Intelligent snail inducing system device and method

Abstract

The invention discloses an intelligent snail inducing system device and method, and relates to the technical field of agricultural pest control. The device comprises an artificial intelligence visual identification module, an intelligent trapping device, a solar power supply module and a cloud data processing and monitoring module. According to the method, eggs and adult pomacea canaliculata are visually recognized through AI, the cloud module predicts the pomacea canaliculata situation trend based on historical data, attractant release and pomacea canaliculata discharging operation of the trapping device are dynamically controlled, and low-power-consumption endurance is achieved through the solar module. The defects that a traditional trapping device is low in efficiency and depends on manual work are overcome, outbreak of the pomacea canaliculata groups can be predicted 72 hours in advance, the recognition accuracy is larger than or equal to 95%, energy consumption is reduced by 70%, and the method is suitable for accurate prevention and control of pomacea canaliculata in rice fields, river channels and other scenes.

Description

Technical Field

[0001] The present invention relates to the technical field of agricultural pest control, and in particular to an intelligent snail trapping system device and method based on artificial intelligence and the Internet of Things. Background Art

[0002] As an invasive species, the golden apple snail is extremely harmful to crops such as rice. Existing control technologies have the following shortcomings:

[0003] Traditional trapping devices (such as the buoy monitor in CN202311742359) can only trap snails passively and cannot actively identify snail conditions;

[0004] Manual inspections are inefficient and difficult to predict snail outbreak trends (for example, the eight-way cage in CN202410817085 can only trap snails statically);

[0005] Existing devices rely on external power supplies or disposable batteries and have poor applicability in remote waters (e.g., CN201821627524 does not involve energy-saving design).

[0006] In response to the above problems, the present invention constructs an integrated intelligent system of "identification-prediction-trapping-monitoring-energy saving" to achieve precise prevention and control. Summary of the Invention

[0007] Technical problems to be solved: How to achieve real-time monitoring, trend prediction, efficient trapping and remote management of apple snails through intelligent means while reducing energy consumption.

[0008] Technical solution: An intelligent snail luring system and method, characterized by including: an artificial intelligence visual recognition module, the module including a high-definition camera that can rotate 360 ​​degrees, an infrared fill light and an edge computing unit, for collecting water area images and identifying the number of golden apple snail eggs and adult snails; an intelligent trapping device, the device including a trapping container, an attractant placement compartment, a one-way inlet structure and an electric snail discharge valve, the one-way inlet is provided with a pressure sensor for triggering the trapping count; a solar power supply module including a photovoltaic panel, an energy storage battery and a voltage regulation unit, for powering the visual recognition module and the intelligent trapping device; a cloud data processing and monitoring module, connected to the visual recognition module and the intelligent trapping device via wireless communication, for receiving image data, generating a snail situation prediction model and remotely controlling the trapping device.

[0009] The edge computing unit of the artificial intelligence visual recognition module has a built-in deep learning model, which can identify apple snail eggs (red features) and adult snails (shell morphology) with an identification accuracy rate of ≥95%.

[0010] The attractant placement chamber of the smart trapping device is provided with a temperature and humidity sensor, which can adjust the attractant release rate according to environmental parameters.

[0011] The cloud data processing and monitoring module includes: a data storage unit for storing historical snail data and images; a prediction unit for predicting the growth trend of the number of golden apple snails in the next 72 hours based on a time series algorithm; and an alarm unit for triggering an audible and visual alarm when the predicted density exceeds a threshold (≥5 snails / m2).

[0012] An intelligent snail luring method, applied to the device of the above scheme, is characterized by comprising the following steps:

[0013] S1. The AI ​​visual recognition module collects water images every hour, uses the edge computing unit to identify the number of golden apple snail eggs and adult snails, and generates real-time snail data.

[0014] S2. Real-time snail data is uploaded to the cloud data processing module, which combines it with historical data and generates a growth trend curve through the prediction unit.

[0015] S3. When the trend curve shows that the density exceeds the threshold, the cloud module sends a start command to the smart trapping device, increasing the attractant release rate by 50% and activating the one-way inlet pressure sensor.

[0016] S4. Every time the trapping device captures 50 golden apple snails, the electric snail expulsion valve automatically opens and expel the snails, simultaneously uploading the expulsion record to the cloud.

