-time audio signal of the quasi-cleaning area can be implemented by the following steps: S603-1: Collect a standard mosquito-fly audio signal and obtain standard mosquito-fly acoustic features of the standard mosquito-fly audio signal, wherein the standard mosquito-fly acoustic features include a standard short-time average zero-crossing rate, a standard short-time average energy, a standard short-time average amplitude, and a standard Mel-frequency cepstrum coefficient; S603-2: Acquire the quasi-clean area acoustic features of the real-time audio signal of the quasi-clean area, wherein the quasi-clean area acoustic features include the short-time average zero-crossing rate of the quasi-clean area, the short-time average energy of the quasi-clean area, the short-time average amplitude of the quasi-clean area, and the Mel-frequency cepstrum coefficient of the quasi-clean area. The expression of the short-time average zero-crossing rate of the quasi-clean area is: , where Z nis the short-time average zero-crossing rate of the quasi-clean area, n is the nth sample point in the Hanning window function, N is the total length of the Hanning window function, x(n) is the real-time audio signal in the quasi-clean area, w(n) is the Hanning window function, and the expression of the short-time average energy of the quasi-clean area is: , where E n is the short-time average energy of the quasi-clean area, and the expression of the short-time average amplitude of the quasi-clean area is: , where M n is the short-time average amplitude of the quasi-clean area, and the expression of the Mel frequency cepstrum coefficient of the quasi-clean area is: , where m is the Mel frequency cepstrum coefficient of the quasi-cleaning area, f Real-time audio signal frequency of the quasi-cleaning area; S603-3: The mosquito and fly audio recognition result is obtained by comparing the standard mosquito and fly acoustic characteristics with the quasi-clean area acoustic characteristics, and a preset overlap threshold is set. When the overlap between the standard mosquito and fly acoustic characteristics and the quasi-clean area acoustic characteristics is greater than or equal to the overlap threshold, the mosquito and fly audio result is output; when the overlap between the standard mosquito and fly acoustic characteristics and the quasi-clean area acoustic characteristics is less than the overlap threshold, the non-mosquito and fly audio result is output.

[0025] In this embodiment, the mosquito and fly expulsion in the quasi-cleaning area is performed according to the azimuth angle of the mosquito and fly sound source, which can be specifically implemented by the following steps: S604-1: Randomly select two of the acoustic mosquito repellent devices, and obtain the azimuth of the mosquito and fly sound source through the acoustic mosquito repellent devices; Obtain the arrival time difference of the sound source to the acoustic mosquito repellent device; calculate the azimuth of the mosquito and fly sound source according to the arrival time difference, and the calculation formula is: , where θ is the azimuth of the mosquito and fly sound source, t is the arrival time difference, v is the speed of light, and D is the distance between the two acoustic mosquito repellent devices; S604-2: Emitting mosquito and fly repellent sound waves in the direction of the mosquito and fly sound source according to the azimuth angle of the mosquito and fly sound source to expel mosquitoes and flies in the quasi-clean area, judging whether there are mosquitoes and flies in the current scene, and locating the direction of mosquitoes and flies by the azimuth angle of the mosquito and fly sound source, thereby emitting mosquito and fly repellent sound waves to repel mosquitoes, thereby accurately locating mosquitoes and flies, and there is no need to emit mosquito and fly repellent sound waves in the entire area, thus saving resources and not making workers feel uncomfortable. At the same time, physical sound waves will not contaminate food.

[0026] In this embodiment, the high-pressure airflow expulsion instruction is used to expel mosquitoes and flies from the clean area, which can be specifically implemented by the following steps: S701: The sonic mosquito repellent device arranged in the cleaning area receives the sonic expulsion instruction and collects the real-time audio signal of the cleaning area; frames and windows the real-time audio signal of the cleaning area; outputs the mosquito and fly audio recognition result according to the real-time audio signal of the cleaning area; when the mosquito and fly audio result is output, the mosquito and fly in the cleaning area are expelled according to the azimuth of the mosquito and fly sound source; S702: The terminal receives the high-pressure airflow expulsion instruction and turns on the high-pressure airflow device arranged at the exit of the clean area. The high-pressure airflow device absorbs the gas in the clean area and forms a high-pressure airflow to discharge the gas in the clean area out of the clean area. Combined with the sonic expulsion, mosquitoes and flies in the clean area are effectively expelled.

