Unmanned aerial vehicle-based village environment intelligent monitoring and early warning system and method

Abstract

The invention belongs to the technical field of traditional village environment monitoring, and discloses a village environment intelligent monitoring and early warning system and method based on an unmanned aerial vehicle. The data acquisition layer and the edge data processing layer are integrated on the unmanned aerial vehicle, the acquired data are processed in real time, long-distance transmission of the data is not needed, the data transmission time is shortened, data processing is directly carried out on the unmanned aerial vehicle, only necessary results are transmitted, and bandwidth resources are saved; data processing is carried out while data acquisition is carried out, potential risks can be found in time, and the risks are controlled, so that the inspection and monitoring efficiency of the unmanned aerial vehicle is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of traditional village environment monitoring, and in particular relates to a village environment intelligent monitoring and early warning system and method based on unmanned aerial vehicles. Background Art

[0002] The value of a traditional village lies in its integrity, the organic integration of material space and intangible culture, and the traditional lifestyles that villagers continue to live within it. It is a living, evolving community, not a static museum.

[0003] However, at present, due to the relatively backward transportation and facilities in traditional villages, it is difficult to protect the environment and ecology; and the transformation of traditional villages is also relatively complicated, so environmental monitoring and early warning in traditional villages are difficult to carry out. Summary of the Invention

[0004] The purpose of the present invention is to provide an intelligent monitoring and early warning system and method for village environment based on drones to address the above-mentioned problems, hoping to improve the ecological environment protection of existing traditional villages, which requires manpower for inspection and is therefore inefficient; and after existing drones detect the target object, they need to upload the data for processing, which results in high latency, heavy network burden, and low bandwidth utilization.

[0005] The technical solution adopted in the present invention is as follows: A drone-based intelligent village environment monitoring and early warning system. The drone is used to monitor and provide early warning for traditional village environments. The system includes the following: In the data acquisition layer, the drone is equipped with a multispectral camera and a lidar. The multispectral camera is used to scan and photograph the module to be inspected; the lidar is used to scan the module to be inspected and obtain three-dimensional data. The edge data processing layer, which is carried on the drone, is used to perform real-time analysis on the data collected by the data collection layer; The cloud layer is used to build the traditional village model based on the data collected by the data collection layer, and generate an annotated map to mark the risky modules to be detected; The control layer is used to transmit warning information to the staff and control the equipment in each module to be detected to respond; Visualization layer, which displays the visualization results of monitoring and early warning; Among them, the data collection layer transmits data to the edge data processing layer and the cloud layer for processing. The control layer transmits warning information based on the labeled map generated by the cloud layer, and transmits the monitoring and warning results to the visualization layer for visual display.

[0006] Furthermore, the modules to be detected include building complexes, roads and vegetation groups, the building complexes include ancient buildings and building land, the roads include roads and rivers used for construction, and the vegetation groups include fields and landscape plants.

[0007] It should be noted that ancient buildings are existing buildings in traditional villages, including but not limited to buildings with cultural relics and old buildings.

[0008] Furthermore, the multispectral camera is provided with multiple lenses, including but not limited to a thermal infrared camera and a visible light camera; the thermal infrared camera is used to identify the module to be detected with abnormal temperature, and the visible light camera is used to capture high-definition images to assist the thermal infrared camera in confirming the module to be detected with abnormal temperature.

[0009] Furthermore, the drone is also equipped with sensors, which are connected to the edge data processing layer and the cloud layer signal. The sensors include distance sensors, smoke sensors and multispectral sensors. The distance sensors are set around the drone to identify the distance between the drone and obstacles, the smoke sensors are used to detect the smoke content in the environment, and the multispectral sensors are used to monitor the modules to be detected.

[0010] It should be noted that the multispectral sensor, smoke sensor and distance sensor are existing technologies; the smoke sensor is used to detect the smoke components of gaseous pollutants, organic compounds and fire pollutants generated by burning objects.

[0011] Furthermore, the edge data processing layer stores the data in the edge data processing layer through the edge computing framework.