[0017] S5. The solar power module monitors the energy storage battery power in real time. When the power is less than 20%, the camera sampling frequency is automatically reduced to once every 3 hours.

[0018] 6. The method according to claim 5, characterized in that the prediction unit in step S2 adopts an LSTM neural network algorithm, the input parameters include the temperature of the day, the number of snail eggs, and the historical capture volume, and outputs the predicted density for the next three days.

[0019] Beneficial effects:

[0020] 1. The accuracy rate has been increased to over 95%, which is 10 times more efficient than traditional manual recognition.

[0021] 2. Predict snail outbreaks within 72 hours, reducing prevention and control response time by 60%;

[0022] 3. Power supply lasting ≥15 days, energy consumption reduced by 70%;

[0023] 4. Remote monitoring reduces manual inspection costs by 80%. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Attached photos:

[0025] Figure 1: Structural block diagram of the intelligent snail trapping system; Figure 2: Cross-sectional view of the intelligent trapping device (including a one-way inlet, attractant chamber, and pressure sensor); Figure 3: Flowchart of the method (recognition-prediction-trapping closed loop).

[0026] Markings in the figure: artificial intelligence visual recognition module 100, smart trapping device 200, trapping container 21, one-way inlet structure 22, electric screw expulsion valve 23, pressure sensor 24, attractant placement chamber 25, humidity sensor 26, solar power supply module 300, cloud data processing and monitoring module 400. DETAILED DESCRIPTION

[0027] like Figure 1 Figure 2 As shown: The system device includes: an artificial intelligence visual recognition module 100, the edge computing unit of the artificial intelligence visual recognition module 100 has a built-in deep learning model, which can identify golden apple snail eggs (red feature) and adult snails (shell morphology) with an identification accuracy of ≥95%;

[0028] The module includes a 360-degree rotating high-definition camera with a 1080P resolution, an infrared fill light, and an edge computing unit. It is used to collect water images and identify the number of golden apple snail eggs and adult snails. The infrared fill light adapts to the night environment and is 360-degree rotated by a stepper motor (rotation angle 0-350°, accuracy ±1°).

[0029] The smart trapping device 200 has an attractant storage compartment 25 equipped with a temperature and humidity sensor 26, which can adjust the attractant release rate according to environmental parameters;

[0030] The device includes a trapping container 21, an attractant placement chamber 25, a one-way inlet structure 22, and an electric snail expulsion valve 23. The one-way inlet is equipped with a pressure sensor 24 for triggering the trapping count. The trapping container 21 is a cylindrical structure with a diameter of 30 cm, and the one-way inlet is an inverted cone (with a diameter of 5 cm). The attractant placement chamber has a built-in slow-release cotton (which can be loaded with the snail's sex pheromone).

[0031] A solar power supply module 300, comprising a photovoltaic panel, an energy storage battery, and a voltage regulator, is used to power the visual recognition module and the intelligent trapping device. The photovoltaic panel has a power of 50W, the energy storage battery has a capacity of 12V / 100Ah, and the voltage regulator supports a 5V-24V adaptive output.

[0032] A cloud-based data processing and monitoring module 400 includes: a data storage unit for storing historical snail data and images; a prediction unit for predicting the growth trend of the number of golden apple snails in the next 72 hours based on a time series algorithm; and an alarm unit for triggering an audible and visual alarm when the predicted density exceeds a threshold (≥5 snails / m2). The module is deployed on an Alibaba Cloud server, supports web and app access, and updates data once per hour. The module is connected to the visual recognition module 100 and the smart trapping device 200 via wireless communication to receive image data, generate a snail prediction model, and remotely control the trapping device.

[0033] like Figure 3 The intelligent snail trapping method is applied to any of the above-mentioned devices, comprising the following steps:

[0034] S1. The AI ​​visual recognition module collects water images every hour, uses the edge computing unit to identify the number of golden apple snail eggs and adult snails, and generates real-time snail data.

[0035] S2. Real-time snail data is uploaded to the cloud data processing module, which combines it with historical data and generates a growth trend curve through the prediction unit.

[0036] S3. When the trend curve shows that the density exceeds the threshold, the cloud module sends a start command to the smart trapping device, increasing the attractant release rate by 50% and activating the one-way inlet pressure sensor.

