Intelligent forest fire prevention system based on unmanned aerial vehicle

Abstract

The invention discloses an intelligent forest fire prevention system based on unmanned aerial vehicles, and the system comprises an unmanned aerial vehicle cluster module which supports the access of various types of unmanned aerial vehicles, and each unmanned aerial vehicle has a unique identifier; the task planning module is provided with a visual interface, and parameters such as a flight path can be set; the data processing module can receive, store and analyze unmanned aerial vehicle data and has edge computing capability; the hovering control module enables the unmanned aerial vehicle to hover by optimizing a flight control system; the emergency response module is used for automatically triggering an alarm and generating a fire extinguishing scheme when a fire disaster is monitored; a communication system supporting a plurality of communication protocols; a sensor including a multispectral camera and the like; the flight control system has an intelligent obstacle avoidance function; the data processing module has a machine learning capability and can optimize a fire prediction model; and the image preprocessing module is used for preprocessing the image data. The system can realize insertion of unmanned aerial vehicle deployment, real-time viewing of unmanned aerial vehicle data and unmanned aerial vehicle hovering work, and is especially suitable for the field of forest fire prevention.

Description

Technical Field

[0001] The present invention relates to the technical field of forest fire prevention by drones, and particularly to an intelligent forest fire prevention system based on drones. Background Art

[0002] Early forest fire prevention methods mainly relied on manual observation and ground patrol. This method was not only inefficient but also had a very limited coverage area. With the gradual progress of technology, aerial patrol and satellite remote sensing technology began to be applied to the field of forest fire prevention. Aerial patrol can provide a wider monitoring range, while satellite remote sensing has a macroscopic perspective and data collection capabilities, bringing innovation to forest fire prevention work. Entering the 21st century, the rapid development of drone technology has opened up new ways for forest fire prevention. Drones, with their significant advantages such as flexibility, low cost, and real-time monitoring, have quickly been widely applied in the field of forest fire prevention and become a key force in the modern forest fire prevention system.

[0003] In current forest fire prevention practices, the application forms of drones are diverse and powerful. First, drones are equipped with advanced sensors such as high-definition cameras and infrared thermal imagers, which can obtain high-resolution visible light images and thermal imaging data in real time, so as to timely and accurately detect potential fire sources and monitor the dynamic changes of fire situations. Second, when a fire occurs, drones can quickly reach the scene and provide detailed information about the fire scene to the emergency command center, including key data such as the intensity of the fire and the spreading direction of the fire line, providing strong support for formulating a scientific and reasonable fire extinguishing plan. In addition, drones also have a fire warning function. They can conduct regular patrols according to preset routes and, combined with meteorological data and the state of forest vegetation, achieve early warning of fires, greatly improving the timeliness of fire prevention and control. Finally, some models of drones have the ability to carry fire extinguishing agents and can directly participate in fire extinguishing operations in the initial stage of the fire, effectively curbing the spread of the fire and buying precious time for subsequent fire extinguishing work.

[0004] However, there are still some problems to be solved urgently in the existing drone forest fire prevention technology. Currently, the endurance of drones is generally limited, and the continuous flight time is usually between 30 minutes and 1 hour, which is obviously difficult to meet the needs of long-term patrols. The environmental adaptability of drones in adverse weather conditions is poor. For example, in complex meteorological conditions such as fog and strong winds, their flight stability and monitoring effects will be greatly affected. Moreover, the massive data collected by drones needs to be processed and analyzed quickly, but the existing data processing capabilities are often lagging, resulting in untimely information transmission. Therefore, the existing forest fire prevention effect of drones is not ideal. Summary of the Invention

[0005] The present invention provides an intelligent forest fire prevention system based on unmanned aerial vehicles to solve the technical problems of low flight stability and monitoring effect, lagging data processing ability, and inability to view the status in real time when existing unmanned aerial vehicles are used for forest fire prevention.

[0006] To solve the above technical problems, the present invention provides the following technical solutions:

[0007] The present invention provides an intelligent forest fire prevention system based on unmanned aerial vehicles, including:

[0008] Unmanned aerial vehicle cluster module: Supports the access of multiple types of unmanned aerial vehicles, covering fixed-wing, rotary-wing, and unmanned airships, etc. Each unmanned aerial vehicle has a unique identifier and can be seamlessly docked with the system to achieve plug-and-play;

[0009] Task planning module: Provides a visual interface. Users can set flight paths, monitoring areas, and data collection parameters to generate task instructions. It supports three-dimensional map display, and users can directly draw flight paths and mark monitoring areas on the map;

[0010] Data processing module: Receives, stores, and analyzes the data transmitted in real time by unmanned aerial vehicles, including images, temperature, humidity, and gas concentration, etc. It has edge computing capabilities and can perform data preprocessing at the unmanned aerial vehicle end to reduce the amount of data transmitted;

[0011] Hover control module: Enables unmanned aerial vehicles without hover capabilities to achieve hovering by optimizing the flight control system. It adopts a closed-loop control algorithm based on sensor feedback, combined with airflow compensation and attitude adjustment technologies to ensure the stable hovering of unmanned aerial vehicles in complex environments;

[0012] Emergency response module: When a forest fire is detected, it automatically triggers an alarm, generates a fire extinguishing plan, and coordinates unmanned aerial vehicles to carry out fire extinguishing operations;

[0013] Communication system: Supports multiple communication protocols, including Zigbee, Wi-Fi, and 4G / 5G, to ensure stable data transmission;

[0014] Sensors: Include multispectral cameras, infrared thermal imagers, meteorological sensors, and gas detectors;

[0015] Flight control system: Has an intelligent obstacle avoidance function and can sense the surrounding environment in real time and automatically avoid obstacles;

[0016] Data processing module: Has machine learning capabilities and can analyze historical data to optimize the fire prediction model;

[0017] Image preprocessing module: Performs preprocessing on the image data collected by unmanned aerial vehicles, including operations such as grayscale conversion and binarization, to improve image quality and reduce the amount of data;

[0018] Fire segmentation algorithm module: Extract flame features from the preprocessed image, and adopt spatial domain processing methods, including filter processing, Otsu image threshold segmentation algorithm, morphological operation and flame feature extraction.