[0012] Furthermore, a method for intelligent monitoring and early warning of village environment based on drones includes the following steps: S100: Collect historical data of traditional villages, find data from professional data systems and import the data into the cloud layer as initial data of traditional villages; The professional data system is an existing database platform.

[0013] Among them, historical data includes building area, building height, inclination angle of each building, vegetation coverage area, village precipitation, etc. The existing relevant data of the village are tested through historical data, and the village environment is intelligently monitored using historical data as a reference.

[0014] S200: Collecting real-time data of the traditional village and planning a path for the drone to conduct a comprehensive scan of the traditional village; using the multispectral camera of the data acquisition layer to capture the module to be detected, generating a captured image; using the laser radar of the data acquisition layer to detect the tilt angle and depth of the module to be detected, generating detection data; The traditional village is divided into multiple sub-areas, and the UAV path planning adopts the grid coverage algorithm: ; in, is the area of the subregion, is the flight speed of the drone, The flight time of the drone, is the weight coefficient; Multispectral imaging resolution control: ; in, is the flight altitude of the drone, is the focal length, is the pixel size, is the ground resolution; LiDAR tilt angle detection formula: ; in, is the coordinate change of the module to be detected; S300: Preliminary data processing: transmitting the real-time data collected in step S200 to the edge data processing layer, and obtaining abnormal information by processing the real-time data; Build a dynamic threshold model: ; in, is a constant coefficient, is the time decay factor, is the base time, is the initial standard deviation, is the initial mean; Calculate the anomaly index: ; in, is the standard deviation, is the dynamic mean, is an abnormal point, is the weight coefficient; S400: Deep data processing, transmitting the abnormal information obtained in step S300 to the cloud layer, which constructs a traditional village model based on historical information and annotates the abnormal information obtained in step S300 on the traditional village model to form an annotated map; Constructing a traditional village model: ; in, For the reconstructed 3D model, is the Laplace operator of the model, The vector field generated for the lidar point cloud, is the regularization term, is a bounded region in space, is the regularization coefficient; S500: Risk warning. According to the abnormal information marked on the map, the control layer sends a control signal to the equipment in the corresponding module to be detected, and performs corresponding processing on the module to be detected according to the abnormal information.

[0015] The risk level assessment formula is: ; in, is the risk type, is the risk weight coefficient.

[0016] Furthermore, in step S200, the edge data processing layer performs preliminary processing on the data, including the following: S201: Setting the dynamic standard deviation ; ; in, is the number of real-time data samples, For the Real-time data, is the mean of real-time data; S202: Perform a multi-dimensional comparison of the real-time data and the initial data. If there is a deviation between the real-time data and the initial data, and the deviation is greater than the set standard deviation, the real-time data is considered risky data. Conversely, if there is no deviation between the real-time data and the initial data, the real-time data is considered risk-free data. S203: Concentrate risk data into abnormal information, and store the abnormal information in the edge data processing layer; ; in, is the average length after encoding, For the The probability of occurrence of risk data, For the The encoding length of each risk data, is the information entropy.

[0017] It should be noted that multi-dimensional comparison requires all environmental data of real-time data and initial data, including but not limited to building inclination, field fire conditions, river water volume, and road traffic conditions.

[0018] Furthermore, the traditional village model is constructed based on historical data and average data of traditional villages, and the traditional village model is used to represent the general form of traditional villages.

[0019] The historical data of traditional villages include building area, building height and inclination angle of each building, vegetation coverage area, village precipitation, etc.

[0020] The average data of traditional villages is obtained based on the average of historical data of existing traditional villages across China.

[0021] The general form is the standard form, that is, when the traditional village has no risks such as natural disasters and the buildings have no tendencies such as tilt, it means that the current environment of the traditional village is good.

[0022] Furthermore, the risk points, risk types and specific real-time data are marked on the marked map, and the abnormal information is compressed and transmitted to the control layer via the cloud layer.