[0027] A mosquito and fly monitoring system based on food processing scenarios, including a partitioning module, a mosquito and fly identification module, and a mosquito and fly expulsion module; The partition module is used to divide the food processing scene into a non-clean area, a buffer area, a quasi-clean area, and a clean area, and the buffer area includes a first buffer area and a second buffer area; The mosquito and fly identification module is used to construct a mosquito and fly image recognition model; mosquito and fly killing lamps and sticky boards are arranged in the food processing scene, a mosquito and fly monitoring cycle is preset, a mosquito and fly monitoring image of the sticky board is obtained according to the mosquito and fly monitoring cycle, the mosquito and fly monitoring image is input into the mosquito and fly image recognition model and a mosquito and fly image recognition result is output, the mosquito and fly image recognition result includes a mosquito and fly result and a non-mosquito and fly result, and the mosquito and fly results are counted to obtain a number of mosquito and fly identifications, and the number of mosquito and fly identifications includes a number of mosquito and fly identifications in a non-clean area, a number of mosquito and fly identifications in a buffer area, a number of mosquito and fly identifications in a quasi-clean area, and a number of mosquito and fly identifications in a clean area; The mosquito and fly expulsion module is used to preset a mosquito and fly threshold for a non-clean area. When the number of mosquito and flies identified in the non-clean area is greater than or equal to the mosquito and fly threshold for the non-clean area, an instruction to add a mosquito and fly killing lamp and an instruction to add a sticky board are output; a mosquito and fly threshold for a buffer area is preset. When the number of mosquito and flies identified in the buffer area is greater than or equal to the mosquito and fly threshold for the buffer area, an instruction to spray an inert powder aerosol is output; a mosquito and fly threshold for a quasi-clean area is preset. When the number of mosquito and flies identified in the quasi-clean area is greater than or equal to the mosquito and fly threshold for the quasi-clean area, a sound wave expulsion instruction is output to expel mosquito and flies from the quasi-clean area; a mosquito and fly threshold for a clean area is preset. When the number of mosquito and flies identified in the clean area is greater than or equal to the mosquito and fly threshold for the clean area, the sound wave expulsion instruction and the high-pressure airflow expulsion instruction are output to expel mosquito and flies from the clean area.

[0028] The computer storage medium of the embodiment of the present invention may adopt any combination of one or more computer-readable media. The computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples (non-exhaustive list) of computer-readable storage media include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this document, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in combination with an instruction execution system, device or device.

[0029] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, which carry computer-readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0030] The program code included on the computer readable medium can be transmitted with any appropriate medium, including but not limited to wireless, electric wire, optical cable, RF, etc., or any suitable combination of the above. The computer program code for performing the operation of the present invention can be written in one or more programming languages ​​or their combinations, and the programming language includes object-oriented programming languages-such as Java, Smalltalk, C++, and also includes conventional procedural programming languages-such as "C" language or similar programming languages. The program code can be executed completely on the user's computer, partially on the user's computer, as an independent software package, partially on the user's computer and partially on the remote computer, or completely on the remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computer (for example, using an Internet service provider to connect through the Internet).

[0031] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any form. Although the present invention has been disclosed as a preferred embodiment as above, it is not used to limit the present invention. Any technical personnel in this field can make some changes or modify the technical contents disclosed above into equivalent embodiments without departing from the scope of the technical solution of the present invention. However, any simple modification, equivalent change and modification made to the above embodiments according to the technical essence of the present invention without departing from the content of the technical solution of the present invention still fall within the scope of the technical solution of the present invention.

