Rodent monitoring system and method

Abstract

The invention relates to a rodent monitoring system which comprises a first sensor unit, a second sensor unit, a master control edge computing unit and a cloud server. The infrared sensor and the pressure sensor are in signal connection with the main control edge calculation module, and the cloud server is configured to be used for receiving the image information, the weight information and the body length information in real time. The rodent information can be monitored in real time, dynamic information such as types, distribution and quantity of rodents in a monitored target area can be remotely mastered, the rodent monitoring system has the advantages of being unattended, reducing manual investigation cost, achieving all-weather monitoring and the like, and the growth conditions of the rodents in part of areas can be analyzed more accurately.

Description

Technical Field

[0001] The invention relates to the technical field of rodent prevention and control, and in particular to a rodent monitoring system and method. Background Art

[0002] my country has a vast territory and a diverse and complex ecological environment, which leads to a large number of rodent species. The current rodent monitoring methods lack the technical means and capabilities for large-scale monitoring, cannot effectively and uniformly analyze monitoring data, and are difficult to establish an effective dynamic prediction model for rodent pests.

[0003] Traditional methods for monitoring rat density include the trap (cage) day method, sticky mouse board method, path method, powder method, etc., which have the following shortcomings and problems: lack of accuracy, extensiveness and big data, and it is difficult to form a prediction and forecasting model based on big data; lack of safety and timeliness, backward monitoring methods, low monitoring efficiency, high cost and difficult management. Summary of the invention

[0004] In view of the above analysis, the object of the present invention is to provide a rodent monitoring system and method to solve the technical problems thereof.

[0005] The object of the present invention is achieved in that:

[0006] A rodent monitoring system includes a first sensor unit, a second sensor unit, a master edge computing module, and a cloud server;

[0007] The first sensor unit is configured to perform real-time detection on the monitoring channel;

[0008] The second sensor unit is configured to monitor the weight data of the monitoring channel in real time;

[0009] The master edge computing module is configured to fuse the data collected from the first sensor unit and the second sensor unit, and based thereon determine whether the object in the monitoring channel is a rodent;

[0010] The cloud server is configured to send control instructions to the master edge computing module, manage and configure the operating parameters of the master edge computing module, receive the data uploaded by the master edge computing module and the preliminary analysis results, perform global data management and advanced analysis, and send the results or new control instructions back to the master edge computing module;

[0011] The master edge computing module is connected to the cloud server via a wireless communication circuit or a Beidou positioning circuit.

[0012] Furthermore, the main control edge computing module is also configured to determine that an object has invaded the monitoring channel when the monitoring value collected by the first sensor unit at the current moment exceeds a set first threshold, and then the main control edge computing module sends a closing control instruction to the automatic door in real time to close the automatic door.

[0013] Furthermore, the main control edge computing module is also configured to send a door closing command when the monitoring value of the monitoring channel collected by the first sensor unit (11) at the current moment exceeds a set first threshold, and the weight value collected by the second sensor unit (12) is greater than a set second threshold.

[0014] Furthermore, the main control edge computing module is also configured that when the monitoring value of the monitoring channel collected by the first sensor unit at the current moment exceeds the set first threshold, and the weight value collected by the second sensor unit is greater than the set second threshold, after sending the door closing command, if the weight value of the second sensor is continuously monitored to be less than the second threshold, then a door opening command is sent, and it is judged that the object has escaped. At the same time, the second sensor unit collects the remaining weight of the bait box and sends the remaining weight data to the main control edge computing module, the main control edge computing module preliminarily processes and analyzes it, and sends the results and data or instructions that need further processing to the cloud server.

[0015] Furthermore, it also includes a third sensor unit, which is configured to collect length data of objects entering the monitoring channel and send the object length data to the main control edge computing module. The main control edge computing module preliminarily processes and analyzes the data and sends the results and data or instructions that need further processing to the cloud server.

