Ignition point positioning and navigation method and device

Abstract

The invention is suitable for the technical field of fire rescue, and provides an ignition point positioning and navigation method and device. In the embodiment, for an area where a fire occurs, firstly, fire source image data of the area are obtained through a binocular camera and a sensor, and coordinates in a three-dimensional space of the area are determined through parameter information calibrated by the camera and image coordinates of all ignition points in the fire source image data; then determining the three-dimensional coordinates of the obstacles in the area through a laser radar; and finally, determining a target path from the inspection robot to the ignition point according to the current position of the inspection robot, the ignition point and the three-dimensional coordinates of the obstacle, and performing navigation. The problems that a traditional vision or single sensor technology is difficult to effectively apply and the ignition point positioning precision is low in a fire area under the conditions of insufficient smoke, insufficient light or complex scenes are solved.

Description

Technical Field

[0001] The present application relates to the field of fire rescue technology, and in particular to a fire point positioning and navigation method and device. Background Art

[0002] With the acceleration of urbanization and the increase in population density, fire safety issues have become increasingly serious. The dense buildings and concentrated population brought about by urbanization have increased the risk of fire and the difficulty of fire response.

[0003] Fire scenes are often characterized by complex environments such as thick smoke, high temperatures, flames, and collapsed buildings. These environmental factors not only directly threaten the safety of rescue workers but also severely interfere with traditional location methods. For example, smoke obscures vision, making it difficult for rescue workers to directly observe the fire's origin; high temperatures can render rescue equipment ineffective and even cause secondary disasters. The path of a fire is often influenced by numerous factors, including the distribution of combustible materials, building structure, ventilation conditions, and the characteristics of the fire source. These factors make the spread of a fire unpredictable and increase the difficulty of rescue efforts. Rapidly locating the fire's origin is crucial in fire rescue operations. Only by accurately identifying the fire's origin can targeted firefighting operations be implemented to control its spread. Furthermore, rapid fire location provides rescue workers with a safe evacuation route, minimizing casualties.

[0004] At present, fire detection and positioning technology mainly relies on thermal sensors and manual operation, which has the following shortcomings: low positioning accuracy. Thermal sensors are greatly affected by temperature distribution and cannot accurately obtain the three-dimensional spatial position of the fire point; they have poor environmental adaptability. Traditional vision or single sensor technology is difficult to effectively apply in smoke, insufficient light or complex scenes. Summary of the Invention

[0005] In view of this, the present application provides a fire point positioning and navigation method and device to solve the problems of low fire point positioning accuracy and poor environmental adaptability in existing fire detection and positioning technologies.

[0006] In a first aspect, the present application provides a fire point positioning and navigation method, which is applied to a fire treatment center. The fire treatment center includes a binocular camera, a sensor, a laser radar, and an inspection robot. The sensor includes a thermal imaging sensor and an environmental detection sensor. The method includes: Acquire fire source image data of a target area using a pre-calibrated binocular camera and the thermal imaging sensor, determine the image coordinates of the fire point in the fire source image data, and determine the three-dimensional coordinates of the fire point based on parameter information of the binocular camera and the image coordinates, wherein the three-dimensional coordinates of the fire point are the coordinates of the fire point in three-dimensional space, and the parameter information of the binocular camera includes a camera intrinsic parameter matrix and a camera extrinsic parameter matrix; Determine point cloud data of the fire point and obstacles in the target area using the laser radar to obtain point cloud coordinates of the fire point and the obstacles, and then determine the three-dimensional coordinates of the thermal imaging center point of the fire point using the thermal imaging sensor; Determine, by means of the environmental detection sensor, weight values ​​corresponding to the point cloud coordinates, the three-dimensional coordinates, and the three-dimensional coordinates of the thermal imaging center point of the fire point, and determine, based on the weight values, a fusion coordinate of the three-dimensional coordinates of the fire point, the point cloud coordinates, and the three-dimensional coordinates of the thermal imaging center point; A target path from the inspection robot to the fire point is determined according to the current position of the inspection robot, the fused coordinates of the fire point, and the point cloud coordinates of the obstacle, and the inspection robot is navigated according to the target path.

[0007] Optionally, before navigating the inspection robot according to the target path, the method further includes: Collect historical images and video data of fire scenes, and train the historical images and video data using a preset target detection algorithm and model to obtain a fire source recognition model; The fire intensity prediction result of the fire point spreading trend is predicted by the fire source recognition model and the fire source image data, and the target path is optimized by the fire intensity prediction result.

