Intelligent fire-fighting trolley system based on multi-module cooperation

Abstract

This invention discloses an intelligent fire truck system based on multi-module collaboration, including a vehicle body and integrated on the vehicle body a main control module, an image acquisition module, a fire extinguishing module, a fire source detection module, a motor drive module, and a drive motor. The fire source detection module detects the presence of fire sources in the surrounding area and returns the acquired fire source signals to the main control module for judgment and location. The image acquisition module assists in acquiring image information related to the fire scene environment and fire sources, feeding it back to the main control module to optimize positioning accuracy. The fire extinguishing module receives instructions from the main control module and performs initial fire extinguishing operations. The main control module receives information collected by each module and feedback on its working status, processes the signals, sends them to designated modules, and controls the truck to perform corresponding actions. The collaborative operation of all modules in this invention enables automatic detection, location, and extinguishing of fire sources, with reliable performance and wide applicability to initial fire prevention in unattended scenarios.

Description

Technical Field

[0001] This invention relates to the field of intelligent fire trucks, and particularly to an intelligent fire truck system based on multi-module collaboration. Background Technology

[0002] The use of fire has permeated the entire process of human societal development, but its improper use has led to frequent fires, posing a serious threat to human life, property, and the natural environment. Following the Second Industrial Revolution, the widespread availability of electricity further increased the probability of fires. To address this issue, humanity has invested significant human and material resources in developing fire protection engineering, and the advancement of fire protection technology has, to some extent, reduced the major losses caused by fires.

[0003] With the advent of the information age, the drawbacks of traditional firefighting models, which rely excessively on manpower, have become increasingly apparent. This not only results in enormous resource consumption, but also poses significant safety hazards in key fire-fighting units such as chemical industrial parks and hazardous materials warehouses due to the presence of large quantities of flammable and explosive materials, requiring long-term personnel deployment. Therefore, the transformation of firefighting towards intelligent systems has become an inevitable trend. Against this backdrop, intelligent firefighting mobile devices have emerged. Among them, mini intelligent fire trucks, with their flexibility and mobility, can perform functions such as fire source detection, automatic tracking, obstacle avoidance, fire extinguishing, and fire early warning. They can contain fires in their early stages, effectively saving social resources and becoming a research hotspot in the firefighting field.

[0004] Currently, the rapid development of high-tech fields such as microcomputer technology, automatic control technology, and intelligent AI has provided support for the technological upgrade of mini intelligent fire trucks. However, the development level of this technology varies globally, with core technological bottlenecks concentrated in two main areas: accurate fire source detection and automatic tracking. Existing fire source detection technologies mostly use single sensors for data collection, coupled with relatively outdated processing algorithms, resulting in low detection accuracy, frequent false alarms and missed alarms, seriously affecting the timeliness of rescue operations. Meanwhile, related technologies abroad started earlier, have higher technological maturity, and have long held a leading position, meaning that my country still faces significant challenges in technological research and development in this field.

[0005] From the perspective of actual fire protection needs, the fire prevention and control situation remains severe. According to statistics, from January to June 2022, my country received 449,000 reports of fires, resulting in 1,025 deaths and 3.31 billion yuan in direct property losses. The harm caused by fires cannot be ignored. At the same time, the scope of rescue missions undertaken by fire and rescue teams is constantly expanding, often requiring them to deal with harsh environments such as high altitudes, underground locations, high temperatures, and smoky heat. The operational scenarios faced by traditional fire fighting methods are becoming increasingly complex, and the dangers and difficulties of rescue operations are continuously increasing. The safety risks for firefighters entering fire scenes are difficult to effectively avoid.

[0006] In response, the Chinese government has increased its investment in the field of intelligent firefighting and actively encouraged researchers to conduct relevant theoretical research and technological innovation. However, existing mini intelligent fire trucks still cannot fully meet the actual rescue needs of complex fire scenes, and there is still considerable room for improvement in terms of detection accuracy, environmental adaptability, and rescue efficiency. Therefore, carrying out innovative designs for mini intelligent fire trucks used for firefighting and rescue, addressing the pain points of existing technologies, reducing the frequency of firefighters entering fire scenes, and improving the efficiency and safety of fire scene rescue are of significant practical importance and application value. Summary of the Invention

[0007] The purpose of this invention is to overcome at least one of the shortcomings of the prior art and provide an intelligent fire truck system based on multi-module collaboration, which realizes automatic fire source detection, accurate positioning and rapid fire extinguishing, and solves the problems of large detection blind spots, inaccurate positioning and delayed response of existing fire trucks. It is suitable for initial fire prevention and control in unattended scenarios.

