Forest fire early warning method and system based on small airship

Abstract

The invention relates to the technical field of forest fire protection, and discloses a forest fire early warning method and system based on a small airship. The method comprises the following steps: firstly, controlling the small airship to hover at a specified position over a monitoring area; then, a data detection module arranged on the small airship is used for collecting fire behavior related data; the fire related data comprises an infrared image of a monitoring area and a sensor signal set; the sensor signal set comprises a temperature sensor signal, a rainfall sensor signal, a wind speed sensor signal, a humidity sensor signal and a wind direction sensor signal; then fire behavior related data is preprocessed; and finally, the fire early warning level is judged according to the preprocessing result of the fire related data, and whether an early warning signal is sent to a remote forest fire monitoring center or not is selected according to the judgment result. According to the invention, the use and distribution of human resources are greatly optimized.

Description

Technical Field

[0001] This invention relates to the field of forest fire prevention technology, specifically to a forest fire early warning method and system based on a small airship. Background Technology

[0002] With the increasing severity of forest fires, forest fire prevention and control technologies and measures are constantly evolving to cope with increasingly complex fire prevention situations. The application of modern meteorological early warning systems helps to forecast extreme weather conditions, such as droughts and high temperatures, allowing for timely fire prevention measures. For example, commonly used methods currently include regional fire risk assessment systems based on satellite remote sensing and fixed ground-based monitoring devices. While these systems can provide wide-area early warning information, they can only assess the risk of forest fires over large areas and have detection blind spots due to orbital cycles. This system can comprehensively assess natural and human factors in high-risk areas on a smaller scale, with higher accuracy than remote sensing. After a forest fire occurs, it analyzes the trend and spread of the fire, primarily targeting the refined management of key protected areas.

[0003] In addition, although some regions control fire risk by restricting human activities, especially by prohibiting open fires in high-risk areas (such as the boundary between forests and grasslands), such measures are difficult to achieve comprehensive and effective management in areas with vast open spaces and weak human control. Summary of the Invention

[0004] To address the technical problem that existing forest fire early warning schemes are not timely and comprehensive enough, this invention provides a forest fire early warning method and system based on a small airship.

[0005] To achieve the above objectives, the present invention provides the following technical solution: This invention discloses a forest fire early warning method based on a small airship, comprising the following steps: S1. Control the small airship to hover at a designated position above the monitoring area; S2. Collect fire-related data using a data detection module mounted on a small airship; the fire-related data includes infrared images of the monitored area and a set of sensor signals; the set of sensor signals includes: temperature sensor signals, rainfall sensor signals, wind speed sensor signals, humidity sensor signals, and wind direction sensor signals; S3. Preprocess fire-related data; S4. Determine the fire warning level based on the preprocessing results of fire-related data, and select whether to send a warning signal to a remote forest fire monitoring center based on the determination results.

[0006] As a further improvement to the above scheme, step S4 includes the following specific steps in determining the fire warning level: S41. Based on the sensor signal set, obtain the initial risk level values ​​of five environmental meteorological factors: temperature, rainfall, wind speed, humidity, and wind direction; wherein, the sensor signal set includes: temperature sensor signal, rainfall sensor signal, wind speed sensor signal, humidity sensor signal, and wind direction sensor signal; S42. Identify the vehicle type based on the infrared image and obtain the initial risk level value of the vehicle identification factor; S43. Assign weights to environmental meteorological factors and vehicle identification factors, and perform a linear mapping on the initial risk level values ​​of the two to obtain the corresponding linear mapping results; S44. Extract human behavior characteristics from infrared images of the monitored area to obtain the initial risk level value of the human behavior tracking factor; S45. Extract fire features from the infrared images of the monitored area to obtain the initial risk level values ​​of the actual image analysis factors; S45. Assign weights to the personnel behavior tracking factor and the actual image analysis factor, and perform a nonlinear mapping on the initial risk level values ​​of the two to obtain the corresponding nonlinear mapping results; S46. Based on the linear and nonlinear mapping results, the Fire Index is calculated:

[0007] In the formula, W The corresponding mapping result for each initial risk level value. s for W The corresponding weights; S47. Compare the fire index with the preset judgment threshold to obtain the fire warning level.