[0016] Furthermore, the master edge computing module is further configured to send a command to start the third sensor unit if the weight value of the second sensor unit is continuously monitored to be greater than the second threshold value when the weight information of the second sensor unit received is greater than the set second threshold value and a door closing command is sent, and the third sensor unit measures the length of the object. When the length measurement of the object is completed, a door opening command is sent, and a third sensor closing command is sent at the same time;

[0017] And / or when the weight information received by the second sensor unit is greater than the set second threshold and a door closing command is sent, if the weight value of the second sensor is continuously monitored to be less than the second threshold, a door opening command is sent, and the target object is judged to have escaped. At the same time, the second sensor unit collects the remaining weight of the bait box and sends the remaining weight data to the main control edge computing module, which preliminarily processes and analyzes it and sends the results and data or instructions that need further processing to the cloud server.

[0018] Furthermore, it also includes a fourth sensor unit, which is configured to collect image data of objects entering the monitoring channel and send the object image data to the main control edge computing module. The main control edge computing module preliminarily processes and analyzes the data and sends the results and data or instructions that need further processing to the cloud server.

[0019] Specifically, the master edge computing module is further configured to: when the weight information of the second sensor unit received is greater than the set second threshold value and a door closing instruction is sent, if the weight value of the second sensor unit is continuously monitored to be greater than the second threshold value, send an instruction to start the fourth sensor, the fourth sensor unit performs image acquisition on the target object, and synchronously measures the weight of the object, and when the image acquisition of the target object is completed, send a door opening instruction, and send a command to close the fourth sensor unit at the same time;

[0020] And / or when the weight information received by the second sensor unit is greater than the set second threshold and a door closing command is sent, if the weight value of the second sensor unit is continuously monitored to be less than the second threshold, a door opening command is sent, and the target object is judged to have escaped. At the same time, the second sensor unit collects the remaining weight of the bait box and sends the remaining weight data to the main control edge computing module, which preliminarily processes and analyzes it and sends the results and data or instructions that need further processing to the cloud server.

[0021] Furthermore, it also includes a coprocessor module, and the coprocessor module is configured to: pre-process the image information collected by the fourth sensor unit;

[0022] The master edge computing unit is connected to the coprocessor unit by signal, and is used to perform non-uniform correction on the image information according to the deep learning algorithm; smooth the corrected image using the filtering algorithm; separate the target from the background information in the smoothed image according to the background difference method, and detect the moving target; and visually track the moving target using the tracking algorithm;

[0023] The preprocessing includes graying, temperature conversion, filtering and noise reduction of the image information collected by the infrared camera.

[0024] A rodent monitoring method comprises the following steps:

[0025] The first sensor unit detects the target environment in real time. When the infrared radiation information of the target environment detected by the first sensor unit is greater than the set first threshold, it is determined that an object has invaded the target environment; when the weight information fed back by the second sensor unit is greater than the set second threshold, a door closing command is sent, and the electric door closes within 50 milliseconds; when the weight information fed back by the second sensor unit is less than the set second threshold, no door closing command is sent, and the electric door always remains in the open state;

[0026] After the electric door is closed, if the second sensor unit continuously feeds back weight information greater than the second threshold, the main control edge computing module sends an instruction to start the third sensor unit and / or the fourth sensor unit to measure the length and / or capture images of the object entering, and synchronously collect the weight of the object. After the length information, image information and weight information are collected, the main control edge computing module sends a door opening instruction, and the electric door opens. At the same time, the main control edge computing module sends an instruction to close the third sensor unit and / or the fourth sensor unit. When the weight information of the second sensor is less than the set second threshold, it is judged that the object has left, and the remaining weight of the bait in the bait box is synchronously collected; if the second sensor continuously feeds back weight information less than the second threshold, it is judged that the object has escaped, and the main control edge computing module sends a door opening instruction, and the automatic door opens;