[0008] To achieve the above objectives, the technical solution adopted by the present invention is as follows: An intelligent fire truck system based on multi-module collaboration includes a vehicle body, and a main control module, a power supply module, an image acquisition module, a fire extinguishing module, a fire source detection module, a motor drive module, and a drive motor integrated on the vehicle body.

[0009] The main control module, as the core of the system, is electrically connected to the image acquisition module, fire extinguishing module, fire source detection module, and motor drive module. It is used to receive information collected by each module and feedback on the working status, process the signals and send control commands to control the trolley to complete corresponding actions such as fire source positioning, movement and fire extinguishing.

[0010] The fire source detection module employs multiple high-sensitivity flame sensors, positioned at the front and sides of the vehicle body, to achieve 360° fire source detection with no blind spots and a wide angle. The collected fire source signals are returned to the main control module to determine the presence of the fire source and assist in its location.

[0011] The image acquisition module is used to acquire real-time images of the fire environment and fire source-related information around the vehicle body, and feed them back to the main control module. It integrates fire source detection signals to optimize fire source positioning accuracy and avoid positioning deviation.

[0012] The motor drive module is electrically connected to the drive motor and is used to convert the control signal of the main control module into a power signal to drive the drive motor to run, thereby controlling the forward, backward, and steering actions of the vehicle.

[0013] The fire extinguishing module is used to receive fire extinguishing commands from the main control module, execute initial fire extinguishing operations, and contain the fire in its early stages.

[0014] The power module provides a stable DC power supply to the entire device, ensuring the continuous and stable operation of the main control module, image acquisition module, fire extinguishing module, fire source detection module, motor drive module and drive motor, while also providing both battery life and portability.

[0015] The power module includes a rechargeable lithium battery. The power module provides a stable DC power supply to the entire device, ensuring the continuous and stable operation of the main control module, image acquisition module, fire extinguishing module, fire source detection module, motor drive module, and drive motor, while also offering both long battery life and portability.

[0016] The present invention also provides a fire extinguishing method.

[0017] This invention achieves comprehensive fire source detection and timely and accurate fire suppression: the fire source detection module adopts a multi-sensor layout to achieve 360° coverage without blind spots, enabling 24-hour continuous detection, effectively avoiding detection blind spots, and accurately capturing initial sporadic fire sources; the main control module integrates image acquisition and fire source detection signals to achieve precise fire source location and plan the optimal path; the motor drive module drives the vehicle to quickly arrive at the scene, and the fire suppression module promptly executes fire suppression operations to minimize fire losses.

[0018] Compared with the prior art, the present invention has the following advantages: 1. No blind spots in detection: It adopts a multi-sensor 360° layout for all-round fire source detection, reducing the rate of missed and false alarms, and can capture initial sporadic fire sources.

[0019] 2. High positioning accuracy: By integrating image information and fire source signals, the positioning is accurate and the path is reasonable, thus improving fire extinguishing efficiency.

[0020] 3. Fast response speed: The system is highly responsive, quickly identifies, moves, and responds, buying time for initial firefighting.

[0021] 4. Wide range of applications: Modular integration, simple structure, and reliable performance make it suitable for unattended high-risk scenarios such as warehouses, computer rooms, power distribution rooms, and underground pipe corridors.

[0022] 5. High safety: It replaces manual entry into dangerous areas, reduces the safety risks for firefighters, reduces manpower input, and saves fire-fighting resources. Attached Figure Description

[0023] Figure 1 This is a schematic diagram of the composition of an intelligent fire truck system based on multi-module collaboration in an embodiment of the present invention.

[0024] Figure 2 This is a schematic diagram illustrating the coordination of the various modules in an embodiment of the present invention.

[0025] Figure 3This is a schematic diagram of the microcontroller interface in an embodiment of the present invention. Detailed Implementation

[0026] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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, 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.

[0027] The present invention provides an intelligent fire truck system based on multi-module collaboration, as shown in Figure 1, which includes a vehicle body and a main control module, a power supply module, an image acquisition module, a fire extinguishing module, a fire source detection module, a motor drive module, and a drive motor integrated on the vehicle body.