[0023] Furthermore, the annotated map and abnormal information are displayed through a visualization layer.

[0024] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are: The present invention integrates a data acquisition layer and an edge data processing layer on the drone to process the collected data in real time. There is no need to transmit the data over long distances, which reduces the time of data transmission. The data is processed directly on the drone and only necessary results are transmitted, saving bandwidth resources. Data processing is performed while data is collected, which can timely discover potential risks and control risks, thereby improving the efficiency of drone inspections and monitoring.

[0025] The multispectral camera of the present invention is equipped with multiple lenses, which are used to capture images and detect the temperature of the modules to be detected with abnormal temperatures. Multiple lenses assist in multi-directional detection of the environment, making it easy to quickly discover problems during drone inspections and improve the monitoring efficiency of the monitoring system.

[0026] The present invention combines the model with the collected data to draw an annotated map, issue an early warning of risks, and conduct real-time risk processing of the traditional village environment based on the risk points, risk types and specific real-time data marked on the annotated map. Risks can be processed by controlling the processing equipment of the drone or the module to be detected, thereby improving the risk processing efficiency of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 This is a system architecture diagram of the present invention; Figure 2 is a flow chart of the method of the present invention; Figure 3 This is a schematic diagram of the building complex and vegetation complex of the present invention; Figure 4 This is an aerial top view of the building of the present invention. DETAILED DESCRIPTION

[0028] The present invention will be described in detail below with reference to the accompanying drawings.

[0029] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0030] like Figure 1 As shown, a drone-based intelligent village environment monitoring and early warning system is used to monitor and provide early warning for traditional village environments. The system includes the following: In the data acquisition layer, the drone is equipped with a multispectral camera and a lidar. The multispectral camera is used to scan and photograph the module to be inspected; the lidar is used to scan the module to be inspected and obtain three-dimensional data. The modules to be detected include building complexes, roads and vegetation complexes. The building complexes include ancient buildings and construction land. The roads include roads and rivers used for construction. The vegetation complexes include fields and landscape plants.

[0031] The multispectral camera is equipped with multiple lenses, including but not limited to a thermal infrared camera and a visible light camera; the thermal infrared camera is used to identify the module to be detected with abnormal temperature, and the visible light camera is used to capture high-definition images to assist the thermal infrared camera in confirming the module to be detected with abnormal temperature.

[0032] The edge data processing layer, which is carried on the drone, is used to perform real-time analysis on the data collected by the data collection layer; The drone is also equipped with sensors, which are connected to the edge data processing layer and the cloud layer signal. The sensors include distance sensors, smoke sensors and multispectral sensors. The distance sensors are set around the drone to identify the distance between the drone and obstacles, the smoke sensors are used to detect the smoke content in the environment, and the multispectral sensors are used to monitor the modules to be detected.

[0033] The edge data processing layer stores data in the edge data processing layer through the edge computing framework.

[0034] The cloud layer is used to build the traditional village model based on the data collected by the data collection layer, and generate an annotated map to mark the risky modules to be detected; The control layer is used to transmit warning information to the staff and control the equipment in each module to be detected to respond; Visualization layer, which displays the visualization results of monitoring and early warning; Among them, the data collection layer transmits data to the edge data processing layer and the cloud layer for processing. The control layer transmits warning information based on the labeled map generated by the cloud layer, and transmits the monitoring and warning results to the visualization layer for visual display.