[0019] The beneficial effects brought by the technical solution provided by the present invention at least include:

[0020] The task planning module of the present invention provides a visual interface and supports three-dimensional map display. Users can directly draw flight paths and mark monitoring areas on the map. This makes task planning more intuitive and accurate.

[0021] According to the task requirements set by the user, the present invention automatically optimizes and verifies the task to ensure the feasibility and safety of the task. This not only improves the efficiency of task planning but also reduces human errors.

[0022] The data processing module of the present invention has edge computing capabilities and can perform data preprocessing on the drone side to reduce the amount of data transmission. This not only improves the efficiency of data processing but also reduces the requirement for communication bandwidth.

[0023] The hover control module of the present invention adopts a closed-loop control algorithm based on sensor feedback, combined with airflow compensation and attitude adjustment technologies, enabling drones without hover capabilities to hover stably in complex environments. This significantly improves the applicability and stability of drones in complex environments.

[0024] The system of the present invention can monitor forest fires in real time through multiple sensors (such as multispectral cameras, infrared thermal imagers, etc.), and can quickly detect the fire source and accurately locate the fire. Description of the Drawings

[0026] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0027] Figure 1 It is a flowchart of a method for predicting the carbon content and temperature of molten steel at the end of a converter provided by an embodiment of the present invention. Detailed Embodiments

[0028] To make the objectives, technical solutions, and advantages of the present invention clearer, the following will further describe the embodiments of the present invention in detail with reference to the drawings.

[0029] This embodiment provides an intelligent forest fire prevention system based on drones.

[0030] I. UAV Cluster Module

[0031] This system supports the access of various types of UAVs, including fixed-wing, rotary-wing, and unmanned airships, etc. Each UAV has a unique identifier and can be seamlessly docked with the system to achieve plug-and-play. The system establishes a communication connection with the UAV through the communication interface unit and supports the conversion and adaptation of multiple communication protocols. The UAV identification and configuration unit automatically identifies the type and parameters of the accessed UAV and assigns a unique identifier and corresponding task configuration to it. The task assignment unit reasonably assigns tasks to UAVs according to the task requirements set by the user and optimizes the task scheduling. The status monitoring unit monitors the flight status, battery power, and data transmission information of the UAV in real time to ensure the smooth execution of tasks. Users can flexibly select and replace UAVs without reconfiguring the entire system.

[0032] In specific implementation, the system manages and assigns tasks to the UAV cluster through the following steps:

[0033] 1. Establishment of communication connection: The communication interface unit supports multiple communication protocols such as Zigbee, Wi-Fi, 4G / 5G, etc. When a UAV accesses the system, the communication interface unit automatically detects the communication protocol type of the UAV and performs an adaptive connection. For example, for a rotary-wing UAV using Wi-Fi communication, the communication interface unit will automatically identify its communication frequency and signal strength and complete the connection.

[0034] 2. UAV identification and configuration: The UAV identification and configuration unit obtains parameter information such as its type, model, and sensor configuration by sending an identification signal to the accessed UAV. The system assigns a unique identifier to the UAV based on this information and configures corresponding task parameters according to its performance characteristics. For example, for a fixed-wing UAV, the system will configure a large-area inspection task according to its long flight range and high flight speed; for a rotary-wing UAV, a fine monitoring task for a local area will be configured.

[0035] 3. Task assignment and optimization: The task assignment unit reasonably assigns tasks according to the task requirements set by the user and combines the performance parameters of the UAV. The system adopts a task scheduling algorithm to optimize the task assignment according to the battery power, flight status, and task priority of the UAV. For example, when multiple UAVs access the system simultaneously, the system will give priority to assigning tasks to UAVs with sufficient battery power and good performance, and reasonably allocate the number of tasks and flight paths according to the size and complexity of the task area.

[0036] 4. Status Monitoring and Feedback: The status monitoring unit continuously monitors the flight status of the drone, including parameters such as flight altitude, speed, attitude, as well as battery power and data transmission information. The system adjusts the flight parameters of the drone in real-time through sensor data feedback to ensure the smooth execution of the task. For example, when the battery power of the drone is lower than the set threshold, the system will automatically adjust its flight path to return to the charging station for charging and reassign tasks to other drones.

[0037] II. Mission Planning Module

[0038] The mission planning module provides a visual interface where users can set flight paths, monitoring areas, and data collection parameters to generate mission instructions. It supports 3D map display, allowing users to directly draw flight paths and mark monitoring areas on the map. The system generates a 3D terrain model based on the map data. Users can intuitively plan the flight path of the drone on the 3D terrain model, set the starting point, ending point, and waypoints, and mark the monitoring areas. The system provides various mission parameter setting options, including flight altitude, speed, data collection frequency, etc. After the mission is generated, the system automatically optimizes and validates the mission to ensure its feasibility and safety.

[0039] In specific implementation, the mission planning module realizes the planning and optimization of the mission through the following steps:

[0040] 1. Map Data Import and 3D Modeling: The map data management unit supports the import of various map data formats, such as satellite imagery maps, terrain elevation data, etc. The system generates a 3D terrain model based on the imported map data, and users can visually view the terrain and landforms on the 3D terrain model. For example, users can clearly see terrain features such as mountains and valleys on the 3D terrain model through the imported elevation data.

[0041] 2. Flight Path Planning and Optimization: The path planning unit allows users to draw flight paths on the 3D terrain model and supports automatic path optimization algorithms. Users can set the starting point, ending point, and waypoints on the map by clicking with the mouse or using gesture operations. The system automatically generates the optimal flight path according to the terrain features and mission requirements. For example, when users plan a flight path in mountainous areas, the system will automatically avoid high-altitude peaks and select relatively gentle paths to reduce the energy consumption and flight risks of the drone.