[0037] S4. Every time the trapping device captures 50 golden apple snails, the electric snail expulsion valve automatically opens and expel the snails, simultaneously uploading the expulsion record to the cloud.

[0038] S5. The solar power module monitors the energy storage battery power in real time. When the power is less than 20%, the camera sampling frequency is automatically reduced to once every 3 hours.

[0039] The prediction unit in step S2 adopts an LSTM neural network algorithm, and the input parameters include the temperature of the day, the number of snail eggs, and the historical capture volume, and outputs the predicted density for the next three days.

[0040] Method flow:

[0041] 1. Startup phase: After the system is powered on, the camera performs a self-check and uploads the initial image to the cloud;

[0042] 2. Identification phase: The edge computing unit runs the YOLOv5 algorithm every hour to identify snail eggs (red pixel clustering) and adult snails (shell outline matching), and outputs quantity data.

[0043] 3. Prediction phase: The cloud-based LSTM model uses the data from the past seven days as input and generates a density curve for the next three days (with a threshold of 5 animals / ㎡).

[0044] 4. Trapping phase: When the threshold is exceeded, the attractant release rate increases from 0.5g / h to 1.0g / h, and the unidirectional inlet pressure sensor (accuracy ±1g) records the number of traps;

[0045] Screw removal and alarm: When 50 screws are removed, the screw removal valve opens (opening time 10s), and the cloud pushes alarm information to the manager's mobile phone.

Claims

1. An intelligent snail trapping system and method, characterized in that: include: An artificial intelligence visual recognition module (100), comprising a high-definition camera capable of 360° rotation, an infrared fill light, and an edge computing unit, for collecting images of water areas and identifying the number of golden apple snail eggs and adult snails; an intelligent trapping device (200), comprising a trapping container, an attractant placement chamber, a one-way inlet structure, and an electric snail discharge valve, wherein the one-way inlet is provided with a pressure sensor for triggering trapping and counting; a solar power supply module (300), comprising a photovoltaic panel, an energy storage battery, and a voltage regulating unit, for powering the visual recognition module and the intelligent trapping device; and a cloud data processing and monitoring module (400), connected to the visual recognition module (100) and the intelligent trapping device (200) via wireless communication, for receiving image data, generating a snail situation prediction model, and remotely controlling the trapping device.

2. The device according to claim 1, characterized in that The edge computing unit of the artificial intelligence visual recognition module (100) has a built-in deep learning model, which can identify golden apple snail eggs (red feature) and adult snails (shell morphology), with an identification accuracy of ≥95%.

3. The device according to claim 1, characterized in that The attractant placement chamber of the smart trapping device (200) is provided with a temperature and humidity sensor, which can adjust the attractant release rate according to environmental parameters.

4. The device according to claim 1, characterized in that The cloud data processing and monitoring module (400) comprises: a data storage unit for storing historical snail data and images; a prediction unit for predicting the growth trend of the number of golden apple snails in the next 72 hours based on a time series algorithm; and an alarm unit for triggering an audible and visual alarm when the predicted density exceeds a threshold (≥5 snails / m2).

5. An intelligent snail trapping method, applied to the device according to any one of claims 1 to 4, characterized in that: The following steps are involved: S1. The AI ​​visual recognition module collects water images every hour, uses the edge computing unit to identify the number of golden apple snail eggs and adult snails, and generates real-time snail data. S2. Real-time snail data is uploaded to the cloud data processing module, which combines it with historical data and generates a growth trend curve through the prediction unit. S3. When the trend curve shows that the density exceeds the threshold, the cloud module sends a start command to the smart trapping device, increasing the attractant release rate by 50% and activating the one-way inlet pressure sensor. S4. Every time the trapping device captures 50 golden apple snails, the electric snail expulsion valve automatically opens and expel the snails, simultaneously uploading the expulsion record to the cloud. S5. The solar power module monitors the energy storage battery power in real time. When the power is less than 20%, the camera sampling frequency is automatically reduced to once every 3 hours.

6. The method according to claim 5, characterized in that The prediction unit in step S2 adopts an LSTM neural network algorithm, and the input parameters include the temperature of the day, the number of snail eggs, and the historical capture volume, and outputs the predicted density for the next three days.

Images