[0008] Optionally, determining a target path from the inspection robot to the fire point includes: By using a preset algorithm, the shortest path is determined according to the current position of the inspection robot, the three-dimensional coordinates of the fire point and the three-dimensional coordinates of the obstacle, and the shortest path is determined as the target path.

[0009] Optionally, before navigating the inspection robot according to the target path, the method further includes: The current position of the inspection robot, the three-dimensional coordinates of the fire point and the three-dimensional coordinates of the obstacle are updated in real time, the second path from the inspection robot to the fire point is recalculated using the updated data, and the second path is determined as the target path.

[0010] Optionally, the thermal imaging sensor is a FLIR thermal imager.

[0011] A second aspect of the present application provides a fire point positioning and navigation device, which is applied to a fire treatment center. The fire treatment center includes a binocular camera, a sensor, a laser radar, and an inspection robot. The sensor includes a thermal imaging sensor and an environmental detection sensor. The device includes: a fire point coordinate determination unit, configured to acquire fire source image data of a target area using a pre-calibrated binocular camera and the thermal imaging sensor, determine the image coordinates of the fire point in the fire source image data, and determine the three-dimensional coordinates of the fire point based on parameter information of the binocular camera and the image coordinates, wherein the three-dimensional coordinates of the fire point are the coordinates of the fire point in three-dimensional space, and the parameter information of the binocular camera includes a camera intrinsic parameter matrix and a camera extrinsic parameter matrix; an obstacle coordinate determination unit, configured to determine point cloud data of the fire point and obstacles in the target area using the laser radar to obtain point cloud coordinates of the fire point and the obstacles, and then determine the three-dimensional coordinates of the thermal imaging center point of the fire point using the thermal imaging sensor; a coordinate fusion unit, configured to determine weight values ​​corresponding to the point cloud coordinates, three-dimensional coordinates, and three-dimensional coordinates of the fire point and the thermal imaging center point through the environmental detection sensor, and determine fused coordinates of the three-dimensional coordinates of the fire point, the point cloud coordinates, and the three-dimensional coordinates of the thermal imaging center point according to the weight values; A path determination unit is used to determine a target path from the inspection robot to the fire point based on the current position of the inspection robot, the fused coordinates of the fire point, and the point cloud coordinates of the obstacle, and to navigate the inspection robot according to the target path.

[0012] Optionally, before navigating the inspection robot according to the target path, the device further includes: A path optimization unit is used to collect historical images and video data of fire scenes, and train the historical images and video data using a preset target detection algorithm and model to obtain a fire source recognition model; The fire intensity prediction result of the fire point spreading trend is predicted by the fire source recognition model and the fire source image data, and the target path is optimized by the fire intensity prediction result.

[0013] Optionally, determining the target path from the inspection robot to the fire point in the path determination unit includes: By using a preset algorithm, the shortest path is determined according to the current position of the inspection robot, the three-dimensional coordinates of the fire point and the three-dimensional coordinates of the obstacle, and the shortest path is determined as the target path.

[0014] Optionally, after the path determination unit determines the target path from the inspection robot to the fire point, the device further includes: A path updating unit is used to update the current position of the inspection robot, the three-dimensional coordinates of the fire point and the three-dimensional coordinates of the obstacle in real time, recalculate the second path from the inspection robot to the fire point through the updated data, and determine the second path as the target path.

[0015] Optionally, the thermal imaging sensor in the fire point coordinate determination unit is a FLIR thermal imager.

[0016] In the embodiments provided herein, for a fire area, binocular cameras and sensors are first used to acquire image data of the fire source in that area. The coordinates of each fire point in the fire source image data are then determined in three-dimensional space using the camera calibration parameters and the image coordinates of each fire point in the fire source image data. A lidar radar is then used to determine the three-dimensional coordinates of obstacles in the area. Finally, the inspection robot's target path to the fire point is determined and navigation is performed based on the inspection robot's current position, the fire point, and the three-dimensional coordinates of the obstacles. This solves the problem of the difficulty of effectively applying traditional vision or single-sensor technologies in fire areas in smoke, low light conditions, or complex scenes, as well as the low accuracy of fire point positioning. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 A flow chart of the method provided in the embodiment of the present application; Figure 2 A diagram of the device structure provided in an embodiment of the present application; Figure 3 A schematic diagram of the internal structure of a computer device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0018] Exemplary embodiments will be described in detail herein, with examples illustrated in the accompanying drawings. In the following description, when referring to the drawings, identical numerals in different figures represent identical or similar elements, unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all embodiments consistent with the present application. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present application, as detailed in the appended claims.