[0028] The image acquisition module, fire extinguishing module, fire source detection module, and motor drive module are electrically connected to the main control module via wires, receiving commands from the main control module and providing feedback on their operating status. The motor drive module is electrically connected to the drive motor, converting the control signals from the main control module into power signals to drive the drive motor, thereby controlling the vehicle's forward, reverse, and steering movements. The power supply module provides a stable DC power supply to the entire device, ensuring the continuous and stable operation of all modules.

[0029] The drive motor is a DC motor.

[0030] The fire source detection module includes multiple high-sensitivity infrared flame sensors to detect the presence of fire sources in the surrounding area and transmit the fire source signals to the main control module in real time. These sensors are positioned at the front, sides, and rear of the vehicle body, creating a 360° omnidirectional coverage area for comprehensive fire source detection of the vehicle's surroundings. During combustion, flames emit characteristic infrared light with a center wavelength of approximately 4.3μm. The infrared flame sensors transmit the collected characteristic infrared light intensity signals to the main control module in real time. The main control module has a built-in AD conversion unit that converts the analog voltage signals input from the infrared flame sensors into digital signals. A fire source light intensity signal judgment threshold is preset. When the collected light intensity signal value exceeds the preset threshold, a fire source is determined to exist in the corresponding detection direction. The location of the infrared flame sensor with the highest signal strength determines the initial azimuth angle range of the fire source. Simultaneously, based on the signal intensity attenuation gradient, the distance range between the fire source and the vehicle body is preliminarily determined, completing the initial coarse location of the fire source.

[0031] In one embodiment, the light intensity signal output by the infrared flame sensor exhibits an exponential decay relationship with the distance to the fire source:

[0032] In the formula, I is the light intensity signal output by the infrared flame sensor in real time; I0 is the light intensity of a nearby standard fire source; k is the environmental attenuation coefficient; and d is the distance between the fire source and the infrared flame sensor.

[0033] The image acquisition module includes a camera, which is used to acquire image information of the fire source location and the fire scene environment in real time, and transmit the image data to the main control module.

[0034] To achieve the fusion of fire source signals and image information: First, using the initial azimuth angle range and distance range of the fire source output by the fire source detection module as constraints, and based on the vehicle's heading angle and the camera's field of view, the initial azimuth angle range of the fire source is mapped to the column coordinates of the image, and the distance range is mapped to the row coordinate range of the image. The corresponding rectangular target detection area is then delineated in the image, thereby narrowing the range of image recognition. Subsequently, the main control module identifies the image within the defined target detection area using the built-in fire source recognition model, and extracts feature information such as the pixel coordinates and contour size of the fire source. Next, combining the camera's intrinsic parameters, installation height, and pitch angle, a monocular vision ranging algorithm is used to convert the fire source pixel coordinates extracted from the image into spatial coordinates in the vehicle coordinate system, obtaining an estimated fire source location based on image information. Then, using the monocular vision ranging algorithm, the fire source pixel coordinates extracted from the image are combined with the camera's intrinsic parameters, installation height, pitch angle, and field of view parameters to convert them into three-dimensional spatial coordinates in the vehicle coordinate system, simultaneously outputting the azimuth angle and straight-line distance values ​​of the fire source, thus completing the image-based fire source distance calculation. Simultaneously, the light intensity gradient data output by the fire source detection module also provides another independent distance and azimuth estimate. Based on the signal intensity distribution of multiple sensors, the centroid algorithm and triangulation principle are used to collect the light intensity signals output by multiple flame sensors and construct a light intensity distribution matrix. Using the position of each flame sensor as the base point and the light intensity as the weight, the initial centroid coordinates of the fire source are calculated using the centroid algorithm to determine the azimuth estimate. Combining the signal intensity difference and installation angle of adjacent flame sensors, the distance estimate between the fire source and the vehicle body is calculated based on the triangulation principle. A unique value is output within the initial azimuth angle and distance range, and the fire source azimuth estimate and distance estimate are determined within the initial azimuth angle and distance range of the fire source. To optimize positioning, the two sets of information are weighted and fused to effectively compensate for the errors in monocular visual ranging, improve positioning accuracy, and achieve positioning optimization.