[0027] The cloud server analyzes and processes the data sent by the master edge computing module, and classifies them into four categories: invalid trigger, foreign object trigger, non-target object, and target trigger;

[0028] If it is determined to be an invalid trigger, the invalid trigger data will be directly deleted;

[0029] If it is determined to be triggered by foreign objects and non-target objects, the data triggered by foreign objects and non-target objects are analyzed and processed and then classified and stored;

[0030] If it is determined to be a target trigger, the data and images of the entry and exit trajectory, weight, body length, and food theft weight generated by the target trigger will be analyzed and processed again to generate relevant charts of daily activity frequency, activity intensity, weight, body length, food theft amount, quantity, and images and store them in the big database.

[0031] Daily activity frequency = activity trajectory every 24 hours ÷ 24 hours, daily activity intensity = daily activity frequency ÷ 24 hours, number of objects = the number of times an object is captured every 24 hours.

[0032] Beneficial effects of the present invention:

[0033] 1. The present invention avoids the influence of human factors on the activities of rodents and thus the influence on the authenticity of data. The rat data is obtained through the monitoring device and the data is recorded in real time, which improves the timeliness of monitoring and further improves the authenticity of the monitoring data.

[0034] 2. The present invention can monitor the activities of rats in real time by setting up an infrared camera, the radar rangefinder can measure the actual size of rats, the pressure sensor can monitor the weight of rats, and the monitoring data can be transmitted to the server wirelessly, so as to remotely grasp the dynamic information such as the species, distribution and number of rats in the monitoring target area. It has the advantages of unmanned operation, reduced manual investigation costs, and all-weather monitoring, and can more accurately analyze the growth status of rats in some areas.

[0035] 3. The present invention can capture live mice, and can release them after testing the data of live mice as needed, and continue to monitor them. When they are not needed to be released, the live mice can be stored for later use. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Exemplary embodiments are described below with reference to the following drawings.

[0037] Figure 1 To schematically and exemplarily illustrate an embodiment of a monitoring system according to the present disclosure;

[0038] Figure 2 The figure schematically and exemplarily shows the overall structure of the monitoring device according to the present disclosure;

[0039] Figure 3 The cross-sectional structure diagram of the monitoring device according to the present disclosure is schematically and exemplarily shown;

[0040] Figure 4 It is a schematic diagram of a line graph of the amount of food stolen by rodents in the monitoring area of ​​the present invention;

[0041] Figure 5 It is a schematic diagram of a line graph of the frequency of random rodent activities within the monitoring area of ​​the present invention for 24 hours;

[0042] Figure 6 It is a line graph for analyzing the intensity of rodent activity in the monitoring area of ​​the present invention.

[0043] Figure: 1. Monitoring system; 2. Main control edge computing unit; 10. Automatic door; 11. First sensor unit; 12. Second sensor unit; 13. Third sensor unit; 14. Fourth sensor unit. DETAILED DESCRIPTION

[0044] The specific implementation of the present invention is described below in conjunction with the accompanying drawings and embodiments:

[0045] The present invention is further described in detail below through specific examples. The following examples are only illustrative and not restrictive, and the protection scope of the present invention cannot be limited thereto.

[0046] Figure 1A monitoring system is shown, a rodent monitoring system includes a first sensor unit 11, a second sensor unit 12, a master edge computing module and a cloud server. The first sensor unit 11 and the second sensor unit 12 are different types of sensor units. The first sensor unit 11 can be an infrared sensor or a photoelectric sensor, and the second sensor unit 12 can be a weighing sensor or a pressure sensor.