[0008] As a further improvement to the above scheme, in step S45, the expression formula for the nonlinear mapping is:

[0009] In the formula, This represents the risk level value after nonlinear mapping; α To adjust the parameters; is the risk level value before nonlinear mapping; tanh(·) is the hyperbolic tangent activation function.

[0010] As a further improvement to the above scheme, step S3 includes the following specific steps in the preprocessing: S31. Sort all infrared images acquired within a preset time period in chronological order to generate a video stream; S32. The following formula is used to preprocess the data of the sensor signal set:

[0011]

[0012] In the formula, For the first t Sensor at each sampling time i Signal, i =1,2,3,4,5 correspond to temperature, rainfall, wind speed, humidity, and wind direction, respectively; For sensors i The summation of the differences between adjacent acquired signals; m This represents the total difference between adjacent signals; For sensors i Threshold value; f (·) represents the unit step function; R i For sensors i The corresponding local decision results.

[0013] As a further improvement to the above scheme, in step S4, if it is necessary to send an early warning signal to the forest fire monitoring center, an early warning signal packet containing the fire location coordinates, environmental meteorological parameters, and video stream is generated based on the preprocessing results of fire-related data and the coordinates recorded by the GPS module on the small airship; the early warning signal packet is transmitted to the backend server of the forest fire monitoring center in real time through the LoRa protocol, thereby triggering the audible and visual alarm and visual pop-up prompt of the forest fire monitoring center.

[0014] This invention also discloses a forest fire early warning system based on a small airship, which applies the forest fire early warning method based on a small airship as described above. The system includes: The airship control module is used to control the small airship to hover at a designated position above the monitoring area; A data detection module is installed on a small airship and is used to collect fire-related data. The fire-related data includes infrared images of the monitored area and a set of sensor signals. The set of sensor signals includes: temperature sensor signals, rainfall sensor signals, wind speed sensor signals, humidity sensor signals, and wind direction sensor signals. The data processing module is used to preprocess fire-related data; determine the fire warning level based on the preprocessed data, and select whether to send a warning signal to a remote forest fire monitoring center based on the determination result.

[0015] As a further improvement to the above solution, the data detection module includes an infrared camera, a temperature sensor, a rainfall sensor, a wind speed sensor, a humidity sensor, and a wind direction sensor.

[0016] Compared with the prior art, the beneficial effects of the present invention are: 1. This invention uses various sensors integrated on a small airship to measure in real time various factors that induce forest fires, and quantifies and integrates natural factors such as humidity and wind speed as well as many human factors, thereby determining the fire warning level of the monitoring area, and choosing whether to report to a remote forest fire monitoring center based on the determination result, which greatly optimizes the use and allocation of human resources.

[0017] 2. This invention comprehensively considers several factors within the monitoring area, including environmental meteorology, vehicle type, personnel behavior characteristics, and actual imagery. Based on real-time collected fire-related data, it determines the level of each factor and processes the level values ​​using linear and nonlinear mapping. Through a low-cost airship platform, a multi-factor fusion evaluation model, and a closed-loop emergency communication architecture, it not only achieves a full-chain upgrade of prevention and control from response to early warning to source tracing, but also breaks through existing technological bottlenecks in three dimensions: cost controllability, early warning accuracy, and real-time response. The multi-source data synchronous feedback mechanism adopted in this invention, while meeting core early warning requirements, can also generate high-precision fire source tracing capabilities, providing a systematic solution for forest fire prevention and control.

[0018] 3. This invention, upon determining the presence of fire risk in a monitored area, can simultaneously package and transmit the area's location coordinates, environmental meteorological parameters, and infrared image / video streams to a remote forest fire monitoring center in real time. This significantly shortens the response and deployment time of firefighting resources during a disaster. By continuously monitoring and recording environmental meteorological changes and abnormal activities of personnel / vehicles in high-risk areas in real time, this invention provides detailed data support for fire cause analysis and post-disaster responsibility tracing, improving the scientific rigor and accuracy of disaster management. Attached Figure Description

[0019] Figure 1 This is a flowchart of the forest fire early warning method based on a small airship in Embodiment 1 of the present invention. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] Example 1

[0022] Please see Figure 1This embodiment provides a forest fire early warning method based on a small airship, including the following steps, namely S1 to S4.

[0023] S1. Control the small airship to hover at a designated position above the monitoring area.