[0035] like Figure 2 As shown, a method for intelligent monitoring and early warning of village environment based on drones includes the following steps: S100: Collect historical data of traditional villages, find data from professional data systems and import the data into the cloud layer as initial data of traditional villages; S200: Collecting real-time data of the traditional village and planning a path for the drone to conduct a comprehensive scan of the traditional village; using the multispectral camera of the data acquisition layer to capture the module to be detected, generating a captured image; using the laser radar of the data acquisition layer to detect the tilt angle and depth of the module to be detected, generating detection data; The traditional village is divided into multiple sub-areas, and the UAV path planning adopts the grid coverage algorithm: ; in, is the area of the subregion, is the flight speed of the drone, The flight time of the drone, is the weight coefficient; Multispectral imaging resolution control: ; in, is the flight altitude of the drone, is the focal length, is the pixel size, is the ground resolution; LiDAR tilt angle detection formula: ; in, is the coordinate change of the module to be detected; In the above step S200, the edge data processing layer performs preliminary processing on the data, including the following: S201: Setting the dynamic standard deviation ; ; in, is the number of real-time data samples, For the Real-time data, is the mean of real-time data; S202: Perform a multi-dimensional comparison of the real-time data and the initial data. If there is a deviation between the real-time data and the initial data, and the deviation is greater than the set standard deviation, the real-time data is considered risky data. Conversely, if there is no deviation between the real-time data and the initial data, the real-time data is considered risk-free data. S203: Concentrate risk data into abnormal information, and store the abnormal information in the edge data processing layer; ; in, is the average length after encoding, For the The probability of occurrence of risk data, For the The encoding length of each risk data, is the information entropy.

[0036] S300: Preliminary data processing: transmitting the real-time data collected in step S200 to the edge data processing layer, and obtaining abnormal information by processing the real-time data; Build a dynamic threshold model: ; in, is a constant coefficient, is the time decay factor, is the base time, is the initial standard deviation, is the initial mean; Calculate the anomaly index: ; in, is the standard deviation, is the dynamic mean, is an abnormal point, is the weight coefficient; S400: Deep data processing, transmitting the abnormal information obtained in step S300 to the cloud layer, which constructs a traditional village model based on historical information and annotates the abnormal information obtained in step S300 on the traditional village model to form an annotated map; Constructing a traditional village model: ; in, For the reconstructed 3D model, is the Laplace operator of the model, The vector field generated for the lidar point cloud, is the regularization term, is a bounded region in space, is the regularization coefficient; The marked map is marked with risk points, risk types and specific real-time data, and the abnormal information is compressed and transmitted to the control layer via the cloud layer.

[0037] S500: Risk warning. According to the abnormal information marked on the map, the control layer sends a control signal to the equipment in the corresponding module to be detected, and performs corresponding processing on the module to be detected according to the abnormal information.

[0038] The risk level assessment formula is: ; in, Risk type is set by management personnel based on the specific situation of the village; is the risk weight coefficient.

[0039] The annotated map and abnormal information are displayed through a visualization layer.

[0040] By integrating the data collection layer and edge data processing layer on the drone, the collected data can be processed in real time, eliminating the need for long-distance data transmission, reducing data transmission time, and processing the data directly on the drone, transmitting only necessary results, saving bandwidth resources. By processing data while collecting data, potential risks can be discovered in a timely manner and controlled, thereby improving the efficiency of drone inspections and monitoring.

[0041] Example 1 like Figure 3 As shown, one embodiment of the present invention is that when the module to be detected is a building complex (black box), the system first collects and organizes the initial data of the building and plans a path for the drone to collect the building. The drone uses a multispectral camera to capture the building image and uses a lidar to scan the building to obtain three-dimensional data; the data acquisition layer summarizes the captured building image and the three-dimensional data of the laser scan to form detection data; the edge data processing layer analyzes the detection data. If there is a deviation between the detection data and the initial data, and the deviation is greater than the set standard deviation, the detection data is risk data. Otherwise, the detection data is risk-free data; the risk data is converted into abnormal information and stored in the edge data processing layer.

[0042] The edge data processing layer determines whether the abnormal information can be processed. If it can be processed in time, the abnormal information will be transmitted to the cloud layer for alarm. The cloud layer sends a control signal to the drone or the corresponding device of the module to be detected, and the drone or the corresponding device of the module to be detected will perform risk processing. If the building is tilted, the tilt angle will be set and the cause of the tilt will be checked through the corresponding device of the module to be detected, and then the building tilt will be processed in time.

[0043] The visualization layer visualizes abnormal information, including specific risk locations, risk types, real-time data, etc., to facilitate staff to view and process data in a timely manner.