[0047] Figure 2 and Figure 3 The specific structure of the monitoring device is shown. The monitoring device 1 is provided with a control chamber 101 on the upper part and a monitoring channel 102 on the lower part. The monitoring channel 102 is penetrated. Automatic doors 103 are respectively provided at both ends of the monitoring channel 102. A first sensor unit 11 is installed at the two ends of the monitoring channel 102. A second sensor unit 12 is provided inside the monitoring channel 102. The second sensor unit 12 is arranged in the control chamber 101. A pressure plate 104 is provided on the top of the second sensor unit 12. A carrier 105 is vertically connected to both sides of the pressure plate 104. The carrier 105 is vertically arranged in the monitoring channel 102. A gap is provided between the carrier 105 and the bottom wall of the monitoring channel 102. A movable bait box 106 is provided on the carrier 105. A third sensor unit 13 and a third sensor unit 14 are provided on the top of the monitoring channel 102.

[0048] The first sensor unit 11 is configured to perform real-time detection on the monitoring channel 102. The first sensor unit 11 may be an infrared sensor, which is disposed at the entrance of the monitoring channel 102 to detect in real time whether an object enters the monitoring channel 102 and transmit data to the main control edge computing module.

[0049] The second sensor unit 12 is configured to monitor the weight data of the monitoring channel in real time; the second sensor unit 12 can be a weighing sensor, which is arranged in the monitoring channel 102, for detecting the weight change of the monitoring channel 102 in real time and transmitting the data to the main control edge computing module.

[0050] The master edge computing module is configured to fuse the data collected from the first sensor unit 11 and the second sensor unit 12, and based on this, determine whether the object in the monitoring channel is a rodent; the master edge computing module determines whether the object in the monitoring channel is a leaf, stone, mouse, hare and other rodents, snakes and other reptiles, and other types of animals based on the collected infrared sensor data and weighing sensor data.

[0051] The cloud server is configured to send control instructions to the master edge computing module, manage and configure the operating parameters of the master edge computing module, receive the data uploaded by the master edge computing module and the preliminary analysis results, perform global data management and advanced analysis, and send the results or new control instructions back to the master edge computing module;

[0052] The master edge computing module is connected to the cloud server via a wireless communication circuit or a Beidou positioning circuit.

[0053] The wireless communication circuit is based on the NB-IoT communication mode. Through the NB-IoT base station and application server, a "grid" management mode is adopted to realize the monitoring and management of the big data cloud platform for monitoring data and geographic information. For example, the wireless communication circuit adopts the NB86-G low-power multi-band NB-IoT wireless communication module, which is based on the HiSilicon Boudica150 chipset and meets the needs of most operators in the world. The NB-IoT wireless communication mode is adopted to establish a network connection with the user end through the NB-IoT base station and application server to provide data transmission services for monitoring data and geographic information.

[0054] Beidou circuit can start short message communication mode, taking advantage of the characteristics of Beidou system with both positioning and communication functions, without the support of other communication systems, using Beidou global short message communication mode, and sending relevant data back through satellite to achieve uninterrupted data monitoring. For example, Beidou positioning circuit 7 uses Beidou RDSS four-in-one module HS201, with a receiving sensitivity of -127dBm.

[0055] Positioning and communication success rate is greater than 95%

[0056] Specifically, the main control edge computing module is also configured such that if the monitoring value of the monitoring channel collected by the first sensor unit 11 at the current moment exceeds the set first threshold, it is judged that an object has invaded the monitoring channel, and then the main control edge computing module 3 sends a closing control instruction to the automatic door in real time to close the automatic door.

[0057] Compare the monitoring value of the monitoring channel at the current moment with the set first threshold, obtain the comparison result, and obtain the automatic door closing instruction according to whether the comparison result meets the preset requirements. The monitoring value is the first sensor unit value collected at the monitoring channel at the current moment, and the first threshold is a pre-set reference value.

[0058] Specifically, the main control edge computing module is also configured to send a door closing command when the monitoring value of the monitoring channel collected by the first sensor unit 11 at the current moment exceeds a set first threshold, and the weight value collected by the second sensor unit 12 is greater than a set second threshold.

[0059] When the first sensor unit 11 is triggered, the second sensor unit 12 is turned on and on standby. When the second sensor unit 12 is triggered, the automatic door is on standby. When a predetermined second threshold is reached, the automatic door is closed.

