An early warning method for forest fires based on lidar and drones

Abstract

The present invention relates to a forest fire early warning method and system based on laser radar and unmanned aerial vehicle. The early warning method obtains laser radar scanning images by controlling the laser radar to perform regular full-cycle cyclic scanning and monitoring of the forest. The smoke plume area in the laser radar scanning image is determined as a suspected fire area. At the same time, the scanning area of ​​the laser radar is reduced according to the azimuth information of the suspected fire area, a local scan is performed, an inspection path is planned, and an unmanned aerial vehicle is dispatched to fly along the inspection path to collect a bird's-eye inspection image in real time. Then, image recognition technology is used to determine whether there are signs of fire in the image. If so, the corresponding suspected fire area is marked as a fire area. Then, the smoke plume information of the fire area in the laser radar scanning image is obtained to evaluate the fire situation in the fire area. Finally, the alarm information and fire information are sent to the decision-making end. The method improves the visibility, timeliness and reliability of the early warning.

Description

Technical Field

[0001] The present invention relates to the field of meteorological monitoring technology, and in particular to a method and system for early warning of forest fires based on laser radar and unmanned aerial vehicles. Background Art

[0002] Forest fire is a natural disaster that is sudden, destructive, and extremely difficult to handle and extinguish. The initial stage of a forest fire is the most favorable time to extinguish the fire, and fire detection is an important part of fire prevention and control. Timely warning of early fire hazards, early detection and early treatment can buy more time for fire fighting, rescue and personnel evacuation, and minimize fire losses.

[0003] The combustion of combustibles is divided into three stages: pyrolysis, smoldering, and open flame. The early pyrolysis and smoldering stages are mainly characterized by the release of a large number of smoke particles, so accurate detection of smoke particles is more conducive to early fire warning. At present, in the application of forest fire monitoring, the monitoring methods mainly include manual monitoring, camera video monitoring, and drone inspection. Manual monitoring is the most common forest fire monitoring method. It relies on human resources for inspection, consumes manpower and material resources, and is prone to missed inspections. Camera video monitoring has strong timeliness, but poor anti-obstruction ability. It is difficult to distinguish fire smoke at night or in foggy weather conditions. Drone inspections have the advantages of flexibility and accurate monitoring results, but they are restricted by inspection methods such as routes, time, and weather, and cannot be continuously monitored around the clock.

[0004] As one of the latest remote sensing technologies, LiDAR has the advantages of high temporal and spatial resolution, high detection accuracy, all-weather, and large scale, and is widely used in the field of three-dimensional detection and research of atmospheric particulate matter. LiDAR can comprehensively monitor the temporal and spatial distribution of atmospheric particulate matter over forests, but the generation and sources of atmospheric particulate matter are rich and varied. Currently, LiDAR has difficulty in determining whether the particles are smoke particles generated by fire. Summary of the invention

[0005] Based on this, it is necessary to address the technical problems in the existing technology of forest fire monitoring and early warning, such as visibility, timeliness and reliability, which need to be improved. The present invention provides a forest fire early warning method and system based on lidar and drone.

[0006] The present invention discloses a forest fire early warning method based on laser radar and unmanned aerial vehicle, comprising the following steps, namely steps S1 to S7.

[0007] S1. Control the laser radar to perform regular full-cycle scanning and monitoring of a forest to obtain a laser radar scanning image that reflects the distribution characteristics of atmospheric particulate matter above the forest.

[0008] S2. Determine the smoke plume area in the laser radar scanning image as a suspected fire area. When a suspected fire area appears, steps S3 and S4 are performed simultaneously.

[0009] S3. Narrow the scanning area of ​​the laser radar according to the azimuth information of the suspected fire area, and perform a local scan that can cover all suspected fire areas.

[0010] S4. Plan an inspection route corresponding to all suspected fire areas based on the location information, and dispatch the drone to fly along the inspection route, collecting overhead inspection images in real time during the flight.

[0011] S5. Use image recognition technology to determine whether there are signs of fire in the overhead inspection image. If so, mark the corresponding suspected fire area as a fire area.