[0035] In one embodiment, the fire source identification model employs an improved YOLO lightweight target detection network. Based on the original YOLO lightweight network, the prior box size and non-maximum suppression strategy are optimized, and an infrared feature channel for flames and a small target detection head are added. The infrared feature channel for flames is used to match the 4.3μm characteristic infrared band of flames, enhancing flame feature extraction capabilities, reducing interference from complex environments, and improving flame identification accuracy. The small target detection head optimizes the detection receptive field for the small pixel size of initial sporadic fire sources, improving the detection rate and positioning accuracy of small flame targets. This improved YOLO lightweight target detection network can improve the identification rate and positioning accuracy of small-sized fire sources in complex environments.

[0036] In one embodiment, the two information streams are weighted and fused together. Based on the spatial coordinates obtained from monocular vision transformation, the signal strength gradient data from the fire source detection module is introduced as a correction term. The two streams are weighted and calculated using preset weights. The fusion formula is as follows:

[0037] In the formula: P is the final spatial location of the fire source; P vis For visual localization results; P sen This is the location result from the flame sensor; w1 and w2 are weighting coefficients that are adaptively adjusted based on the environment and distance, and w1 + w2 = 1. The location result includes both the fire source azimuth and straight-line distance, providing three-dimensional spatial positioning information that can be directly used for path planning and fire suppression execution.

[0038] The main control module is used to receive and process the signals input by the fire source detection module and the image acquisition module, complete the fire source judgment, location positioning and path planning based on the A* algorithm, and output corresponding control commands to the motor drive module to control the trolley to complete the corresponding actions such as forward movement, driving and turning, and output corresponding fire extinguishing commands to the fire extinguishing module to execute the fire extinguishing action.

[0039] In one embodiment, the fire extinguishing module employs a jet-type fire extinguishing mechanism consisting of a water pump, a water tank, and a spray pipe. The water tank is filled with a water-based extinguishing agent, and the spray pipe is sequentially connected to the water pump and the water tank. The control terminal of the water pump is connected to the main control module. The fire extinguishing module receives the fire extinguishing command from the main control module, executes initial fire suppression, and extinguishes the fire source. This fire truck system is mainly used for early fire monitoring and suppression in unattended scenarios. It can quickly extinguish small-scale fires with low temperatures and limited spread, preventing the fire from expanding into large-scale open flames, dense smoke, and high temperatures, thereby reducing the difficulty of suppression and equipment load, and improving the fire extinguishing success rate and system safety.

[0040] In one embodiment, the power module uses a rechargeable lithium battery and a voltage regulator circuit, which features stable battery life, portability and ease of use, and is suitable for long-term inspection in unattended scenarios.

[0041] In one embodiment, the motor drive module uses the L298N, and the main control module includes a microcontroller. Table 1 shows the control logic diagram of the L298N motor drive module. IN1 and IN2 are connected to the input pins of motor driver A, controlling the rotation and angle of motor A. A high level (HIGH) on IN1 and a low level (LOW) on IN2 correspond to forward rotation of motor A; a low level (LOW) on IN1 and a high level (HIGH) on IN2 correspond to reverse rotation of motor A; and both IN1 and IN2 simultaneously being either high (HIGH) or low (LOW) correspond to stopping motor A. Rotation speed adjustment involves changing the duty cycle of the high levels on IN1 and IN2. IN3 and IN4 are connected to the input pins of motor driver B, controlling the rotation and angle of motor B. A high level (HIGH) input on IN3 and a low level (LOW) input on IN4 correspond to forward rotation of drive motor B. A low level (LOW) input on IN3 and a high level (HIGH) input on IN4 correspond to reverse rotation of drive motor B. Simultaneous high (HIGH) or low (LOW) inputs on both IN3 and IN4 correspond to stopping drive motor B. Rotation speed is adjusted by changing the duty cycle of the high level inputs on IN3 and IN4. Two drive motors, A and B, are connected between OUT1 and OUT2, and between OUT3 and OUT4, respectively. Drive motor A drives the left rear wheel of the vehicle, and drive motor B drives the right rear wheel. The forward, backward, and turning movements of the vehicle are achieved through the speed difference between the two drive motors and coordination with the steering mechanism. When both drive motors rotate synchronously in the forward direction, the vehicle moves forward; when both drive motors rotate synchronously in the reverse direction, the vehicle moves backward; when the speed of drive motor A is lower than that of drive motor B, the vehicle turns left; when the speed of drive motor A is higher than that of drive motor B, the vehicle turns right; when the two drive motors rotate in opposite directions, the vehicle achieves a stationary turn. The IN1, IN2, IN3, and IN4 pins are connected to the microcontroller's control level. The IN1, IN2, IN3, and IN4 pins of the L298N motor drive module are connected to the microcontroller's PB0, PB1, PB5, and PB4 pins respectively to control the motor's forward and reverse rotation. The ENA and ENB pins are connected to the microcontroller's PA6 and PA7 pins respectively. The OUT1, OUT2, OUT3, and OUT4 pins of the motor drive module are connected to drive motor A (driving the left rear wheel) and drive motor B (driving the right rear wheel) respectively. The six circumferentially arranged flame sensors of the fire source detection module are connected to the microcontroller's PA0, PA1, PA2, PA3, PC0, and PC1 pins respectively. The OV7670 camera's D0-D7 data pins of the image acquisition module are connected to the microcontroller's PD0-PD7 pins. The SDA and SCL pins are connected to the microcontroller's PB6 and PB7 pins, respectively. The VSYNC, HREF, and PCLK pins are connected to the microcontroller's PB8, PB9, and PA8 pins, respectively.