[0060] Specifically, the main control edge computing module is also configured that when the monitoring value of the monitoring channel collected by the first sensor unit 11 at the current moment exceeds the set first threshold, and the weight value collected by the second sensor unit 12 is greater than the set second threshold, after sending the door closing command, if the weight value of the second sensor is continuously monitored to be less than the second threshold, then a door opening command is sent, and it is judged that the object has escaped. At the same time, the second sensor unit 12 collects the remaining weight of the bait box and sends the remaining weight data to the main control edge computing module, the main control edge computing module preliminarily processes and analyzes it, and sends the results and data or instructions that need further processing to the cloud server.

[0061] Specifically, it also includes a third sensor unit 13, which can be a laser ranging sensor, an infrared ranging sensor, an ultrasonic ranging sensor, and a micron radar distance sensor. The third sensor unit 13 is configured to collect length data of objects entering the monitoring channel, and send the object length data to the main control edge computing module. The main control edge computing module preliminarily processes and analyzes the data, and sends the results and data or instructions that need further processing to the cloud server.

[0062] Specifically, the master edge computing module is further configured to send a command to start the third sensor unit 13 if the weight value of the second sensor unit 12 is continuously monitored to be greater than the second threshold value when the weight information of the second sensor unit 12 received is greater than the set second threshold value and a door closing command is sent, and the third sensor unit 13 measures the length of the object. When the length measurement of the object is completed, a door opening command is sent, and a third sensor closing command is sent at the same time;

[0063] And / or when the received weight information of the second sensor unit 12 is greater than the set second threshold and a door closing command is sent, if the weight value of the second sensor is continuously monitored to be less than the second threshold, a door opening command is sent, and the target object is judged to have escaped. At the same time, the second sensor unit 12 collects the remaining weight of the bait box and sends the remaining weight data to the main control edge computing module, which preliminarily processes and analyzes it and sends the results and data or instructions that need further processing to the cloud server.

[0064] When monitoring mice, the first sensor unit 11 and the second sensor unit 12 are cross-verified, and by setting the second threshold, the triggering situation of foreign body interference is excluded, thereby further improving the accuracy.

[0065] The second sensor unit determines whether the object entering is a mouse by judging the weight of the object. If the weight is obviously higher than that of a mouse, it can be automatically released.

[0066] Specifically, it also includes a fourth sensor unit 14, which can be an optical image sensor. The fourth sensor unit 14 is configured to collect image data of objects entering the monitoring channel and send the object image data to the main control edge computing module. The main control edge computing module preliminarily processes and analyzes the data and sends the results and data or instructions that need further processing to the cloud server.

[0067] Specifically, the master edge computing module is further configured to: when the weight information of the second sensor unit 12 received is greater than the set second threshold value and a door closing command is sent, if the weight value of the second sensor unit is continuously monitored to be greater than the second threshold value, send a command to start the fourth sensor, the fourth sensor unit collects images of the target object, and simultaneously measures the weight of the object, and when the image collection of the target object is completed, send a door opening command and send a command to close the fourth sensor unit at the same time;

[0068] And / or when the received weight information of the second sensor unit 12 is greater than the set second threshold value and a door closing command is sent, if the weight value of the second sensor unit is continuously monitored to be less than the second threshold value, a door opening command is sent, and the target object is judged to have escaped. At the same time, the second sensor unit 12 collects the remaining weight of the bait box and sends the remaining weight data to the main control edge computing module, which preliminarily processes and analyzes it and sends the results and data or instructions that need further processing to the cloud server.