Claims

1. A mosquito and fly monitoring method based on food processing scenarios, characterized in that: The implementation of the mosquito and fly monitoring method comprises the following steps: Divide the food processing scene into non-clean area, buffer area, quasi-clean area, and clean area; Build a mosquito and fly image recognition model; In the food processing scene, mosquito and fly killing lamps and sticky boards are arranged, a mosquito and fly monitoring cycle is preset, a mosquito and fly monitoring image of the sticky board is obtained according to the mosquito and fly monitoring cycle, the mosquito and fly monitoring image is input into the mosquito and fly image recognition model and a mosquito and fly image recognition result is output, the mosquito and fly image recognition result includes a mosquito and fly result and a non-mosquito and fly result, and the mosquito and fly results are counted to obtain a number of mosquito and fly identifications, and the number of mosquito and fly identifications includes a number of mosquito and fly identifications in a non-clean area, a number of mosquito and fly identifications in a buffer area, a number of mosquito and fly identifications in a quasi-clean area, and a number of mosquito and fly identifications in a clean area; A mosquito and fly threshold in a non-clean area is preset, and when the number of mosquitoes and flies identified in the non-clean area is greater than or equal to the mosquito and fly threshold in the non-clean area, an instruction to add a mosquito and fly killing lamp and an instruction to add a sticky board are output; A mosquito and fly threshold in the buffer zone is preset, and when the number of mosquito and flies identified in the buffer zone is greater than or equal to the mosquito and fly threshold in the buffer zone, an inert powder aerosol spraying instruction is output; A mosquito and fly threshold in the quasi-clean area is preset, and when the number of mosquito and flies identified in the quasi-clean area is greater than or equal to the mosquito and fly threshold in the quasi-clean area, a sonic wave expulsion instruction is output to expel mosquito and flies in the quasi-clean area; A mosquito and fly threshold in the clean area is preset, and when the number of mosquitoes and flies identified in the clean area is greater than or equal to the mosquito and fly threshold in the clean area, the sound wave expulsion instruction and the high-pressure airflow expulsion instruction are output to expel mosquitoes and flies in the clean area.

2. The method for monitoring mosquitoes and flies according to claim 1, characterized in that: The construction of the mosquito and fly image recognition model comprises: Construct a standard image set of mosquitoes and flies; The mosquito and fly image recognition model is constructed through the GCHCNet network model.

3. The mosquito and fly monitoring method according to claim 2, characterized in that: The mosquito and fly image recognition model constructed by the GCHCNet network model includes: The mosquito and fly monitoring image is scaled to a size of 224×224×3 and then input into the mosquito and fly image recognition model, and a 112×112×16 feature map is obtained through a 3×3 ordinary convolution layer; the 112×112×16 feature map passes through a convolution block attention mechanism and sequentially passes through G-bneck1, G-bneck2, G-bneck3, G-bneck4, and G-bneck5, and finally a 7×7×960 feature map is output through a 1×1 ordinary convolution layer; the 7×7×960 feature map is capsule-encoded into 8 groups of 32×3×3 capsules and dynamic routing is performed between layers, and finally the mosquito and fly image recognition result is output.

4. The method for monitoring mosquitoes and flies according to claim 1, characterized in that: The output of the sound wave expulsion instruction to expel mosquitoes and flies from the quasi-clean area includes: The sonic mosquito repellent device arranged in the quasi-clean area receives the sonic mosquito repellent instruction and collects the real-time audio signal of the quasi-clean area; Framing and windowing the real-time audio signal of the quasi-clean area, and performing the windowing by using a Hanning window function; Outputting a mosquito and fly audio recognition result according to the real-time audio signal of the quasi-cleaning area, the mosquito and fly audio recognition result includes a mosquito and fly audio result and a non-mosquito and fly audio result; When the mosquito and fly audio result is output, the mosquito and fly expulsion in the quasi-clean area is performed according to the azimuth angle of the mosquito and fly sound source.

5. The method for monitoring mosquitoes and flies according to claim 4, characterized in that: Outputting the mosquito and fly audio recognition result according to the real-time audio signal of the quasi-cleaning area includes: Collecting a standard mosquito-fly audio signal and obtaining standard mosquito-fly acoustic features of the standard mosquito-fly audio signal, wherein the standard mosquito-fly acoustic features include a standard short-time average zero-crossing rate, a standard short-time average energy, a standard short-time average amplitude, and a standard Mel-frequency cepstrum coefficient; Acquiring the quasi-clean area acoustic features of the real-time audio signal of the quasi-clean area, wherein the quasi-clean area acoustic features include the short-time average zero-crossing rate of the quasi-clean area, the short-time average energy of the quasi-clean area, the short-time average amplitude of the quasi-clean area, and the Mel-frequency cepstrum coefficient of the quasi-clean area; The mosquito and fly audio recognition result is obtained by comparing the standard mosquito and fly acoustic characteristics with the quasi-clean area acoustic characteristics.