[0019] The terms used in this application are for the purpose of describing specific embodiments only and are not intended to limit this application. As used in this application and the appended claims, the singular forms "a," "an," "the," and "the" are intended to include the plural forms, unless the context clearly indicates otherwise. It should also be understood that the term "and / or" as used herein refers to and encompasses any and all possible combinations of one or more of the associated listed items.

[0020] It should be understood that although the terms first, second, third, etc. may be used in this application to describe various information, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from each other. For example, without departing from the scope of this application, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Depending on the context, the word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining".

[0021] The present application provides a fire point positioning and navigation method and device to solve the problems of low fire point positioning accuracy and poor environmental adaptability in existing fire detection and positioning technologies.

[0022] The following specific embodiments are used to describe the technical solution of the present application in detail. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments.

[0023] like Figure 1 The figure shows a flow chart of a fire point positioning and navigation method provided by the present application. The method is applied to a fire treatment center. The fire treatment center includes a binocular camera, a sensor, a lidar, and an inspection robot. The sensor includes a thermal imaging sensor and an environmental detection sensor. The process may include the following steps: Step S101, obtain fire source image data of the target area through a pre-calibrated binocular camera and the thermal imaging sensor, determine the image coordinates of the fire point in the fire source image data, and determine the three-dimensional coordinates of the fire point based on the parameter information of the binocular camera and the image coordinates.

[0024] In this embodiment, image data of the target area where the fire occurs can be obtained through a binocular camera, such as a high-resolution industrial binocular structured light camera, and the temperature distribution of the hot spot area of ​​the target area can be collected through a sensor. The image data and the temperature distribution of the hot spot area constitute the fire source image data of the target area.

[0025] The SIFT algorithm can be used to process the fire source image data. The specific processing method is as follows: 1. For any frame A of the fire source image data, extract the key points in the image. For each key point, generate a corresponding feature description using the SIFT algorithm to represent the local characteristics of the key point.

[0026] 2. Match each key point with the next frame image of image A in the fire source image data. The matching process can be completed by comparing the similarity between feature descriptions.

[0027] 3. After the match is successful, the coordinates of the matching key points in the two images can be determined. These coordinates can be used to identify the same features in the images and then determine the location of the fire point, that is, the image coordinates of the fire point.

[0028] After determining the image coordinates of the ignition point, the formula X world (x i ,y i ) =K -1 *(R*X image +T) calculate the three-dimensional coordinates of the fire point, which are the coordinates of the fire point in three-dimensional space. In the above formula, K is the intrinsic parameter matrix of the camera, R, T are the extrinsic parameter matrices of the camera, and can be determined by laboratory calibration of the binocular camera, such as the checkerboard calibration method. image (x i ,y i) 为着火点的图像坐标Xworld (x i ,y i ) is the coordinate of the ignition point in the world coordinate system, that is, the coordinate in three-dimensional space, (x i,yi ) is the coordinate of the i-th pixel point on the fire source image data.

[0029] In step S102, the laser radar is used to determine the point cloud data of the fire point and the obstacles in the target area to obtain the point cloud coordinates of the fire point and the obstacles, and then the thermal imaging sensor is used to determine the three-dimensional coordinates of the thermal imaging center point of the fire point.

[0030] In this embodiment, the LiDAR calculates the target's distance and position by emitting a laser beam and measuring the time difference between its reflection and return from the target object, generating a high-precision 3D point cloud dataset. Each point cloud in this dataset contains the 3D coordinates and laser reflection intensity. This 3D point cloud dataset is then grouped into multiple clusters using a pre-defined clustering algorithm, such as K-means or DBSCAN. Each cluster is then analyzed to identify the cluster representing the obstacle. The 3D coordinates of the fire point and the obstacle can then be determined using their center point or bounding box.

[0031] For the three-dimensional coordinates of the thermal imaging center point, the geometric center point of the main fire source area can be calculated based on the fire source image data collected by the thermal imaging sensor as the thermal imaging center point. The center point can be calculated using the following formula: thermal (x i ,y i ) = , where (x i ,y i) is the coordinate of the i-th pixel point on the fire source image data, and N is the total number of pixels in the fire source image data. Step S103, determine the weight values ​​corresponding to the point cloud coordinates, three-dimensional coordinates and three-dimensional coordinates of the fire point and the thermal imaging center point through the environmental detection sensor, and determine the fusion coordinates of the three-dimensional coordinates of the fire point, the point cloud coordinates and the three-dimensional coordinates of the thermal imaging center point according to the weight values.