[0069] Specifically, it also includes a coprocessor module, and the coprocessor module is configured to: pre-process the image information collected by the fourth sensor unit 14;

[0070] The master edge computing unit is connected to the coprocessor unit by signal, and is used to perform non-uniform correction on the image information according to the deep learning algorithm; smooth the corrected image using the filtering algorithm; separate the target from the background information in the smoothed image according to the background difference method, and detect the moving target; and visually track the moving target using the tracking algorithm;

[0071] The preprocessing includes graying, temperature conversion, filtering and noise reduction of the image information collected by the infrared camera.

[0072] A rodent monitoring method comprises the following steps:

[0073] The first sensor unit 11 detects the target environment in real time. When the infrared radiation information of the target environment detected by the first sensor unit 11 is greater than the set first threshold, it is determined that an object has invaded the target environment; when the weight information fed back by the second sensor unit 12 is greater than the set second threshold, a door closing command is sent, and the electric door closes within 50 milliseconds; when the weight information fed back by the second sensor unit 12 is less than the set second threshold, no door closing command is sent, and the electric door always remains in the open state;

[0074] After the electric door is closed, if the second sensor unit 12 continuously feeds back weight information greater than the second threshold, the main control edge computing module sends an instruction to start the third sensor unit 13 and / or the fourth sensor unit 14 to measure the length and / or capture images of the object entering, and synchronously collect the weight of the object. After the length information, image information and weight information are collected, the main control edge computing module sends a door opening instruction, and the electric door opens. At the same time, the main control edge computing module sends an instruction to close the third sensor unit 13 and / or the fourth sensor unit 14. When the weight information of the second sensor 12 is less than the set second threshold, it is judged that the object has left, and the remaining weight of the bait in the bait box is synchronously collected; if the second sensor 12 continuously feeds back weight information less than the second threshold, it is judged that the object has escaped, and the main control edge computing module sends a door opening instruction, and the automatic door opens;

[0075] The cloud server analyzes and processes the data sent by the master edge computing module, and classifies them into four categories: invalid trigger, foreign object trigger, non-target object, and target trigger;

[0076] If it is determined to be an invalid trigger, the invalid trigger data will be directly deleted;

[0077] If it is determined to be triggered by foreign objects and non-target objects, the data triggered by foreign objects and non-target objects are analyzed and processed and then classified and stored;

[0078] If it is determined to be a target trigger, the data and images of the entry and exit trajectory, weight, body length, and food theft weight generated by the target trigger are analyzed and processed again, and relevant charts of daily activity frequency, activity intensity, weight, body length, food theft amount, quantity, and images are generated and stored in the big database. The main control edge computing module is connected to the cloud server through a wireless communication circuit or a Beidou positioning circuit, wherein:

[0079] Daily activity frequency = activity trajectory per 24 hours ÷ 24 hours;

[0080] Daily activity intensity = daily activity frequency ÷ 24 hours;

[0081] Number of objects = number of times an object is captured per 24 hours.

[0082] Figure 4This is a line graph showing the amount of food stolen by rodents in the monitoring channel from 2022-09-23 16:42:57 to 2022-09-24 13:59:40;

[0083] Figure 5 It is a schematic diagram of the line graph of the frequency of random rodent activities in the monitoring area over 24 hours;

[0084] Figure 6 This is the situation of field tests in the grassland. The triggering frequency of mice entering the monitoring system gradually decreases from 0 o'clock to 5 o'clock, drops to a low value at 5 o'clock, then gradually increases to reach a small peak at 6 o'clock, then gradually decreases to 9 o'clock and reaches a low value again, after 9 o'clock and 10 o'clock, it reaches a small peak again, then gradually decreases to a low value at 14 o'clock, after 14 o'clock it increases again to 19 o'clock, reaches the highest peak, and then gradually decreases.

[0085] Table I is a statistical table of rodent activity from field tests in the grassland monitoring area from September 17, 2022 to September 24, 2022;

[0086] date frequency 2022-09-17 257 2022-09-18 295 2022-09-19 166 2022-09-20 47 2022-09-21 128 2022-09-22 1308 2022-09-23 1769 2022-09-24 293

[0087] Table I

[0088] The above description is only a preferred embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Those skilled in the art should understand that these embodiments can be modified and improved without departing from the principle and spirit of the present invention whose scope is defined by the claims and their equivalents, and these modifications and improvements should also be within the protection scope of the present invention.