[0044] In order to protect the ancient buildings in traditional villages, drones are used to monitor the village environment. When abnormal information is found, timely warnings are issued to facilitate risk management. At the same time, drone inspections can improve the efficiency of inspections and the accuracy of detection.

[0045] Example 2 like Figure 3 As shown, one embodiment of the present invention is that when the module to be detected is a vegetation group (orange box), the drone flies according to the planned path. When it reaches above the vegetation group, the temperature of the vegetation area is detected by the thermal infrared camera on the drone, and the alarm temperature of the thermal infrared camera is set according to the temperature environment; the thermal infrared camera transmits the detected temperature data to the edge data processing layer, and compares the detected temperature data with the alarm temperature; if the temperature of a certain area of the vegetation is detected to be higher than the alarm temperature, alarm data is generated, and the alarm data is transmitted to the control layer and the visualization layer, and the control layer transmits the control information; conversely, if the temperature of a certain area of the vegetation is detected to be lower than the alarm temperature, the detected temperature data is stored in the edge data processing layer, and the edge data processing layer transmits the temperature data to the visualization layer for data display.

[0046] This embodiment monitors and provides early warnings for vegetation fire prevention. If a high temperature is detected, the potential fire location is addressed and a timely warning is issued to personnel, allowing them to extinguish the fire and evacuate personnel. Drone inspections enable timely detection of critical situations, and the edge data processing layer performs preliminary processing on the detection data to further verify the reliability of the results.

[0047] Example 3 Another embodiment of the present invention is that the modules to be detected include but are not limited to building complexes, roads and vegetation groups existing in traditional villages, the building complexes include ancient buildings and building land, the roads include roads and rivers used for construction, and the vegetation groups include fields and landscape plants.

[0048] The multispectral camera carried on the drone is equipped with multiple cameras. The cameras are installed and used according to the type of module to be detected, so as to collect data in a targeted manner for different types of modules to be detected.

[0049] This embodiment is based on the portability of the drone, and the camera has the characteristics of being light in weight, fast in imaging, and easy to replace.

[0050] Example 4 Another embodiment of the present invention is that the initial data of the traditional village is data compiled from public information such as historical images, data records, and statistical data, and the resulting initial data serves as a control group for collecting real-time data; the initial data covers all types of specific data in the module to be detected, including precipitation, vegetation types and specific dimensions, river water volume, building dimensions and inclination angles, etc., and is used to calculate precipitation and river water storage capacity. When precipitation increases, drainage is carried out in a timely manner to avoid disasters such as floods; and the status of ancient buildings can be monitored. If the ancient buildings have problems such as tilt or damage, an alarm can be issued in a timely manner to facilitate maintenance by staff.

[0051] The collected real-time data is analyzed based on long-term statistical data as the initial data, and the conclusions drawn are more accurate. The initial data is updated in real time based on risk-free real-time data.

[0052] Example 5 Another embodiment of the present invention is that the drone is also equipped with sensors, which are connected to the edge data processing layer and the cloud layer signal, and the sensors include distance sensors, smoke sensors and multispectral sensors; the distance sensors are arranged around the drone to identify the distance between the drone and obstacles, the smoke sensors are used to detect the smoke content in the environment, and the multispectral sensors are used to monitor the modules to be detected.

[0053] The functions of the multispectral sensor in this embodiment include but are not limited to the following areas: In agriculture, multispectral sensors can sense the growth and nutritional status of plants, helping farmers apply fertilizers and manage crops accurately, thereby improving crop yield and quality. By monitoring crop chlorophyll content, moisture levels, and pest and disease conditions, farmers can adjust agricultural operations in a timely manner to ensure healthy crop growth. In the field of water quality, multispectral sensors can sense water quality and other information for monitoring and management. By using parameters such as the content of pollutants in water, environmental problems can be discovered and reported in a timely manner. Therefore, the sensors installed on drones can adapt to a variety of usage scenarios and broaden the monitoring range of drones.