[0012] S6. Obtain smoke plume information of the fire area in the laser radar scanning image, and evaluate the fire situation in the fire area based on the smoke plume information. The smoke plume information includes the number of smoke plumes, the geometric shape of the smoke plumes, the location of the smoke plumes, the color value of the smoke plumes, the area of ​​the smoke plumes, and the frequency of smoke plumes. The types of fire situation include the initial stage of fire, the development stage of fire, the intense stage of fire, and the decline and extinguishing stage of fire.

[0013] S7. Send alarm information and fire scene information to the decision-making end, wherein the fire scene information includes the location of the fire area and the corresponding fire situation.

[0014] As a further improvement of the above scheme, a scanning platform is used to drive the laser radar to perform full-cycle scanning and monitoring of the forest at a preset rotation speed. The laser radar is mounted on the scanning platform, and the scanning platform is used to adjust the scanning posture and scanning speed of the laser radar. The scanning platform has a vertical axis and a horizontal axis for adjusting the pitch angle and azimuth angle respectively.

[0015] As a further improvement of the above scheme, in S5, the fire sign characteristics include forest flame characteristics, smoke characteristics, and fire line boundaries.

[0016] As a further improvement of the above scheme, in step S6, if the geometric shape of the smoke plume in the lidar scanning image is long and narrow or diffusely diffused around a certain point, the position of the smoke plume has the characteristics of concentrated source and uncertain smoke plume tail, and the smoke plume continues to appear in multiple scanning cycles, then this stage is determined to be the initial stage of the fire.

[0017] As a further improvement of the above solution, in step S6, if the number of smoke plumes in the laser radar scanning image increases and the color value of the smoke plume gradually increases in continuous scanning cycles, it is determined that this stage is the fire development stage.

[0018] As a further improvement of the above scheme, in step S6, if the increasing speed of the smoke plume color value in the laser radar scanning image and / or the increasing speed of the smoke plume area are higher than the corresponding growth rate threshold, it is determined that this stage is an intense fire stage.

[0019] As a further improvement of the above solution, in step S6, if the color value and area of ​​the smoke plume in the laser radar scanning image both show a downward trend, it is determined that this stage is the fire decline and extinction stage.

[0020] As a further improvement of the above scheme, during the laser radar scanning monitoring process, the particle distribution profile is obtained by inverting the laser radar echo signal at each scanning angle, and then the spatiotemporal distribution of the particles scanned by the laser radar is obtained.

[0021] Among them, according to the laser radar position coordinates, scanning angle and detection distance, the particle periodic distribution image is integrated with the geographic information, the temporal and spatial distribution of the particle matter is specifically obtained on the map, and then the laser radar scanning image is obtained.

[0022] The invention also discloses an early warning system for forest fires, which comprises a laser radar module, a drone and an early warning control module.

[0023] The lidar module is used to scan and monitor a forest and collect the distribution characteristics of particulate matter in the atmosphere above the forest.

[0024] Drones are used to fly over forests for inspections and collect bird’s-eye inspection images.

[0025] The early warning control module applies any of the above-mentioned forest fire early warning methods based on lidar and drones to control the lidar module and drones to complete early warning of forest fires.

[0026] As a further improvement of the above solution, the UAV includes a flying carrier. The flying carrier is equipped with a fire extinguisher for spraying a fire extinguishing agent on the fire point.

[0027] Compared with the prior art, the technical solution disclosed in the present invention has the following beneficial effects:

[0028] 1. The forest fire early warning method disclosed in the present invention locates the suspected fire area according to the range of the smoke plume in the spatiotemporal distribution of particulate matter above the forest, narrows the scanning range according to the suspected fire area, and formally confirms the fire in combination with drone inspections. At the same time, the laser radar regularly performs full-cycle scanning to fully perceive new suspected fire areas in the environmental space. If a fire is confirmed to exist in the suspected fire area, the fire development stage and combustion situation are determined based on the spatiotemporal distribution of particulate matter. This method improves the visibility, timeliness and reliability of the early warning, and can remotely sense the distribution of smoke in the fire field above the forest in real time on a large scale around the clock, identify and warn of early fire conditions, and provide refined data support for fire risk forecasting, fire monitoring, and fire control.