6. The method for monitoring mosquitoes and flies according to claim 4, characterized in that: The method of expelling mosquitoes and flies from the quasi-cleaning area by using the azimuth of the mosquito and fly sound source comprises: Randomly select two of the acoustic mosquito repellent devices, and obtain the azimuth of the mosquito and fly sound source through the acoustic mosquito repellent devices; According to the azimuth angle of the mosquito and fly sound source, mosquito and fly repelling sound waves are emitted in the direction of the mosquito and fly sound source to expel mosquitoes and flies from the quasi-cleaning area.

7. The method for monitoring mosquitoes and flies according to claim 1, characterized in that: The high-pressure airflow expulsion instruction for expelling mosquitoes and flies from the clean area includes: The terminal receives the high-pressure airflow expulsion instruction and turns on the high-pressure airflow device arranged at the exit of the clean area. The high-pressure airflow device absorbs the gas in the clean area and forms a high-pressure airflow to discharge the gas in the clean area out of the clean area.

8. A mosquito and fly monitoring system based on food processing scenarios, characterized in that: The system is applied to the mosquito and fly monitoring method as described in any one of claims 1 to 7, including a partitioning module, a mosquito and fly identification module, and a mosquito and fly expulsion module; The partition module is used to divide the food processing scene into a non-clean area, a buffer area, a quasi-clean area, and a clean area, and the buffer area includes a first buffer area and a second buffer area; The mosquito and fly identification module is used to construct a mosquito and fly image recognition model; mosquito and fly killing lamps and sticky boards are arranged in the food processing scene, a mosquito and fly monitoring cycle is preset, a mosquito and fly monitoring image of the sticky board is obtained according to the mosquito and fly monitoring cycle, the mosquito and fly monitoring image is input into the mosquito and fly image recognition model and a mosquito and fly image recognition result is output, the mosquito and fly image recognition result includes a mosquito and fly result and a non-mosquito and fly result, and the mosquito and fly results are counted to obtain a number of mosquito and fly identifications, and the number of mosquito and fly identifications includes a number of mosquito and fly identifications in a non-clean area, a number of mosquito and fly identifications in a buffer area, a number of mosquito and fly identifications in a quasi-clean area, and a number of mosquito and fly identifications in a clean area; The mosquito and fly expelling module is used to preset a mosquito and fly threshold in a non-clean area, and when the number of mosquito and flies identified in the non-clean area is greater than or equal to the mosquito and fly threshold in the non-clean area, an instruction to add a mosquito and fly killing lamp and an instruction to add a sticky board are output; A mosquito and fly threshold in the buffer zone is preset, and when the number of mosquito and flies identified in the buffer zone is greater than or equal to the mosquito and fly threshold in the buffer zone, an inert powder aerosol spraying instruction is output; a mosquito and fly threshold in the quasi-clean zone is preset, and when the number of mosquito and flies identified in the quasi-clean zone is greater than or equal to the mosquito and fly threshold in the quasi-clean zone, a sonic wave expulsion instruction is output to expel mosquito and flies in the quasi-clean zone; A mosquito and fly threshold in the clean area is preset, and when the number of mosquitoes and flies identified in the clean area is greater than or equal to the mosquito and fly threshold in the clean area, the sound wave expulsion instruction and the high-pressure airflow expulsion instruction are output to expel mosquitoes and flies in the clean area.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the mosquito and fly monitoring method as described in any one of claims 1-7 is implemented.

10. A storage medium containing computer executable instructions, characterized in that: The computer executable instructions are used to execute the mosquito and fly monitoring method as described in any one of claims 1 to 7 when executed by a computer processor.

Images