[0042]

[0043] The workflow of this invention is as follows: After the system is powered on, it automatically initializes, and the vehicle enters the inspection state. The fire source detection module collects environmental data in real time at a 360° angle, and the image acquisition module simultaneously acquires environmental images. When no fire source is detected, the vehicle inspects along a preset route; upon detecting a fire source, the main control module integrates the two signals to achieve precise positioning and plan a path, then controls the motor drive module to drive the drive motor, enabling the vehicle to quickly move towards the fire source. Upon reaching the fire source, the main control module activates the fire extinguishing module to perform the fire extinguishing operation. After the fire is extinguished, the system confirms that the fire is eliminated, and the vehicle resumes the inspection state, thus achieving 24 / 7 automated fire prevention and control in unattended scenarios.

[0044] In one embodiment, a fire extinguishing method is provided, comprising the following steps: Step 1: After the system is powered on, it will automatically initialize and the vehicle will enter the inspection state; Step 2: Collect the light intensity signal of the fire source through the fire source detection module, and collect image information of the fire source location and fire scene environment through the image acquisition module; Step 3: The main control module receives the fire source light intensity signal, and determines whether there is a fire source in the corresponding detection direction based on the fire source light intensity signal and the preset fire source light intensity signal determination threshold. It also determines the initial azimuth angle range of the fire source and the distance range between the fire source and the vehicle body based on the signal strength. Step 4: After determining that there is a fire source, the main control module obtains the final spatial location of the fire source by weighted fusion of the fire source azimuth estimate and distance estimate determined based on the initial azimuth range of the fire source and the distance range between the fire source and the vehicle body, as well as the fire source distance and fire source azimuth estimate obtained from the image information of the fire scene environment through image recognition and monocular vision ranging. Step 5: The main control module plans the travel path based on the final fire source location and sends control commands to the motor drive module to drive the vehicle to the final fire source location; Step 6: After the vehicle arrives, the main control module sends a fire extinguishing command to the fire extinguishing module, controlling the fire extinguishing module to perform the fire extinguishing operation; Step 7: After the fire is extinguished, the system confirms that the fire is eliminated, the trolley resumes its inspection state, and repeats the above steps to achieve all-weather automated fire prevention and control in unattended scenarios.

[0045] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An intelligent fire truck system based on multi-module collaboration, characterized in that, It includes a vehicle body, and a main control module, an image acquisition module, a fire extinguishing module, a fire source detection module, a motor drive module, and a drive motor installed on the vehicle body; The image acquisition module, fire extinguishing module, fire source detection module, and motor drive module are electrically connected to the main control module to feed back detection information and receive control commands. The motor drive module is electrically connected to the motor it drives, and is used to drive the drive motor to operate, so as to realize the forward, backward and steering actions of the vehicle body. The fire source detection module is used to collect characteristic light intensity signals of fire sources and transmit them to the main control module; the image acquisition module is used to collect real-time image information of the fire scene environment and the location of the fire source, and transmit the image data to the main control module; the main control module is used to receive and fuse the input signals of the fire source detection module and the image acquisition module to determine the location of the fire source and plan the travel path, and send control commands to the motor drive module to drive the vehicle to the location of the fire source, and send fire extinguishing commands to the fire extinguishing module to extinguish the fire source.

2. The intelligent fire truck system based on multi-module collaboration according to claim 1, characterized in that, The fire source detection module includes multiple flame sensors. The detection fields of the multiple flame sensors are stitched together to form a 360° full-area coverage without blind spots, which is used to realize all-round fire source detection of the vehicle's surrounding environment.