Claims

1. A rodent monitoring system, characterized in that: It comprises a first sensor unit (11), a second sensor unit (12), a main control edge computing module (2) and a cloud server; The first sensor unit (11) is configured to perform real-time detection on the monitoring channel; The second sensor unit (12) is configured to monitor the weight data of the monitoring channel in real time; The master edge computing module (2) is configured to fuse the data collected from the first sensor unit (1) and the second sensor unit (2), and based on the data, determine whether the object in the monitoring channel is a rodent; The cloud server is configured to send control instructions to the master edge computing module (2), manage and configure the operating parameters of the master edge computing module (2), receive data uploaded by the master edge computing module (2) and preliminary analysis results, perform global data management and advanced analysis, and send the results or new control instructions back to the master edge computing module (2); The master edge computing module (2) is connected to the cloud server via a wireless communication circuit or a Beidou positioning circuit.

2. The rodent monitoring system according to claim 1, characterized in that: The master control edge computing module (2) is further configured to determine that an object has invaded the monitoring channel when the monitoring value of the monitoring channel collected by the first sensor unit (11) at the current moment exceeds a set first threshold value, and then the master control edge computing module (3) sends a closing control instruction to the automatic door in real time, so that the automatic door is closed.

3. The rodent monitoring system according to claim 1, characterized in that: The main control edge computing module (2) is also configured to send a door closing instruction when the monitoring value of the monitoring channel collected by the first sensor unit (11) at the current moment exceeds a set first threshold value, and the weight value collected by the second sensor unit (12) is greater than a set second threshold value.

4. The rodent monitoring system according to claim 1, characterized in that: The main control edge computing module (2) is also configured such that when the monitoring value of the monitoring channel collected by the first sensor unit (11) at the current moment exceeds a set first threshold value, and the weight value collected by the second sensor unit (12) is greater than a set second threshold value, after sending a door closing command, if the weight value of the second sensor is continuously monitored to be less than the second threshold value, a door opening command is sent, and it is judged that the object has escaped. At the same time, the second sensor unit (12) collects the remaining weight of the bait box and sends the remaining weight data to the main control edge computing module (2). The main control edge computing module (2) performs preliminary processing and analysis, and sends the results and data or instructions that need to be further processed to the cloud server.

5. The rodent monitoring system according to any one of claims 2 to 4, characterized in that: The system also includes a third sensor unit (13), which is configured to collect length data of objects entering the monitoring channel and send the object length data to a main control edge computing module. The main control edge computing module performs preliminary processing and analysis, and sends the results and data or instructions that need further processing to the cloud server.

6. The rodent monitoring system according to claim 5, characterized in that: The master edge computing module (2) is further configured to, when the weight information received from the second sensor unit (12) is greater than a set second threshold value and a door closing command is sent, if the weight value of the second sensor unit (12) is continuously monitored to be greater than the second threshold value, send a command to start the third sensor unit (13), the third sensor unit (13) measures the length of the object, and when the length measurement of the object is completed, send a door opening command and send a third sensor closing command at the same time; And / or when the weight information received from the second sensor unit (12) is greater than a set second threshold value and a door closing command is sent, if the weight value of the second sensor is continuously monitored to be less than the second threshold value, a door opening command is sent, and it is judged that the target object has escaped. At the same time, the second sensor unit (12) collects the remaining weight of the bait box and sends the remaining weight data to the main control edge computing module (2). The main control edge computing module (2) performs preliminary processing and analysis, and sends the results and data or instructions that need to be further processed to the cloud server.