[0054] Example 6 Another embodiment of the present invention is that in step S200, the edge data processing layer performs preliminary processing on the data, including the following: S201: Setting the dynamic standard deviation ; ; in, is the number of real-time data samples, For the Real-time data, is the mean of real-time data; S202: Real-time data With initial data Perform multi-dimensional comparisons, if real-time data With initial data If there is a deviation compared to the real-time data, and the deviation is greater than the set standard deviation, the real-time data will be As risk data; on the contrary, if real-time data With initial data Compared with the absence of deviation, the real-time data Risk-free data; Set the real-time data weight to , the initial data weight is , the formula for comparing real-time data with initial data is as follows: ; This means that real-time data With initial data Compared with the absence of deviation, the real-time data Risk-free data; ; This means that real-time data With initial data If there is a deviation compared to the real-time data, and the deviation is greater than the set standard deviation, the real-time data will be as risk data; For example, the initial tilt angle of a building in a village is 6°. The drone detects that the real-time tilt angles of the building at different times are 10°, 12°, and 13°. The average real-time tilt angle is 11.67 degrees. The dynamic standard deviation is ; set up , , then the real-time tilt angles of 10°, 12° and 13° are all risk data; combined with the soil settlement height, the cause of the building tilt is verified.

[0055] S203: Risk data Concentrate abnormal information and stores the exception information in the edge data processing layer.

[0056] ; in, is the average length after encoding, For the The probability of occurrence of risk data, For the The encoding length of each risk data, is the information entropy.

[0057] like Figure 4 As shown, Figure 4 This is an aerial view of a building. The scale in the image represents the building's top dimensions, measured in meters. When a drone is directly above a building, it can only measure the building's top dimensions, and the resulting data may have certain errors. However, the data measured from the top view can reflect the building's deformation. If the measured data from the top view has errors, there may be problems with the roof collapsing or unstable foundations.

[0058] By comparing real-time data with initial data, risk data can be found, and then alarms and processing can be performed on the corresponding areas in the modules to be detected, thereby increasing the accuracy and reliability of judgment.

[0059] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. The intelligent monitoring and early warning method for village environment based on drones is characterized by: The environmental intelligent monitoring and early warning method includes the following steps: S100: Collect historical data of traditional villages, find data from professional data systems and import the data into the cloud layer as initial data of traditional villages; S200: Collecting real-time data of the traditional village and planning a path for the drone to conduct a comprehensive scan of the traditional village; using the multispectral camera of the data acquisition layer to capture the module to be detected, generating a captured image; using the laser radar of the data acquisition layer to detect the tilt angle and depth of the module to be detected, generating detection data; The traditional village is divided into multiple sub-areas, and the UAV path planning adopts the grid coverage algorithm: ； in, is the area of the subregion, is the flight speed of the drone, The flight time of the drone, is the weight coefficient; Multispectral imaging resolution control: ； in, is the flight altitude of the drone, is the focal length, is the pixel size, is the ground resolution; LiDAR tilt angle detection formula: ； in, is the coordinate change of the module to be detected; S300: Preliminary data processing: transmitting the real-time data collected in step S200 to the edge data processing layer, and obtaining abnormal information by processing the real-time data; Build a dynamic threshold model: ； in, is a constant coefficient, is the time decay factor, is the base time, is the initial standard deviation, is the initial mean; Calculate the anomaly index: ； in, is the standard deviation, is the dynamic mean, is an abnormal point, is the weight coefficient; S400: Deep data processing, transmitting the abnormal information obtained in step S300 to the cloud layer, which constructs a traditional village model based on historical information and annotates the abnormal information obtained in step S300 on the traditional village model to form an annotated map; Constructing a traditional village model: ； in, For the reconstructed 3D model, is the Laplace operator of the model, The vector field generated for the lidar point cloud, is the regularization term, is a bounded region in space, is the regularization coefficient; S500: Risk warning: Based on the abnormal information marked on the map, the control layer sends a control signal to the device in the corresponding module to be detected, and the module to be detected is processed accordingly according to the abnormal information; The risk level assessment formula is: ； in, is the risk type, is the risk weight coefficient.