[0024] In this embodiment, the small airship can be a self-designed and manufactured airship with a volume of 6m³, a load capacity of 2kg, and an appearance and structure similar to the existing h-aero™ hybrid aircraft.

[0025] S2. Collect fire-related data using a data detection module mounted on a small airship; the fire-related data includes infrared images of the monitored area and a set of sensor signals. The set of sensor signals includes: temperature sensor signals, rainfall sensor signals, wind speed sensor signals, humidity sensor signals, and wind direction sensor signals.

[0026] S3. Preprocess fire-related data.

[0027] In step S3, the preprocessing includes the following specific steps, namely S31~S32.

[0028] S31. Sort all infrared images acquired within a preset time period in chronological order to generate a video stream; S32. The following formula is used to preprocess the data of the sensor signal set:

[0029]

[0030] In the formula, For the first t Sensor at each sampling time i Signal, i =1,2,3,4,5 correspond to temperature, rainfall, wind speed, humidity, and wind direction, respectively; For sensors i The summation of the differences between adjacent acquired signals; m This represents the total difference between adjacent signals; For sensors i Threshold value; f (·) represents the unit step function; R i For sensors i The corresponding local decision results.

[0031] S4. Determine the fire warning level based on the preprocessing results of fire-related data, and select whether to send a warning signal to a remote forest fire monitoring center based on the determination results.

[0032] If it is necessary to send an early warning signal to the forest fire monitoring center, an early warning signal packet containing the fire location coordinates, environmental meteorological parameters, and video stream is generated based on the preprocessing results of fire-related data and the coordinates recorded by the GPS module on the small airship. The early warning signal packet is then transmitted to the backend server of the forest fire monitoring center in real time via the LoRa protocol, thereby triggering the sound and light alarm and visual pop-up prompt of the forest fire monitoring center.

[0033] Upon receiving an early warning signal, the forest fire monitoring center actively receives the video stream transmitted by the airship and can view the forest fire's occurrence process in real time through video playback software or the monitoring system. Based on information such as the fire's spread path, smoke direction, and fire source location displayed in the video, combined with on-site meteorological conditions (such as wind speed and direction), the monitoring center staff can determine the fire's starting location and possible causes. Furthermore, by analyzing image information from different time points in the video stream, the entire process of the fire's occurrence and spread can be traced. Based on the video stream analysis results, the monitoring center generates a fire source tracing report, which includes detailed information such as the fire's start time, coordinates, cause (natural / human-caused), spread rate, and affected area.

[0034] In step S4, the method for determining the fire warning level includes the following specific steps, namely S41 to S47.

[0035] S41. Based on the sensor signal set, obtain the initial risk level values ​​of five environmental meteorological factors: temperature, rainfall, wind speed, humidity, and wind direction.

[0036] In this embodiment, the initial risk level classification method for the five environmental meteorological factors is as follows: I. Temperature (°C) When the temperature is below 25℃, it is considered low risk (T1), with an initial risk level value of [value missing]. The value is 1.

[0037] When the temperature is between 25℃ and 35℃, it is classified as medium risk (T2), with an initial risk level value of [value missing]. The value is 2.

[0038] When the temperature is above 35℃, it is classified as high risk (T3), with an initial risk level value of [value missing]. The value is 3.

[0039] II. Rainfall (mm / h)

[0040] When rainfall is less than 2.5 mm, it is considered low risk (R1), with an initial risk level value of [value missing]. The value is 1.

[0041] When rainfall is between 2.5 and 8.0 mm, it is classified as medium risk (R2), with an initial risk level of [value missing]. The value is 2.

[0042] When rainfall exceeds 8.0 mm, it is classified as high risk (R3), with an initial risk level value of [value missing]. The value is 3.

[0043] III. Relative Humidity (RH%)

[0044] When the relative humidity is above 60%, it is considered low risk (H1), with an initial risk level value of [value missing]. The value is 1.

[0045] When the relative humidity is between 30% and 60%, it is classified as medium risk (H2), with an initial risk level of [value missing]. The value is 2.

[0046] When the relative humidity is below 30%, it is considered high risk (H1), with an initial risk level value of [value missing]. The value is 3.

[0047] IV. Wind speed (m / s)

[0048] When the wind speed is below 3 m / s, it is considered low risk (S1), with an initial risk level value of [value missing]. The value is 1.