[0042] 3. Task Parameter Setting and Verification: The task parameter setting unit provides rich task parameter setting options. Users can set parameters such as flight altitude, speed, and data acquisition frequency according to task requirements. The task optimization and verification unit optimizes and verifies the generated task to ensure the feasibility and safety of the task. For example, the system will automatically adjust the flight altitude and speed according to the performance parameters of the UAV and the terrain features of the task area to ensure that the UAV will not collide due to excessive speed or too low altitude during flight.

[0043] 4. Task Instruction Generation and Issuance: After the task planning is completed, the system generates task instructions and issues the task instructions to the corresponding UAV through the communication system. After receiving the task instructions, the UAV automatically starts the task execution program and flies and collects data according to the planned path and parameters.

[0044] III. Data Processing Module

[0045] The data processing module receives, stores, and analyzes the data transmitted in real time by the UAV, including images, temperature, humidity, and gas concentration, etc. It has edge computing capabilities and can perform data preprocessing on the UAV side to reduce the amount of data transmission. An edge computing device, such as a Raspberry Pi or NVIDIA Jetson, is installed on the UAV to perform real-time preprocessing on the collected data. The Faster R-CNN is used to perform real-time recognition on the image data to extract key features and reduce the amount of data. At the same time, the data is compressed to further reduce the amount of data transmission. At the ground station, the system performs fusion processing on the data from multiple UAVs to improve the integrity and accuracy of the data. The historical data is used to train and optimize the fire prediction model to improve the accuracy of the prediction.

[0046] In specific implementation, the data processing module realizes the preprocessing, fusion, and optimization of data through the following steps:

[0047] 1. Data Preprocessing: The edge computing unit performs real-time preprocessing on the data collected on the UAV side. For example, for image data, the Faster R-CNN algorithm is used to perform real-time recognition on the image to extract key features such as flames and smoke. For temperature, humidity, and gas concentration data, simple statistical analysis is performed to extract key information such as maximum value, minimum value, and average value. The preprocessed data is compressed by the data compression unit using compression algorithms such as JPEG or H.264 to reduce the amount of data transmission.

[0048] 2. Data Fusion Processing: At the ground station, the data fusion unit performs fusion processing on data from multiple UAVs. For example, for the image data collected by multiple UAVs in the same area, a multi-source data fusion algorithm is used to splice and fuse the images taken at different angles and times to generate more complete image information. For temperature, humidity, and gas concentration data, the weighted average method is used for fusion to improve the accuracy and reliability of the data.

[0049] 3. Fire Prediction Model Training and Optimization: The model training and optimization unit uses historical data to train and optimize the fire prediction model. The system collects meteorological data, vegetation data, and fire occurrence records over a period of time as training data. Machine learning algorithms such as support vector machine (SVM) or random forest are used to train the fire prediction model. By continuously optimizing the model parameters, the accuracy of fire prediction is improved. For example, the system can predict the probability of fire occurrence and the spreading direction based on historical data, providing a scientific basis for fire prevention and control.

[0050] IV. Hover Control Module

[0051] The hover control module enables UAVs without hover function to hover by optimizing the flight control system. A closed-loop control algorithm based on sensor feedback is adopted, combined with airflow compensation and attitude adjustment technologies, to ensure the UAV hovers stably in a complex environment. The UAV obtains real-time environmental information, including wind speed, wind direction, airflow, etc., through sensors such as lidar and cameras carried on it. The system performs real-time calculations and adjustments based on the environmental information, generates corresponding control commands, compensates according to the airflow information, and enables the UAV to hover stably at the specified position. PID control is used to perform real-time compensation according to the attitude change of the UAV to ensure the stability and accuracy of hovering.

[0052] In specific implementation, the hover control module realizes the stable hover of the UAV through the following steps:

[0053] 1. Environmental Information Collection: The sensor data collection unit obtains the sensor data carried on the UAV in real time. For example, lidar can measure the distance between the UAV and surrounding obstacles, the camera can obtain the visual information around the UAV, and the meteorological sensor can measure environmental parameters such as wind speed, wind direction, and airflow.

[0054] 2. Application of Closed-loop Control Algorithm: The closed-loop control algorithm unit performs real-time calculations and adjustments based on sensor data to generate control instructions. For example, when the sensor detects a strong wind speed, the system will adjust the flight attitude of the drone according to the wind speed and direction information to keep it stable. The PID control algorithm is adopted to perform real-time compensation according to the attitude change of the drone to ensure the stability and accuracy of hovering. For example, when the drone is affected by external interference and causes attitude deviation, the PID controller will quickly adjust the motor speed of the drone according to the deviation amount to make it return to the stable state.

[0055] 3. Airflow Compensation and Attitude Adjustment: The airflow compensation unit compensates the flight attitude of the drone according to the airflow information to improve the stability of hovering. The attitude adjustment unit adopts an advanced control algorithm to perform real-time compensation for the attitude change of the drone to ensure the accuracy of hovering. For example, when the drone hovers in a complex airflow environment such as a valley, the system will adjust the lift and attitude of the drone in real time according to the airflow change to enable it to hover stably at the designated position.

[0056] V. Emergency Response Module

[0057] When the emergency response module detects a forest fire, it automatically triggers an alarm, generates a fire extinguishing plan, and coordinates the drones to carry out fire extinguishing operations. The system monitors the data transmitted by the drones in real time and immediately triggers an alarm when fire characteristics are detected. According to information such as the location, scale, and fire intensity of the fire, the system generates a corresponding fire extinguishing plan, mobilizes drones with fire extinguishing capabilities to carry fire extinguishing agents to the fire scene for fire extinguishing operations. The system monitors the fire extinguishing progress and effect in real time and makes dynamic adjustments and optimizations according to the actual situation.