2. The method for intelligent monitoring and early warning of village environment based on drones according to claim 1 is characterized in that: In step S200, the edge data processing layer performs preliminary processing on the data, including the following: S201: Setting the dynamic standard deviation ; ； in, is the number of real-time data samples, For the Real-time data, is the mean of real-time data; S202: Perform a multi-dimensional comparison of the real-time data and the initial data. If there is a deviation between the real-time data and the initial data, and the deviation is greater than the set standard deviation, the real-time data is considered risky data. Conversely, if there is no deviation between the real-time data and the initial data, the real-time data is considered risk-free data. S203: Concentrate risk data into abnormal information, and store the abnormal information in the edge data processing layer; ； in, is the average length after encoding, For the The probability of occurrence of risk data, For the The encoding length of each risk data, is the information entropy.

3. The method for intelligent monitoring and early warning of village environment based on drones according to claim 2 is characterized in that: The traditional village model is constructed based on the historical data and average data of traditional villages. The traditional village model is used to represent the general form of traditional villages.

4. The method for intelligent monitoring and early warning of village environment based on drones according to claim 3 is characterized in that: The risk points, risk types and specific real-time data are marked on the marked map, and the abnormal information is compressed and transmitted to the control layer through the cloud layer.

5. The method for intelligent monitoring and early warning of village environment based on drones according to claim 4 is characterized in that: Annotated maps and anomaly information are displayed through the visualization layer.

6. A drone-based village environment intelligent monitoring and early warning system, using the drone-based village environment intelligent monitoring and early warning method according to any one of claims 1 to 5, wherein the drone is used to monitor and provide early warning for the traditional village environment, characterized in that: The system includes the following: In the data acquisition layer, the drone is equipped with a multispectral camera and a lidar. The multispectral camera is used to scan and photograph the module to be inspected; the lidar is used to scan the module to be inspected and obtain three-dimensional data. The edge data processing layer, which is carried on the drone, is used to perform real-time analysis on the data collected by the data collection layer; The cloud layer is used to build the traditional village model based on the data collected by the data collection layer, and generate an annotated map to mark the risky modules to be detected; The control layer is used to transmit warning information to the staff and control the equipment in each module to be detected to respond; Visualization layer, which displays the visualization results of monitoring and early warning; Among them, the data collection layer transmits data to the edge data processing layer and the cloud layer for processing. The control layer transmits warning information based on the labeled map generated by the cloud layer, and transmits the monitoring and warning results to the visualization layer for visual display.

7. The intelligent monitoring and early warning system for village environment based on drones according to claim 6 is characterized in that: The modules to be detected include building complexes, roads and vegetation complexes. Building complexes include ancient buildings and construction land. Roads include roads and rivers used for construction. Vegetation complexes include fields and landscape plants.

8. The intelligent monitoring and early warning system for village environment based on drones according to claim 7 is characterized in that: The multispectral camera is equipped with multiple lenses, including but not limited to a thermal infrared camera and a visible light camera; the thermal infrared camera is used to identify the module to be detected with abnormal temperature, and the visible light camera is used to capture high-definition images to assist the thermal infrared camera in confirming the module to be detected with abnormal temperature.

9. The intelligent monitoring and early warning system for village environment based on drones according to claim 8 is characterized in that: The drone is also equipped with sensors, which are connected to the edge data processing layer and the cloud layer signal. The sensors include distance sensors, smoke sensors and multispectral sensors. The distance sensors are set around the drone to identify the distance between the drone and obstacles, the smoke sensors are used to detect the smoke content in the environment, and the multispectral sensors are used to monitor the modules to be detected.

10. The intelligent monitoring and early warning system for village environment based on drones according to claim 9 is characterized in that: The edge data processing layer stores data in the edge data processing layer through the edge computing framework.

Images