[0049] When the wind speed is between 3 m / s and 6 m / s, it is classified as medium risk (S2), with an initial risk level value of [value missing]. The value is 2.

[0050] When the wind speed is higher than 6 m / s, it is classified as high risk (S3), with an initial risk level value of [value missing]. The value is 3.

[0051] V. Wind Direction

[0052] When the wind is blowing towards water or roads, the risk level is low (D1), and the initial risk level is [value missing]. The value is 1.

[0053] When the wind direction is towards densely populated areas, it is considered high risk (D2), with an initial risk level value of [value missing]. The value is 2.

[0054] It should be noted that wind speed and wind direction are dynamically divided into percentages based on statistics over a certain period of time.

[0055] S42. Identify the vehicle type based on the infrared image and obtain the initial risk level value of the vehicle identification factor.

[0056] In this embodiment, the initial risk level value classification method for vehicle identification factors is as follows: When all vehicles appearing within the forest monitoring area are registered fire prevention vehicles (fire trucks, patrol cars), the risk level is low (V1), with an initial risk rating of [value missing]. The value is 1.

[0057] When unauthorized vehicles (such as agricultural vehicles or tourist cars) enter the forest monitoring area, especially when they enter the pre-designated core forest area, the risk level is classified as medium (V2), with an initial risk level value of [value missing]. The value is 2.

[0058] When a blacklisted vehicle (involved in a fire incident) or a vehicle transporting flammable or explosive materials enters the forest monitoring area, it is considered high risk (V3), with an initial risk level value of [value missing]. The value is 3.

[0059] It should be noted that in this embodiment, the infrared images can be identified and the aforementioned vehicle categories can be analyzed using an artificial intelligence host deployed on a small airship. Specifically, a pre-trained model for vehicle type recognition can be used, employing the YOLOv8 vehicle recognition algorithm for vehicle type identification.

[0060] S43. Assign weights to environmental meteorological factors and vehicle identification factors, and perform a linear mapping on the initial risk level values ​​of the two to obtain the corresponding linear mapping results.

[0061] In this embodiment, a linear mapping result can be generated using a linear mapping function. The weights of the above factors and the risk level values ​​after linear mapping are as follows: Temperature factor: =0.1; =0.3; =0.6; weight is 4.

[0062] Rainfall factors: =0.1; =0.3; =0.6; weight is 2.

[0063] Relative humidity factor: =0.1; =0.3; =0.6; weight is 4.

[0064] Wind speed factor: =0.1; =0.3; =0.6; weight is 1.5.

[0065] Wind direction factor: =0.1; =0.3; weight is 2.

[0066] S44. Extract human behavior characteristics from infrared images of the monitored area to obtain the initial risk level value of the human behavior tracking factor.

[0067] In this embodiment, the initial risk level value classification method for personnel behavior tracking factors is as follows: When there is no human activity or compliant behavior (such as ranger patrols) detected within the forest monitoring area, it is considered low risk (L1), with an initial risk level of [value missing]. It is -1.

[0068] When potentially dangerous behaviors (such as tourists smoking) are detected within the forest monitoring area, it is classified as low risk (L2), with an initial risk level value of [value missing]. It is 0.

[0069] When open flame use (burning, picnicking), carrying flammable materials, or damage to fire prevention facilities are detected within the forest monitoring area, it is classified as high risk (L3), with an initial risk level value of [value missing]. It is 0.9.

[0070] In this embodiment, human behavior can be identified using the YOLOv8 human behavior recognition algorithm through a pre-trained model for human behavior recognition (the dataset differs from that of the vehicle category recognition model).

[0071] S45. Extract fire features from infrared images of the monitored area to obtain the initial risk level values ​​of the actual image analysis factors.

[0072] In this embodiment, the initial risk level classification method for actual image analysis factors is as follows: When there are no thermal anomalies, smoke, or open flames within the monitored forest area, it is considered low risk (I1), with an initial risk level value of [value missing]. It is -1.

[0073] When there are localized high temperatures or slight haze within the forest monitoring area, it is classified as medium risk (I2), with an initial risk level value of [value missing]. It is 0.

[0074] When there is a spreading fire line or open flames and dense smoke within the forest monitoring area, it is classified as high risk (I3), with an initial risk level value. It is 0.9.