[0058] In specific implementation, the emergency response module realizes the rapid response and disposal of fires through the following steps:

[0059] 1. Fire Detection and Alarm Triggering: The fire detection unit analyzes the data transmitted by the drones in real time to detect fire characteristics. For example, by the flame characteristics in the image data, the abnormal high-temperature area in the temperature data, and the change in the concentration of harmful gases in the gas concentration data, the occurrence of a fire is judged. When fire characteristics are detected, the alarm triggering unit immediately triggers an alarm to notify relevant personnel.

[0060] 2. Fire Extinguishing Plan Generation: The fire extinguishing plan generation unit generates a reasonable fire extinguishing plan according to information such as the location, scale, and fire intensity of the fire. For example, the system will calculate the required amount of fire extinguishing agent and the number of drones according to the area and fire intensity of the fire. For small-area fires, the system can dispatch one or more rotor drones carrying fire extinguishing agents to go for fire extinguishing; for large-area fires, the system will dispatch multiple fixed-wing drones to carry out fire extinguishing agent delivery and coordinate the ground fire fighting forces for coordinated fire extinguishing.

[0061] 3. UAV Scheduling and Fire Extinguishing Operations: The UAV scheduling unit coordinates UAVs with fire extinguishing capabilities to perform fire extinguishing tasks. The system assigns task areas and flight paths to each UAV according to the fire extinguishing plan. The fire extinguishing operation monitoring unit monitors the fire extinguishing progress and effect in real time and makes dynamic adjustments and optimizations according to the actual situation. For example, when the fire extinguishing agent of a certain UAV runs out, the system will automatically dispatch other UAVs for replenishment; when the fire spread speed accelerates, the system will adjust the dropping frequency and flight path of the UAVs to ensure the fire extinguishing effect.

[0062] 4. Fire Extinguishing Effect Evaluation and Feedback: The system evaluates the fire extinguishing effect in real time through sensor data. For example, it detects the temperature change in the fire area through an infrared thermal imager to judge whether the fire is under control. According to the evaluation result of the fire extinguishing effect, the system will adjust the fire extinguishing plan in real time, optimize the UAV scheduling and operation strategies to ensure that the fire can be extinguished quickly and effectively.

[0063] VI. Communication System

[0064] The communication system supports multiple communication protocols, including Zigbee, Wi-Fi, and 4G / 5G, to ensure stable data transmission. During the communication process, the system automatically detects and switches the signal strength to ensure the stability and reliability of the communication. When the Wi-Fi signal is weak, the system automatically switches to the 4G / 5G network. The AES encryption and forward error correction coding technologies are adopted to ensure the security and integrity of data transmission.

[0065] In specific implementation, the communication system realizes stable and secure data transmission through the following steps:

[0066] 1. Communication Protocol Adaptation and Switching: The communication module management unit is responsible for managing and controlling multiple communication modules, supporting the switching and adaptation of multiple communication protocols. For example, when the UAV is flying in the mountainous area and the Wi-Fi signal is weak, the signal detection and switching unit will automatically detect the change in signal strength and switch the communication protocol to the 4G / 5G network to ensure the stability of data transmission.

[0067] 2. Data Encryption and Error Correction: The data encryption unit uses the AES encryption algorithm to encrypt the transmitted data to ensure data security. The error correction unit uses the forward error correction coding technology to improve the reliability of data transmission. For example, for important fire monitoring data, the system will use the high-encryption-strength AES-256 algorithm for encryption and add error correction codes during the data transmission process to ensure that the data will not be lost or damaged due to interference during transmission.

[0068] 3. Communication Status Monitoring and Feedback: The communication system monitors the communication status in real time, including parameters such as signal strength, data transmission rate, and packet loss rate. When the communication status is abnormal, the system will automatically adjust the communication parameters or switch the communication protocol to ensure the stability and reliability of communication. For example, when the detected data transmission rate drops, the system will automatically adjust the communication frequency or switch to a more stable communication network.

[0069] VII. Sensors

[0070] The sensors include a multispectral camera, an infrared thermal imager, a meteorological sensor, and a gas detector. The multispectral camera is used to obtain the spectral information of vegetation to help identify vegetation types and health conditions. The infrared thermal imager is used to detect fire sources and temperature changes to improve the accuracy of fire monitoring. The meteorological sensor is used to measure meteorological parameters such as wind speed, wind direction, temperature, and humidity to provide data support for fire prediction. The gas detector is used to detect the concentration of harmful gases in the air, such as carbon monoxide and carbon dioxide, to provide air quality information.

[0071] In specific implementation, the sensors achieve comprehensive monitoring of the forest environment through the following steps:

[0072] 1. Multispectral Camera Monitoring of Vegetation: The multispectral camera unit supports spectral acquisition in multiple bands and can obtain the spectral information of vegetation. For example, by analyzing the reflectance of vegetation in the red and near-infrared bands, the health condition of vegetation can be judged. Healthy vegetation has a higher reflectance in the near-infrared band, while damaged vegetation has a lower reflectance. The system generates a vegetation health index map based on the spectral data obtained by the multispectral camera, providing a reference for forest resource management and fire prevention.

[0073] 2. Infrared Thermal Imager Detection of Fire Sources: The infrared thermal imager unit can detect fire sources and temperature changes and provide high-resolution thermal images. For example, when a fire occurs, the infrared thermal imager can quickly detect the high-temperature anomaly in the fire area and determine the location and scope of the fire through image analysis algorithms. The system transmits the fire information detected by the infrared thermal imager to the ground station in real time, providing a basis for fire early warning and emergency response.

[0074] 3. Meteorological Sensor Measurement of Environmental Parameters: The meteorological sensor unit measures multiple meteorological parameters, including wind speed, wind direction, temperature, and humidity. For example, by measuring wind speed and wind direction, the system can predict the spread direction and speed of a fire. The meteorological sensor transmits the measured data to the ground station in real time, providing input data for the fire prediction model and improving the accuracy of fire prediction.