[0029] 2. The beneficial effects of the forest fire early warning system are the same as those of the above-mentioned forest fire early warning method, which will not be repeated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 This is a flow chart of the forest fire early warning method based on lidar and drone in Example 1 of the present invention. DETAILED DESCRIPTION

[0031] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0032] It should be noted that when a component is referred to as being "mounted on" another component, it may be directly on the other component or there may be a central component. When a component is considered to be "set on" another component, it may be directly set on the other component or there may be a central component at the same time. When a component is considered to be "fixed to" another component, it may be directly fixed on the other component or there may be a central component at the same time.

[0033] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which the present invention belongs. The terms used herein in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The term "or / and" used herein includes any and all combinations of one or more of the related listed items.

[0034] Example 1

[0035] See also Figure 1This embodiment provides a forest fire early warning method based on laser radar and drone, including the following steps, namely steps S1~S7.

[0036] S1. Control the laser radar to perform regular full-cycle scanning and monitoring of a forest to obtain a laser radar scanning image that reflects the distribution characteristics of atmospheric particulate matter above the forest.

[0037] In this embodiment, a scanning platform can be used to drive the laser radar to perform full-cycle cyclic scanning and monitoring of the forest at a preset rotation speed. The scanning platform is used to adjust the scanning posture and scanning speed of the laser radar. During the actual detection process, the laser radar is mounted on the scanning platform, and the vertical axis of the scanning platform can adjust the pitch angle, and the horizontal axis can adjust the azimuth angle. After the vertical axis adjusts the elevation angle, the radar scanning field of view is prevented from being blocked by obstacles such as mountains, buildings, and trees. The scanning platform drives the laser radar to perform cyclic scanning and monitoring at a constant rotation speed to obtain the distribution of atmospheric particulate matter.

[0038] Among them, the laser radar may include lasers, detectors, telescopes, scanning platforms and other parts, and may also directly adopt existing particle laser radars, aerosol laser radars or Mie scattering Doppler laser radars.

[0039] Laser radar uses remote sensing detection technology to emit laser into the atmospheric environment. The laser interacts with aerosol and other particles in the atmosphere. The various characteristic parameters of the atmospheric particles on the laser propagation path are inverted through the received echo signal, such as backscattering coefficient, extinction coefficient, optical thickness, depolarization ratio and other parameters.

[0040] S2. Determine the smoke plume area in the laser radar scanning image as a suspected fire area. When a suspected fire area appears, steps S3 and S4 are performed simultaneously.

[0041] The laser radar performs a full-cycle rapid scan of the environmental space and monitors the changes in the spatiotemporal distribution of particulate matter above the forest in real time. The combustion of combustibles is divided into three stages: pyrolysis, smoldering, and open flame. The early pyrolysis and smoldering stages are mainly characterized by the release of a large number of smoke particles, so accurate detection of smoke particles is more conducive to early fire warning. Under normal circumstances, the spatiotemporal distribution of particulate matter above the forest is characterized by a balanced distribution on the scanning plane. When a fire occurs, a large number of smoke particles are generated and released throughout the fire process, and the smoke concentration above the forest increases abnormally. The interaction between the laser and the smoke particles produces a strong backscattering signal. By inverting the laser radar echo signal at each scanning angle, the spatiotemporal distribution of particulate matter is obtained, and then the periodic distribution image of particulate matter scanned by the laser radar is obtained. According to the differences in characteristic parameters such as the extinction coefficient and depolarization ratio of atmospheric particles in the scanned image, the information such as the number, geometry, position, color value, area and occurrence frequency of particulate plumes can be intuitively obtained.

[0042] In this embodiment, the spatiotemporal distribution of particulate matter can be integrated with geographic information based on the laser radar position coordinates (latitude and longitude), scanning angle, detection distance and other information, and specific information such as the number, geometric shape, position, color value, area and frequency of occurrence of smoke plumes can be obtained on the map. The distribution of smoke above the forest can be remotely sensed on a large scale and in real time around the clock, early fire conditions can be identified and warned, and the development stage and trend of the fire can be analyzed, providing refined data support for fire forecasting, fire detection, and fire control.

[0043] S3. Narrow the scanning area of ​​the laser radar according to the azimuth information of the suspected fire area, and perform a local scan that can cover all suspected fire areas.

[0044] Due to the timeliness of fire conditions, and to reduce the impact of atmospheric flow on the location of the smoke plume source, the lidar scanning cycle should be as short as possible.