3. The intelligent fire truck system based on multi-module collaboration according to claim 2, characterized in that, Multiple flame sensors are arranged at the front, sides and rear of the vehicle to achieve 360° fire source detection without blind spots.

4. A multi-module collaborative intelligent fire truck system according to any one of claims 1-3, characterized in that, It also includes a power supply module, which is electrically connected to the main control module, image acquisition module, fire extinguishing module, fire source detection module, motor drive module, and drive motor to provide stable power supply for the entire system.

5. A fire extinguishing method, characterized in that, The system described in any one of claims 1-4 is used, and the method includes the following steps: The system automatically initializes after power-on, and the vehicle enters inspection mode; The fire source light intensity signal is collected by the fire source detection module, and the image information of the fire source location and fire scene environment is collected by the image acquisition module. The main control module receives the fire source light intensity signal, and determines whether there is a fire source in the corresponding detection direction based on the fire source light intensity signal and a preset fire source light intensity signal determination threshold. It also determines the initial azimuth angle range of the fire source and the distance range between the fire source and the vehicle body based on the signal strength. Once a fire source is determined to exist, the main control module obtains the final spatial location of the fire source by weighted fusion of the fire source azimuth estimate and distance estimate determined based on the initial azimuth range of the fire source and the distance range between the fire source and the vehicle body, as well as the fire source distance and fire source azimuth estimate obtained from the image information of the fire scene environment through image recognition and monocular vision ranging. The main control module plans the travel path based on the final fire source spatial location and sends control commands to the motor drive module to drive the vehicle to the final fire source spatial location. After the vehicle arrives, the main control module sends a fire extinguishing command to the fire extinguishing module, controlling the fire extinguishing module to perform the fire extinguishing operation; After the fire is extinguished, the system confirms that the fire is gone, the vehicle resumes its inspection state, and repeats the above steps in a loop to achieve all-weather automated fire prevention and control in unattended scenarios.

6. The method according to claim 4, characterized in that... The light intensity signal from the fire source decays exponentially with respect to the distance from the fire source. In the formula, I is the light intensity signal output by the infrared flame sensor in real time; I0 is the light intensity of a standard fire source at close range; k is the environmental attenuation coefficient; and d is the distance between the fire source and the vehicle body.

7. The method according to claim 4, characterized in that, The steps for determining the estimated azimuth and distance of the fire source based on the initial azimuth range of the fire source and the distance range between the fire source and the vehicle body include: Based on the light intensity signal intensity distribution data collected by multiple flame sensors in the fire source detection module, the centroid algorithm and triangulation principle are used to determine the estimated azimuth and distance of the fire source within the initial azimuth angle range and distance range.

8. The method according to claim 4, characterized in that, The method of obtaining the estimated distance and orientation of the fire source from image information of the fire scene environment through image recognition and monocular visual ranging includes: Using the initial azimuth range of the fire source and the distance range between the fire source and the vehicle body obtained from the fire source detection module as constraints, the initial azimuth range of the fire source is mapped to the column coordinates of the image and the distance range is mapped to the row coordinate range of the image based on the vehicle's heading angle and the camera's field of view, and the corresponding target detection area is delineated in the image. The fire source identification model identifies the image within the defined target detection area, extracts the feature information of the fire source, and uses a monocular vision ranging algorithm to convert the fire source pixel coordinates extracted from the image into spatial coordinates in the vehicle coordinate system, thereby obtaining an estimated fire source location based on the image information. The fire source pixel coordinates are then combined with camera intrinsic parameters, installation height, pitch angle, and field of view parameters to convert them into three-dimensional spatial coordinates in the vehicle coordinate system, and the azimuth angle and straight-line distance of the fire source are output simultaneously.

9. The method according to claim 8, characterized in that, The fire source identification model adopts an improved YOLO lightweight target detection network. The improved YOLO lightweight target detection network adds a flame infrared feature channel and a small target detection head on the basis of the original YOLO lightweight network, and optimizes the prior box size and non-maximum suppression strategy.

10. The method according to any one of claims 4-9, characterized in that, The fusion formula for the weighted fusion is: In the formula, P represents the final spatial location of the fire source. vis Based on the fire source location results obtained from the fire source detection module, P sen The fire source location result is obtained through image recognition and monocular visual ranging. w1 and w2 are weight coefficients that are adaptively adjusted according to the environment and distance.

Images