[0075] 4. Air quality monitoring by gas detector: The gas detector unit is used to detect the concentration of harmful gases in the air, such as carbon monoxide, carbon dioxide, etc. For example, during a fire, the gas detector can monitor the concentration of harmful gases in the fire area in real time, providing protection for the safety of firefighters. The system generates an air quality report based on the air quality data obtained by the gas detector, providing a basis for post-fire environmental assessment.

[0076] VIII. Flight control system

[0077] The flight control system has an intelligent obstacle avoidance function, which can sense the surrounding environment in real time and automatically avoid obstacles. The drone is equipped with sensors such as lidar and cameras to sense the surrounding environment in real time. The system performs real-time analysis and processing on the data obtained by the sensors to determine whether there are obstacles. When an obstacle is detected, the system automatically plans an obstacle avoidance path and generates corresponding control instructions. The Dijkstra path planning algorithm is used to optimize the flight path according to the task requirements and environmental information.

[0078] In specific implementation, the flight control system realizes intelligent obstacle avoidance and path planning through the following steps:

[0079] 1. Environmental perception and obstacle detection: The environmental perception unit obtains the surrounding environmental information in real time through sensors, providing data support for obstacle avoidance. For example, lidar can measure the distance between the drone and surrounding obstacles, and the camera can obtain the visual information of the obstacles. The obstacle detection unit uses deep learning algorithms to accurately detect the position and type of obstacles. For example, the images obtained by the camera are analyzed through a convolutional neural network (CNN) to identify obstacles such as trees and buildings.

[0080] 2. Obstacle avoidance path planning and control instruction generation: The path planning unit generates the optimal obstacle avoidance path according to the task requirements and environmental information. For example, when the drone detects an obstacle ahead during flight, the path planning unit will plan a safe obstacle avoidance path according to the position and shape of the obstacle, combined with the flight speed and task path of the drone. The control instruction generation unit generates corresponding control instructions according to the path planning result to guide the flight of the drone. For example, the control instructions will adjust the flight direction and speed of the drone to bypass the obstacle and continue to execute the task.

[0081] 3. Dynamic obstacle avoidance and path optimization: The flight control system has the ability of dynamic obstacle avoidance and can adjust the obstacle avoidance path in real time. For example, when the drone encounters a sudden obstacle during flight, the system will immediately re-plan the obstacle avoidance path and generate new control instructions. At the same time, the system will continuously optimize the flight path according to the real-time environmental information and task requirements to ensure the flight safety and task efficiency of the drone in a complex environment.

[0082] IX. Image Preprocessing Module

[0083] The image preprocessing module is used to preprocess the image data collected by the drone, including grayscale conversion and binarization operations, to improve the image quality and reduce the data volume. The grayscale conversion process converts a color image into a grayscale image by calculating the grayscale value of each pixel point using the weighted average method. The binarization process converts the grayscale value of the pixel points in the image into 0 or 255 by setting a threshold. The composition of the image preprocessing module includes an image acquisition unit, a grayscale conversion unit, and a binarization unit. The image acquisition unit obtains the original image data from the drone. The grayscale conversion unit calculates the grayscale value of each pixel point using the weighted average method and converts the color image into a grayscale image. The binarization unit converts the grayscale image into a binary image by setting a threshold to further simplify the data. The fire segmentation algorithm module uses spatial domain processing methods, including filter processing, Otsu image threshold segmentation algorithm, morphological operations, and flame feature extraction, to extract flame features from the preprocessed image.

[0084] In specific implementation, the image preprocessing module realizes the preprocessing of image data and flame feature extraction through the following steps:

[0085] 1. Image acquisition and grayscale conversion: The image acquisition unit obtains the original image data from the drone. The grayscale conversion unit calculates the grayscale value of each pixel point using the weighted average method and converts the color image into a grayscale image. For example, for an RGB color image, the grayscale conversion formula is: Gray = 0.299R + 0.587G + 0.114B. Through grayscale conversion, the image data volume is reduced while the main feature information of the image is retained.

[0086] 2. Binarization and image simplification: The binarization unit converts the grayscale image into a binary image by setting a threshold. For example, the global threshold method or the adaptive threshold method is used to divide the grayscale value of the pixel points in the grayscale image into two values, 0 and 255. The binarization process further simplifies the image data, facilitating subsequent image analysis and processing.

[0087] 3. Image denoising and edge enhancement: The filter processing unit uses a Gaussian filter, etc., to perform denoising and edge enhancement processing on the image. For example, the Gaussian filter can effectively remove the noise in the image while retaining the edge information of the image. Through denoising and edge enhancement processing, the features of the image become more obvious, providing a better basis for flame feature extraction.

[0088] 4. Flame Feature Extraction and Image Segmentation: The image threshold segmentation unit uses the Otsu algorithm to automatically determine the optimal threshold and segment the image into foreground and background. The morphological operation unit performs opening and closing operations to eliminate noise in the image and enhance edges. The flame feature extraction unit uses algorithms such as the watershed algorithm to extract flame features from the processed image. For example, the flame region is separated from the background through the watershed algorithm, and feature information such as the shape, size, and position of the flame is extracted. The system compares the extracted flame features with a preset flame feature library to determine whether it is a real flame, thus achieving accurate detection and segmentation of the flame.

[0089] X. Fire Segmentation Algorithm Module

[0090] The fire segmentation algorithm module adopts spatial domain processing methods, including filter processing, Otsu image threshold segmentation algorithm, morphological operations, and flame feature extraction, to extract flame features from the preprocessed image. The filter processing uses Gaussian filters, etc., to eliminate image noise and enhance edge details. The Otsu image threshold segmentation algorithm automatically determines the optimal threshold and segments the image into foreground and background. The morphological operations include opening and closing operations, which are used to eliminate noise in the segmented image and enhance edges. The flame feature extraction uses strategies such as the watershed algorithm to extract flame features from the processed image, achieving accurate segmentation of the flame.