[0045] When the concentration of particulate matter in the scanning range increases, the backscatter signal of the laser radar is enhanced, so that it is displayed as a smoke plume area in the laser radar scanning image, which is a suspected fire area detected by the laser radar. When a suspected fire area appears, the laser radar uses a scanning method that combines full-cycle scanning and local scanning for scanning and monitoring. According to the suspected fire area, the scanning range is narrowed for local scanning. The laser radar adjusts the azimuth according to the direction of the fire hazard, narrows the scanning area, and performs automatic scanning and tracking. Compared with full-cycle scanning, after narrowing the scanning area, the laser radar has a faster scanning cycle and a higher scanning speed, which improves the timeliness of early warning. At the same time, in order to ensure the monitoring of potential fire hazards in other areas, the laser radar regularly performs full-cycle scanning to fully perceive new suspected fire areas in the environmental space. If a new suspected fire area appears, the laser radar adjusts the scanning range to ensure that the monitoring range covers all suspected fire areas. By adopting a scanning method that combines full-cycle scanning and local scanning, the laser radar realizes comprehensive monitoring and timely early warning of fire hazards. In this embodiment, the color value of the smoke plume in the laser radar scanning image can be compared with the preset color value threshold to determine whether it is a high-value area.

[0046] S4. Plan an inspection route corresponding to all suspected fire areas based on the location information, and dispatch the drone to fly along the inspection route, collecting overhead inspection images in real time during the flight.

[0047] It should be noted that the inspection route can be located above the forest and set at a certain horizontal distance near each suspected fire area. This can prevent the high temperature and dense fog at the fire scene from affecting the flight work of the drone during inspection.

[0048] Among them, particles such as smoke, dust, pollen, and cloud droplets are typical types of atmospheric aerosol particles. In order to eliminate interference from factors such as clouds and fog and determine that the particles are generated by fire, after the laser radar has identified the suspected fire area, the drone flight system will inspect the suspected fire area to determine whether there is a fire in the suspected fire area. The drone flight system can transmit dynamic information about the forest area in real time and issue an alarm after discovering fire points, smoke points, and other fire conditions in the forest area. The combination of laser radar and drone flight system improves the reliability of early fire warning.

[0049] It is worth noting that if other technical means, such as video or manual methods, have been used to determine that the cause of the increase in particle concentration in the lidar scanning plane is caused by a fire, then whether to conduct drone inspections can be chosen based on actual needs.

[0050] In this embodiment, the UAV flight system is mainly composed of UAV flight carrier, sensor, data transmission system, ground workstation and other components. When the laser radar scans the suspected fire area, the UAV flight carrier takes off to the suspected fire area. The UAV is equipped with equipment such as a camera module or an infrared thermal imager to perform visual detection by taking images or recording videos. The UAV sensor evaluates and feedbacks the forest area by collecting images or videos to determine whether there are fire points, smoke points and other fire conditions.

[0051] Once the drone confirms the existence of a fire, it will send an alarm message and fire scene information to the relevant departments through the data transmission system. The drone flight carrier can carry fire extinguishing materials and spray fire extinguishing agents on the fire point, which can effectively extinguish or suppress the initial fire. The ground workstation is a special parking place designed for drones. When not in operation, the drone automatically stands by at the ground workstation; when working, the door of the ground workstation is opened and the drone automatically flies out to operate.

[0052] In addition, the inspection path planning can adopt the existing commonly used path planning methods. Path planning has a wide range of applications in many fields, such as autonomous collision-free robot actions, GPS navigation, and vehicle problems in logistics management. For general path planning problems within the continuous domain, such as dynamic path planning problems for robots and aircraft, the general steps mainly include three links: environment modeling, path search, and path smoothing.

[0053] S5. Use image recognition technology to determine whether there are signs of fire in the overhead inspection image. If so, mark the corresponding suspected fire area as a fire area. If not, dispatch the drone back to the above-mentioned ground workstation.

[0054] In this embodiment, the fire sign features may include flame features, smoke features, and fire line boundaries. The deep learning recognition model can be trained with images in different lighting scenes, and the trained model can be used to perform flame recognition on input infrared images or ordinary images. This technology belongs to the conventional technology in the field of image recognition, and this application will not elaborate on it.