[0075] S45. Assign weights to the personnel behavior tracking factor and the actual image analysis factor, and perform a nonlinear mapping on the initial risk level values ​​of the two to obtain the corresponding nonlinear mapping results.

[0076] In this embodiment, nonlinear mapping can be performed using the Tanh function, expressed as follows:

[0077] In the formula, This represents the risk level value after nonlinear mapping; α To adjust the parameters; is the risk level value before nonlinear mapping; tanh(·) is the hyperbolic tangent activation function.

[0078] The weights of the above factors and the risk level values ​​after nonlinear mapping are as follows: Personnel behavior tracking factors: =0.015; =0.5; =1; weight is 12.

[0079] Actual impact analysis factors: =0.015; =0.5; =1; weight is 12.

[0080] S46. Based on the linear and nonlinear mapping results, the Fire Index is calculated:

[0081] In the formula, W The corresponding mapping result for each initial risk level value. s for W The corresponding weights.

[0082] S47. Compare the fire index with the preset judgment threshold to obtain the fire warning level.

[0083] In this embodiment, the determination threshold can be set as follows: Level I (Blue Alert): Fire Index ≤ 4; Level II (Yellow Alert): 4 < Fire Index ≤ 8 Level III (Orange Alert): 8 < Fire Index ≤ 12 Level IV (Red Alert): Fire Index > 12.

[0084] Example 2

[0085] This embodiment provides a forest fire early warning system based on a small airship, which applies the forest fire early warning method based on a small airship as described above. The system includes: an airship control module, a data detection module, and a data processing module.

[0086] The airship control module is used to control a small airship to hover at a designated position above the monitoring area. In this embodiment, the implementation of the airship control module includes the following steps: Step 1: Obtain the system's angular velocity and acceleration using an inertial measurement unit (IMU). The specific process is as follows: Accelerometers detect the capacitance change indirectly caused by the displacement of a mass block (or equivalent physical quantity change) due to inertial force, and convert it into an electrical signal output to detect the dynamic acceleration of the system in real time. Gyroscopes convert angular velocity into electrical signals by detecting inertial effects caused by rotation (such as Coriolis force) or by utilizing the principle of conservation of angular momentum. The signal strength characterizes the precise angular velocity value of the system. The inertial measurement unit can be the Bosch BMI088, which integrates a 16-bit triaxial gyroscope and a 16-bit triaxial accelerometer, and features high vibration resistance and excellent temperature stability.

[0087] Step 2: Obtain the system's altitude and location information using a barometer or GPS. The specific process is as follows: The barometer uses a MEMS capacitive sensing unit to monitor atmospheric pressure by utilizing the piezoresistive effect of silicon (changes in air pressure cause deformation of the silicon diaphragm, resulting in a change in its resistance, which is converted into an electrical signal). Combined with temperature compensation and digital signal processing technology, it outputs high-precision air pressure data. The air pressure value then reflects the system's altitude. Original formula:

[0088] In the formula, Standard atmospheric pressure is equal to 1013.25 mbar; Altitude is the altitude in meters; P is the air pressure in mbars at a certain altitude. Thus, the instantaneous altitude is obtained:

[0089] Step 3: By fusing data from multiple sensors through a microcontroller, errors from a single sensor are eliminated, and the precise flight attitude, three-dimensional position, and velocity are calculated.

[0090] In this embodiment, the microcontroller can be an STM32F405RGT6 with a 32-bit ARM Cortex-M4 core.

[0091] Step 4: The main control development board compares the current state with the target state (such as desired altitude and attitude angle) and generates an error signal. Using control algorithms such as PID (Proportional-Integral-Derivative) to calculate the speed adjustment of each motor, it ensures a fast and stable response.

[0092] Step 5: The main control development board executes preset waypoints or real-time mission commands (such as hovering or flight path) from the ground station by adjusting the speed of the four motors. At the same time, it adjusts the attitude to counteract external interference such as wind disturbances and maintain flight stability.

[0093] The data detection module is mounted on a small airship and is used to collect fire-related data, including infrared images of the monitored area and a set of sensor signals. The data detection module includes an infrared camera, a temperature sensor, a rainfall sensor, a wind speed sensor, a humidity sensor, and a wind direction sensor.

[0094] The data processing module is used to preprocess fire-related data; it determines the fire warning level based on the preprocessing results of the fire-related data, and selects whether to send a warning signal to a remote forest fire monitoring center based on the determination results.