[0091] In specific implementation, the fire segmentation algorithm module realizes the accurate extraction and segmentation of flame features through the following steps:

[0092] 1. Image Denoising and Edge Enhancement: The filter processing unit uses a Gaussian filter to denoise the image and at the same time adopts an edge enhancement algorithm, such as the Sobel operator or the Canny operator, to enhance the edge information in the image. For example, the Gaussian filter smooths the image through convolution operations to remove noise; the Sobel operator highlights edge features by calculating the gradient of the image. After denoising and edge enhancement processing, the flame edges in the image are clearer, facilitating subsequent segmentation operations.

[0093] 2. Image Threshold Segmentation: The image threshold segmentation unit uses the Otsu algorithm to automatically determine the optimal threshold and segment the image into foreground and background. The Otsu algorithm finds the threshold that maximizes the between-class variance of the foreground and background by calculating the gray-level histogram of the image. For example, for the grayscale image, the Otsu algorithm will automatically calculate a threshold to separate the flame region (foreground) from the background region in the image. Through image threshold segmentation, the flame region is initially extracted, providing a basis for further feature extraction.

[0094] 3. Morphological operation optimization: The morphological operation unit performs opening and closing operations to eliminate the noise in the segmented image and enhance the edges. The opening operation can remove small objects and connecting parts, and the closing operation can fill small holes and broken parts. For example, for the segmented flame image, the opening operation can remove small noise points in the flame area, and the closing operation can fill small holes at the flame edge, making the flame area more complete and continuous. After morphological operation optimization, the shape of the flame area is more accurate, facilitating subsequent feature extraction.

[0095] 4. Flame feature extraction and recognition: The flame feature extraction unit uses algorithms such as the watershed algorithm to extract flame features from the processed image. The watershed algorithm separates the flame area from the background by simulating the process of water flooding and extracts feature information such as the shape, size, and position of the flame. The system compares the extracted flame features with a preset flame feature library to determine whether it is a real flame. For example, by comparing the color, shape, and texture features of the flame, the system can accurately identify the flame area and issue a fire alarm. Through this series of image processing and analysis steps, the fire segmentation algorithm module can achieve accurate extraction and segmentation of the flame, providing reliable technical support for the early warning and rapid response of forest fires.

[0096] Through the specific implementation methods of the above modules, this system can achieve efficient monitoring, rapid early warning, and effective disposal of forest fires, improving the intelligent level and emergency response ability of forest fire prevention work.

[0097] In addition, it should be noted that the present invention can be provided as a method, apparatus, or computer program product. Therefore, the embodiments of the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the embodiments of the present invention can take the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program code.

[0098] The embodiments of the present invention are described with reference to the flowcharts and / or block diagrams of methods, terminal devices (systems), and computer program products according to the embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, and the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, an embedded processor, or other programmable data processing terminal device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing terminal device generate a device for implementing the functions specified in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0099] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing terminal device to work in a specific manner, such that the instructions stored in the computer-readable memory produce a manufacture including an instruction device that implements the functions specified in one process Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks. These computer program instructions can also be loaded onto a computer or other programmable data processing terminal device, such that a series of operational steps are executed on the computer or other programmable terminal device to produce a computer-implemented process, thereby providing steps for implementing the functions specified in one process Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks on the computer or other programmable terminal device.

[0100] It should also be noted that in this text, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. The term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, such that a process, method, article or terminal device comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article or terminal device. Without further limitation, an element defined by the statement "comprising an..." does not exclude the presence of additional identical elements in the process, method, article or terminal device comprising the element.

[0101] Finally, it should be noted that the above is the preferred embodiment of the present invention. It should be pointed out that although the preferred embodiments of the present invention have been described, for those skilled in the art of this technology, once the basic creative concept of the present invention is known, several improvements and refinements can be made without departing from the principle of the present invention, and these improvements and refinements should also be regarded as the protection scope of the present invention. Therefore, the appended claims are intended to be construed as including the preferred embodiments and all changes and modifications falling within the scope of the embodiments of the present invention.

Claims

1. An intelligent forest fire prevention system based on drones, characterized in that, It includes: UAV cluster module: Supports the access of various types of UAVs, covering fixed-wing, rotary-wing, and unmanned airships, etc. Each UAV has a unique identifier and can be seamlessly docked with the system to achieve plug-and-play. Mission planning module: Provides a visual interface. Users can set flight paths, monitoring areas, and data acquisition parameters, generate mission instructions, and support 3D map display. Users can directly draw flight paths and mark monitoring areas on the map. Data processing module: Receives, stores, and analyzes the real-time data transmitted by UAVs, including images, temperature, humidity, and gas concentration, etc. It has edge computing capabilities and can perform data preprocessing at the UAV end to reduce the amount of data transmission. Hover control module: Enables UAVs without hover capabilities to hover by optimizing the flight control system. It adopts a closed-loop control algorithm based on sensor feedback, combined with airflow compensation and attitude adjustment technologies to ensure the stable hover of UAVs in complex environments. Emergency response module: When a forest fire is detected, it automatically triggers an alarm, generates a fire extinguishing plan, and coordinates UAVs to perform fire extinguishing operations. Communication system: Supports multiple communication protocols, including Zigbee, Wi-Fi, and 4G / 5G, to ensure stable data transmission. Sensors: Include multispectral cameras, infrared thermal imagers, meteorological sensors, and gas detectors. Flight control system: Has an intelligent obstacle avoidance function and can perceive the surrounding environment in real time and automatically avoid obstacles. Data processing module: Has machine learning capabilities and can analyze historical data to optimize the fire prediction model. Image preprocessing module: Performs preprocessing on the image data collected by UAVs, including grayscale conversion, binarization, etc., to improve image quality and reduce the amount of data. Fire segmentation algorithm module: Extracts flame features from the preprocessed images, using spatial domain processing methods, including filter processing, Otsu image threshold segmentation algorithm, morphological operations, and flame feature extraction.