[0055] S6. Obtain smoke plume information of the fire area in the laser radar scanning image, and evaluate the fire situation in the fire area based on the smoke plume information. The smoke plume information includes the number of smoke plumes, the geometric shape of the smoke plumes, the location of the smoke plumes, the color value of the smoke plumes, the area of ​​the smoke plumes, and the frequency of smoke plumes. The types of fire situation include the initial stage of fire, the development stage of fire, the intense stage of fire, and the decline and extinguishing stage of fire.

[0056] Once it is determined that there is a fire in the suspected fire area, the fire situation is assessed based on the lidar scanning image.

[0057] In the laser radar scanning image, the geometric shape of the smoke plume is long and narrow or diffuses around a certain point. The location of the smoke plume has the characteristics of relatively concentrated source, uncertain smoke plume tail, and the frequency of smoke plume appearance gradually increases in continuous scanning cycles. At this time, the smoke plume generated by the ignition point swings with the flow of the atmosphere. This shows that the fire has not spread at this stage, the burning area is small, and the smoke flow rate is slow. It is the initial stage of the fire. The source position of the smoke plume can be traced back according to the flow direction of the smoke plume, and the source position of the smoke plume is the location of the ignition point. In this embodiment, the aforementioned image recognition technology can be used to identify the geometric shape and movement state of the smoke plume; for example, by identifying the aspect ratio of the geometric shape of the smoke plume, when the aspect ratio meets a certain threshold, the smoke plume is determined to be long and narrow.

[0058] In the laser radar scanning image, the number of smoke plumes increases, and the color value of the smoke plumes gradually increases in the continuous scanning cycle, indicating that the fire is expanding, the burning speed is accelerating, the number of burning points is increasing, and the smoke plume concentration is increasing, which is the fire development stage.

[0059] In the laser radar scanning image, the color value of the smoke plume increases sharply and the area of ​​the smoke plume expands sharply. This indicates that the fire is burning and spreading very quickly, the burning area is expanding rapidly, and the number of burning points is increasing sharply, which is the fierce stage of the fire. In this embodiment, the increasing speed of the smoke plume color value and the increasing speed of the smoke plume area in the laser radar scanning image can be compared with their corresponding growth rate thresholds to determine whether it is the said sharp situation.

[0060] In the laser radar scanning images, the smoke area and smoke color value show a downward trend overall. This indicates that the fire is under control at this stage, the burning area is decreasing, and the smoke concentration is decreasing, which is the descending extinguishing stage. At this stage, if smoke plumes appear from time to time in the laser radar scanning images, it means that Mars has reignited in multiple locations.

[0061] S7. Send alarm information and fire scene information to the decision-making end, wherein the fire scene information includes the location of the fire area and the corresponding fire situation.

[0062] In this embodiment, the decision-making end can be a forest manager's mobile phone, laptop or tablet computer, or other forest management platforms or systems. After receiving the alarm information, the relevant manager can make different decisions and formulate different forest fire extinguishing strategies based on the fire information.

[0063] Example 2

[0064] This embodiment provides a forest fire early warning system, which includes: a laser radar module, a drone and a warning control module.

[0065] The lidar module is used to scan and monitor a forest and collect the distribution characteristics of particulate matter in the atmosphere above the forest.

[0066] The drone is used to fly over the forest for inspection and collect bird's-eye inspection images. The drone includes a flight carrier. The flight carrier is equipped with a fire extinguisher, which can spray fire extinguishing agents on the fire point when necessary.

[0067] The early warning control module can apply the forest fire early warning method based on laser radar and drone in Example 1 to control the laser radar module and drone to complete the early warning of forest fire.

[0068] The technical features of the above-described embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0069] The above-mentioned embodiments only express several implementation modes of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the scope of the invention. It should be pointed out that, for a person of ordinary skill in the art, several modifications and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the attached claims.