7. The rodent monitoring system according to any one of claims 2 to 4, characterized in that: The system further comprises a fourth sensor unit (14), wherein the fourth sensor unit (14) is configured to collect image data of objects entering the monitoring channel and send the object image data to the master edge computing module (2); the master edge computing module (2) performs preliminary processing and analysis, and sends the results and data or instructions that need further processing to the cloud server.

8. The rodent monitoring system according to claim 7, characterized in that: The main control edge computing module is further configured to: when the weight information received from the second sensor unit (12) is greater than a set second threshold value and a door closing instruction is sent, if the weight value of the second sensor unit is continuously monitored to be greater than the second threshold value, send an instruction to start the fourth sensor, the fourth sensor unit collects an image of the target object and simultaneously measures the weight of the object, and when the image collection of the target object is completed, send an open door instruction and simultaneously send an instruction to close the fourth sensor unit; And / or when the weight information received from the second sensor unit (12) is greater than a set second threshold value and a door closing command is sent, if the weight value of the second sensor unit is continuously monitored to be less than the second threshold value, a door opening command is sent, and it is judged that the target object has escaped. At the same time, the second sensor unit (12) collects the remaining weight of the bait box and sends the remaining weight data to the main control edge computing module (2). The main control edge computing module (2) performs preliminary processing and analysis, and sends the results and data or instructions that need to be further processed to the cloud server.

9. The rodent monitoring system according to claim 7, characterized in that: It also includes a coprocessor module, which is configured to: pre-process the image information collected by the fourth sensor unit (14); The master edge computing unit is connected to the coprocessor unit by signal, and is used to perform non-uniform correction on the image information according to the deep learning algorithm; smooth the corrected image using the filtering algorithm; separate the target from the background information in the smoothed image according to the background difference method, and detect the moving target; and visually track the moving target using the tracking algorithm; The preprocessing includes graying, temperature conversion, filtering and noise reduction of the image information collected by the infrared camera.

10. A method for monitoring rodents, characterized in that Using the rodent monitoring system according to any one of claims 1 to 9 comprises the following steps: The first sensor unit (11) detects the target environment in real time. When the infrared radiation information of the target environment detected by the first sensor unit (11) is greater than a set first threshold, it is determined that an object has invaded the target environment. When the weight information fed back by the second sensor unit (12) is greater than a set second threshold, a door closing command is sent, and the electric door closes within 50 milliseconds. When the weight information fed back by the second sensor unit (12) is less than the set second threshold, no door closing command is sent, and the electric door always remains in an open state. When the electric door is closed, if the second sensor unit (12) continuously feeds back weight information greater than a second threshold, the main control edge computing module sends an instruction to start the third sensor unit (13) and / or the fourth sensor unit (14), measures the length and / or collects images of the object entering, and simultaneously collects the weight of the object; after the length information, image information, and weight information are collected, the main control edge computing module sends a door opening instruction, and the electric door opens; at the same time, the main control edge computing module sends an instruction to close the third sensor unit (13) and / or the fourth sensor unit (14); when the weight information of the second sensor (12) is less than a set second threshold, it is determined that the object has left, and the remaining weight of the bait in the bait box is collected synchronously; if the second sensor (12) continuously feeds back weight information less than the second threshold, it is determined that the object has escaped, and the main control edge computing module sends a door opening instruction, and the automatic door opens; The cloud server analyzes and processes the data sent by the master edge computing module, and classifies them into four categories: invalid trigger, foreign object trigger, non-target object, and target trigger; If it is determined to be an invalid trigger, the invalid trigger data will be directly deleted; If it is determined to be triggered by foreign objects and non-target objects, the data triggered by foreign objects and non-target objects are analyzed and processed and then classified and stored; If it is determined to be a target trigger, the data and images of the entry and exit trajectory, weight, body length, and food theft weight generated by the target trigger will be analyzed and processed again to generate relevant charts of daily activity frequency, activity intensity, weight, body length, food theft amount, quantity, and images and store them in the big database.

Images