2. The intelligent forest fire prevention system based on unmanned aerial vehicle according to claim 1, characterized in that The UAV cluster module supports the access of any type of UAV, including fixed-wing, rotary-wing, and unmanned airships, etc. Each UAV has a unique identifier. The system supports Zigbee, Wi-Fi, 4G / 5G communication protocols. UAVs are connected to the system through these protocols. When a UAV accesses the system, the system automatically identifies the type and parameters of the UAV and assigns corresponding tasks and resources to it. During the task execution process, the system monitors the status and data transmission of the UAV in real time to ensure the smooth progress of the task. Users can flexibly select and replace UAVs without reconfiguring the entire system. The composition of the UAV cluster module includes a communication interface unit, a UAV identification and configuration unit, a task assignment unit, and a status monitoring unit. The communication interface unit is responsible for establishing communication connections with UAVs and supports the conversion and adaptation of multiple communication protocols. The UAV identification and configuration unit automatically identifies the types and parameters of the accessed UAVs, assigns unique identifiers and corresponding task configurations to them. The task assignment unit reasonably assigns tasks to UAVs according to the task requirements set by the user, optimizing task scheduling. The status monitoring unit monitors the flight status, battery power, and data transmission information of UAVs in real time to ensure the smooth execution of tasks.

3. The intelligent forest fire prevention system based on unmanned aerial vehicles according to claim 1, wherein, The visualization interface of the task planning module supports 3D map display. Users can directly draw flight paths and mark monitoring areas on the map. When users select a corresponding map area on the visualization interface, the system generates a 3D terrain model based on the map data. Users can intuitively plan the flight paths of UAVs on the 3D terrain model, set the starting point, ending point, and waypoints, and mark the monitoring areas. The system provides various task parameter setting options, including flight altitude, speed, and data acquisition frequency. After the task is generated, the system automatically optimizes and verifies the task to ensure its feasibility and safety. The composition of the task planning module includes a map data management unit, a 3D terrain modeling unit, a path planning unit, a task parameter setting unit, and a task optimization and verification unit. The map data management unit is responsible for storing and managing various map data, supporting the import and conversion of multiple map data formats. The 3D terrain modeling unit generates a 3D terrain model based on the map data, providing an intuitive visual effect. The path planning unit allows users to draw flight paths on the 3D terrain model, supports automatic path optimization algorithms, and generates the optimal flight path. The task parameter setting unit provides rich task parameter setting options to meet the task requirements in different scenarios. The task optimization and verification unit optimizes and verifies the generated task to ensure its feasibility and safety.

4. The intelligent forest fire prevention system based on an unmanned aerial vehicle according to claim 1, wherein, The data processing module has edge computing capabilities and can perform data preprocessing on the UAV side to reduce the amount of data transmission. An edge computing device, such as a Raspberry Pi or NVIDIA Jetson, is installed on the UAV to perform real-time preprocessing on the collected data. Faster R-CNN is used to perform real-time recognition on image data, extract key features, and reduce the amount of data. At the same time, the data is compressed to further reduce the amount of data transmission. At the ground station, the system performs fusion processing on data from multiple UAVs to improve the integrity and accuracy of the data; uses historical data to train and optimize the fire prediction model to improve the accuracy of prediction. The composition of the data processing module includes an edge computing unit, an image recognition unit, a data compression unit, a data fusion unit, and a model training and optimization unit. The edge computing unit performs real-time preprocessing of data on the UAV side to improve data processing efficiency. The image recognition unit uses deep learning algorithms to perform real-time recognition and analysis of image data. The data compression unit compresses the preprocessed data to reduce the amount of data transmission. The data fusion unit performs fusion processing on data from multiple UAVs at the ground station to improve the integrity and accuracy of the data. The model training and optimization unit uses historical data to train and optimize the fire prediction model to improve the accuracy of prediction.

5. The intelligent forest fire prevention system based on an unmanned aerial vehicle according to claim 1, characterized in that, The hovering control module adopts a closed-loop control algorithm based on sensor feedback, combined with airflow compensation and attitude adjustment technologies, to ensure the stable hovering of the UAV in complex environments; The UAV obtains environmental information in real time through the sensors it carries, and the sensors include lidar and cameras. The environmental information includes wind speed, wind direction, and airflow; The system performs real-time calculations and adjustments based on the environmental information, generates corresponding control commands, compensates according to the airflow information, enables the UAV to stably hover at the specified position, and uses PID control to perform real-time compensation according to the attitude changes of the UAV to ensure the stability and accuracy of hovering; The composition of the hovering control module includes a sensor data acquisition unit, a closed-loop control algorithm unit, an airflow compensation unit, and an attitude adjustment unit. The sensor data acquisition unit obtains the data of the sensors carried by the UAV in real time to provide environmental information for hovering control. The closed-loop control algorithm unit performs real-time calculations and adjustments based on the sensor data to generate control commands. The airflow compensation unit compensates the flight attitude of the UAV according to the airflow information to improve the stability of hovering. The attitude adjustment unit uses advanced control algorithms to perform real-time compensation for the attitude changes of the UAV to ensure the accuracy of hovering.