Claims

1. A forest fire early warning method based on laser radar and drone, characterized in that: The following steps are involved: S1. Control the laser radar to perform full-cycle scanning and monitoring of a forest periodically to obtain a laser radar scanning image that reflects the distribution characteristics of atmospheric particulate matter above the forest; S2. Determine the smoke plume area in the laser radar scanning image as a suspected fire area; wherein, when the suspected fire area appears, perform steps S3 and S4 simultaneously; S3. narrowing the laser radar scanning area according to the position information of the suspected fire area, and performing a local scan that can cover all suspected fire areas; S4. Planning an inspection path corresponding to all the suspected fire areas according to the position information, and dispatching a drone to fly along the inspection path, and collecting a bird's-eye inspection image in real time during the flight; S5. Using image recognition technology to determine whether there are signs of fire in the overhead inspection image, if so, marking the corresponding suspected fire area as a fire area; S6. Obtain the smoke plume information of the fire area in the laser radar scanning image, and evaluate the fire situation of the fire area according to the smoke plume information; wherein the smoke plume information includes the number of smoke plumes, the geometric shape of the smoke plumes, the position of the smoke plumes, the color value of the smoke plumes, the area of ​​the smoke plumes, and the frequency of smoke plumes; the types of the fire situation include the initial stage of the fire, the development stage of the fire, the intense stage of the fire, and the decline and extinguishing stage of the fire; S7. Send alarm information and fire scene information to the decision-making end; wherein the fire scene information includes the location of the fire area and the corresponding fire scene situation.

2. The forest fire early warning method based on laser radar and drone according to claim 1 is characterized in that: A scanning platform is used to drive a laser radar at a preset rotation speed to perform the full-cycle cyclic scanning and monitoring of the forest; wherein the laser radar is mounted on the scanning platform, and the scanning platform is used to adjust the scanning posture and scanning speed of the laser radar; the scanning platform has a vertical axis and a horizontal axis for adjusting the pitch angle and the azimuth angle respectively.

3. The forest fire early warning method based on laser radar and drone according to claim 1 is characterized in that: In S5, the fire sign characteristics include forest flame characteristics, smoke characteristics, and fire line boundaries.

4. The forest fire early warning method based on laser radar and drone according to claim 1 is characterized in that: In step S6, if the geometric shape of the smoke plume in the laser radar scanning image is long and narrow or diffusely diffused around a certain point, the position of the smoke plume has the characteristics of concentrated source and uncertain smoke plume tail, and the frequency of occurrence of the smoke plume gradually increases in continuous scanning cycles, then this stage is determined to be the initial stage of the fire.

5. The forest fire early warning method based on laser radar and drone according to claim 4 is characterized in that: In step S6, if the number of smoke plumes in the laser radar scanning image increases and the color value of the smoke plume gradually increases in continuous scanning cycles, it is determined that this stage is the fire development stage.

6. The forest fire early warning method based on laser radar and drone according to claim 5 is characterized in that: In step S6, if the increasing speed of the smoke plume color value in the laser radar scanning image and / or the increasing speed of the smoke plume area are higher than the corresponding growth rate threshold, it is determined that this stage is an intense fire stage.

7. The forest fire early warning method based on laser radar and drone according to claim 6 is characterized in that: In step S6, if the color value and area of ​​the smoke plume in the laser radar scanning image both show a downward trend, it is determined that this stage is the fire decline and extinguishing stage.

8. The forest fire early warning method based on laser radar and drone according to claim 1 is characterized in that: During the laser radar scanning monitoring process, the particle distribution profile is obtained by inverting the laser radar echo signal at each scanning angle, and then the spatiotemporal distribution of the particles scanned by the laser radar is obtained; Among them, according to the laser radar position coordinates, scanning angle and detection distance, the particle periodic distribution image is fused with geographic information, the spatiotemporal distribution of the particle is specifically obtained on the map, and then the laser radar scanning image is obtained.

9. A forest fire early warning system, characterized in that: The early warning system includes: A laser radar module is used to scan and monitor a forest and collect the distribution characteristics of particulate matter in the atmosphere above the forest; Drones, which are used to fly over the forest and collect bird's-eye inspection images; and An early warning control module, which applies the forest fire early warning method based on laser radar and drone as described in any one of claims 1 to 8, so as to control the laser radar module and the drone to complete the early warning of the forest fire.

10. The forest fire early warning system according to claim 9, characterized in that: The unmanned aerial vehicle comprises a flying carrier; the flying carrier is equipped with a fire extinguisher for spraying a fire extinguishing agent on a fire point.

Images