[0095] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A forest fire early warning method based on small airships, characterized in that, Includes the following steps: S1. Control the small airship to hover at a designated position above the monitoring area; S2. Collect fire-related data using a data detection module mounted on a small airship; the fire-related data includes infrared images of the monitored area and a set of sensor signals; the set of sensor signals includes: temperature sensor signals, rainfall sensor signals, wind speed sensor signals, humidity sensor signals, and wind direction sensor signals; S3. Preprocess fire-related data; S4. Determine the fire warning level based on the preprocessing results of fire-related data, and select whether to send a warning signal to a remote forest fire monitoring center based on the determination results.

2. The forest fire early warning system based on a small airship according to claim 1, characterized in that, In step S4, the method for determining the fire warning level includes the following specific steps: S41. Based on the sensor signal set, obtain the initial risk level values ​​of five environmental meteorological factors: temperature, rainfall, wind speed, humidity, and wind direction; S42. Identify the vehicle type based on the infrared image and obtain the initial risk level value of the vehicle identification factor; S43. Assign weights to environmental meteorological factors and vehicle identification factors, and perform a linear mapping on the initial risk level values ​​of the two to obtain the corresponding linear mapping results; S44. Extract human behavior characteristics from infrared images of the monitored area to obtain the initial risk level value of the human behavior tracking factor; S45. Extract fire features from the infrared images of the monitored area to obtain the initial risk level values ​​of the actual image analysis factors; S45. Assign weights to the personnel behavior tracking factor and the actual image analysis factor, and perform a nonlinear mapping on the initial risk level values ​​of the two to obtain the corresponding nonlinear mapping results; S46. Based on the linear and nonlinear mapping results, the Fire Index is calculated: In the formula, W The corresponding mapping result for each initial risk level value. s for W The corresponding weights; S47. Compare the fire index with the preset judgment threshold to obtain the fire warning level.

3. A forest fire early warning system based on a small airship according to claim 2, characterized in that, In step S45, the expression formula for the nonlinear mapping is: In the formula, This represents the risk level value after nonlinear mapping; α To adjust the parameters; is the risk level value before nonlinear mapping; tanh(·) is the hyperbolic tangent activation function.

4. The forest fire early warning system based on a small airship according to claim 2, characterized in that, In step S3, the preprocessing includes the following specific steps: S31. Sort all infrared images acquired within a preset time period in chronological order to generate a video stream; S32. The following formula is used to preprocess the data of the sensor signal set: In the formula, For the first t Sensor at each sampling time i Signal, i =1,2,3,4,5 correspond to temperature, rainfall, wind speed, humidity, and wind direction, respectively; For sensors i The summation of the differences between adjacent acquired signals; m This represents the total difference between adjacent signals; For sensors i Threshold value; f (·) represents the unit step function; R i For sensors i The corresponding local decision results.

5. The forest fire early warning method based on a small airship according to claim 4, characterized in that, In step S4, if it is necessary to send an early warning signal to the forest fire monitoring center, an early warning signal packet containing the fire location coordinates, environmental meteorological parameters, and video stream is generated based on the preprocessing results of the fire-related data and the coordinates recorded by the GPS module on the small airship. The warning signal packets are transmitted in real time to the back-end server of the forest fire monitoring center via the LoRa protocol, thereby triggering the sound and light alarms and visual pop-up prompts of the forest fire monitoring center.

6. A forest fire early warning system based on a small airship, characterized in that, The forest fire early warning method based on a small airship as described in any one of claims 1 to 5, the system comprising: The airship control module is used to control the small airship to hover at a designated position above the monitoring area; A data detection module is installed on a small airship and is used to collect fire-related data. The fire-related data includes infrared images of the monitored area and a set of sensor signals. The set of sensor signals includes: temperature sensor signals, rainfall sensor signals, wind speed sensor signals, humidity sensor signals, and wind direction sensor signals. The data processing module is used to preprocess fire-related data; determine the fire warning level based on the preprocessed data, and select whether to send a warning signal to a remote forest fire monitoring center based on the determination result.

7. The forest fire early warning system based on a small airship according to claim 6, characterized in that, The data detection module includes an infrared camera, a temperature sensor, a rainfall sensor, a wind speed sensor, a humidity sensor, and a wind direction sensor.

Images