6. The intelligent forest fire prevention system based on an unmanned aerial vehicle according to claim 1, wherein For the emergency response module, when a forest fire is detected, it automatically triggers an alarm, generates a fire extinguishing plan, and coordinates the UAVs to carry out fire extinguishing operations. The system monitors the data transmitted by the UAVs in real time. When fire characteristics are detected, it immediately triggers an alarm. According to information such as the location, scale, and fire intensity of the fire, the system generates a corresponding fire extinguishing plan, mobilizes the UAVs with fire extinguishing capabilities to carry fire extinguishing agents to the fire scene, and conducts fire extinguishing operations. The system monitors the progress and effect of fire extinguishing in real time and makes dynamic adjustments and optimizations according to the actual situation; The composition of the emergency response module includes a fire detection unit, an alarm trigger unit, a fire extinguishing plan generation unit, a UAV scheduling unit, and a fire extinguishing operation monitoring unit. The fire detection unit analyzes the data transmitted by the UAVs in real time to detect fire characteristics. The alarm trigger unit immediately triggers an alarm when a fire is detected to notify relevant personnel. The fire extinguishing plan generation unit generates a reasonable fire extinguishing plan according to the fire information. The UAV scheduling unit coordinates the UAVs with fire extinguishing capabilities to execute fire extinguishing tasks. The fire extinguishing operation monitoring unit monitors the progress and effect of fire extinguishing in real time and makes dynamic adjustments and optimizations.

7. The intelligent forest fire prevention system based on an unmanned aerial vehicle according to claim 1, characterized in that, The communication system supports multiple communication protocols, including Zigbee, Wi-Fi, and 4G / 5G, ensuring stable data transmission. During the communication process, the system automatically detects and switches the signal strength to ensure the stability and reliability of communication. When the Wi-Fi signal is weak, the system automatically switches to the 4G / 5G network. It adopts AES encryption and forward error correction coding technologies to ensure the security and integrity of data transmission; The communication system consists of a communication module management unit, a signal detection and switching unit, a data encryption unit, and an error correction unit. The communication module management unit is responsible for managing and controlling multiple communication modules, supporting the switching and adaptation of multiple communication protocols. The signal detection and switching unit real-time detects the communication signal strength and automatically switches to the best communication network. The data encryption unit encrypts the transmitted data using encryption algorithms to ensure data security. The error correction unit adopts error correction coding technologies to improve the reliability of data transmission.

8. The intelligent forest fire prevention system based on an unmanned aerial vehicle according to claim 1, characterized in that The sensors include, but are not limited to, multispectral cameras, infrared thermal imagers, meteorological sensors, and gas detectors. The multispectral camera is used to obtain the spectral information of vegetation to help identify vegetation types and health conditions. The infrared thermal imager is used to detect fire sources and temperature changes to improve the accuracy of fire monitoring. The meteorological sensor is used to measure meteorological parameters such as wind speed, wind direction, temperature, and humidity to provide data support for fire prediction. The gas detector is used to detect the concentration of harmful gases in the air, such as carbon monoxide and carbon dioxide, to provide air quality information; The sensors consist of a multispectral camera unit, an infrared thermal imager unit, a meteorological sensor unit, and a gas detector unit. The multispectral camera unit obtains the spectral information of vegetation and supports spectral collection in multiple bands. The infrared thermal imager unit detects fire sources and temperature changes and provides high-resolution thermal images. The meteorological sensor unit measures multiple meteorological parameters and provides real-time meteorological data. The gas detector unit detects the concentration of harmful gases in the air and supports the detection and analysis of multiple gases.

9. The intelligent forest fire prevention system based on an unmanned aerial vehicle according to claim 1, characterized in that, The flight control system has an intelligent obstacle avoidance function, which can perceive the surrounding environment in real time and automatically avoid obstacles. The unmanned aerial vehicle is equipped with sensors such as lidar and cameras to perceive the surrounding environment in real time. The system performs real-time analysis and processing on the data obtained by the sensors to determine whether there are obstacles. When an obstacle is detected, the system automatically plans an obstacle avoidance path and generates corresponding control instructions. It adopts the Dijkstra path planning algorithm to optimize the flight path according to the task requirements and environmental information; The flight control system consists of an environment perception unit, an obstacle detection unit, a path planning unit, and a control instruction generation unit. The environment perception unit obtains the surrounding environment information in real time through sensors to provide data support for obstacle avoidance. The obstacle detection unit uses deep learning algorithms to accurately detect the position and type of obstacles. The path planning unit generates the optimal obstacle avoidance path according to the task requirements and environmental information. The control instruction generation unit generates corresponding control instructions according to the path planning results to guide the flight of the unmanned aerial vehicle.

10. The intelligent forest fire prevention system based on unmanned aerial vehicle according to claim 1, wherein, The image preprocessing module is used to preprocess the image data collected by the drone, including grayscale conversion and binarization operations, to improve the image quality and reduce the data volume; The grayscale conversion process converts a color image into a grayscale image by calculating the grayscale value of each pixel point using the weighted average method. The binarization process converts the grayscale value of the pixel points in the image into 0 or 255 by setting a threshold; The image preprocessing module consists of an image acquisition unit, a grayscale conversion unit, and a binarization unit. The image acquisition unit obtains the original image data from the drone. The grayscale conversion unit calculates the grayscale value of each pixel point using the weighted average method and converts the color image into a grayscale image. The binarization unit converts the grayscale image into a binary image by setting a threshold to further simplify the data. The fire segmentation algorithm module uses spatial domain processing methods, including filter processing, Otsu image threshold segmentation algorithm, morphological operations, and flame feature extraction, to extract flame features from the preprocessed image. Filter processing uses a Gaussian filter, etc., to eliminate image noise and enhance edge details. The Otsu image threshold segmentation algorithm automatically determines the optimal threshold to segment the image into foreground and background. Morphological operations include opening and closing operations, which are used to eliminate the noise in the segmented image and enhance the edges. Flame feature extraction uses strategies such as the watershed algorithm to extract flame features from the processed image; The fire segmentation algorithm module consists of a filter processing unit, an image threshold segmentation unit, a morphological operation unit, and a flame feature extraction unit. The filter processing unit uses a Gaussian filter, etc., to perform denoising and edge enhancement processing on the image. The image threshold segmentation unit uses the Otsu algorithm to automatically determine the optimal threshold to segment the image into foreground and background. The morphological operation unit performs opening and closing operations to eliminate the noise in the image and enhance the edges. The flame feature extraction unit uses the watershed algorithm, etc., to extract flame features from the processed image to achieve accurate segmentation of the flame.

Images