[0032] In this embodiment, the environmental detection sensor may include a smoke sensor, a temperature sensor, a humidity sensor, and the like, so as to determine the current environment, and then set corresponding weight values ​​for each coordinate according to different environments. For example, when the smoke concentration is low, the reliability of the lidar data is high and the weight value is large. When the lighting conditions are good, the reliability of the binocular camera data is high and the weight value is large. In a high temperature environment, the reliability of the thermal imaging data is high and the weight value is large. Therefore, different intervals can be set for parameters such as temperature, smoke, and humidity, and corresponding weight values ​​can be set within each interval, so as to achieve the point cloud coordinates, three-dimensional coordinates of the fire point, and the weight values ​​corresponding to the three-dimensional coordinates of the thermal imaging center point determined by the environmental detection sensor.

[0033] By formula X fused (x i ,y i ) =X world (x i ,y i) * w1+Xthermal (x i ,y i ) * w2+X lidar* w3 can determine the fusion coordinate X fused (x i ,y i ) .

[0034] This embodiment introduces multi-sensor fusion technology, combining visual data, thermal imaging data and lidar data to improve the accuracy of fire source positioning, solving the problem that single sensor data is not sufficient for accurate positioning in complex fire scenes.

[0035] Step S104 , determining a target path from the inspection robot to the fire point according to the current position of the inspection robot, the fused coordinates of the fire point, and the point cloud coordinates of the obstacle, and navigating the inspection robot according to the target path.

[0036] In this embodiment, the inspection robot's current position is acquired in real time using its own IMU, visual odometry, and GNSS module. A path from the robot's current position to the fire source can be calculated using a pre-defined algorithm, such as the A* or RRT algorithm, and pre-defined path requirements, such as the highest number of fire points or the lowest temperature.

[0037] In another embodiment, a preset algorithm is used to determine the shortest path based on the inspection robot's current position, the three-dimensional coordinates of the fire point, and the three-dimensional coordinates of the obstacle, and this shortest path is determined as the target path. By determining the shortest path as the target path, this embodiment enables the inspection robot to more efficiently and reliably complete inspection tasks in a fire environment, thereby reducing fire losses.

[0038] In another embodiment, before navigating the inspection robot according to the target path, the method further includes: Collect historical images and video data of fire scenes, and train the historical images and video data using a preset target detection algorithm and model to obtain a fire source recognition model; The fire intensity prediction result of the fire point spreading trend is predicted by the fire source recognition model and the fire source image data, and the target path is optimized by the fire intensity prediction result.

[0039] In this embodiment, the collected image and video data of fire scenes can include different types of fire sources, such as open flames, smoldering fires, as well as smoke and high-temperature areas. The collected data is then annotated with information such as fire source location, type, and smoke density. A fire source identification model is then trained using this annotated dataset to optimize the model's detection and classification accuracy. Finally, the trained fire source identification model is deployed on patrol robots or computing equipment in fire response centers to process image data captured by binocular and infrared cameras in real time, identifying and classifying fire sources.

[0040] In another embodiment, after determining the target path of the inspection robot to the fire point, the method further includes: The current position of the inspection robot, the three-dimensional coordinates of the fire point and the three-dimensional coordinates of the obstacle are updated in real time, the second path from the inspection robot to the fire point is recalculated using the updated data, and the second path is determined as the target path.

[0041] In this embodiment, the method for updating the current position of the inspection robot is the same as the method for determining the current position of the inspection robot mentioned above; updating the three-dimensional coordinates of the fire point and the three-dimensional coordinates of the obstacle is the same as the method for determining the three-dimensional coordinates of the fire point and the three-dimensional coordinates of the obstacle mentioned above, and will not be repeated here.

[0042] This embodiment achieves the following effects by updating the robot position and the coordinates of the ignition point and obstacles in real time and recalculating the path: 1. Ensure that the path planning algorithm can adapt to the dynamic changes of the environment, thereby improving the adaptability and flexibility of path planning.

[0043] 2. Detect obstacles in real time and update the environment map. When dynamic obstacles are detected, use the preset algorithm to perform local path planning to effectively avoid obstacles and improve obstacle avoidance capabilities.

[0044] 3. By updating data and recalculating the path in real time, it can ensure that the path taken by the robot is the optimal solution under the current conditions, reduce time delays caused by environmental changes, and improve inspection efficiency.

[0045] So far, completed Figure 1 The process shown.

[0046] In an embodiment of the present application, for a fire area, a binocular camera and sensor are first used to obtain image data of the fire source in that area. The coordinates of each fire point in the fire source image data are then determined in three-dimensional space using the camera calibration parameters and the image coordinates of each fire point in the fire source image data. A lidar is then used to determine the three-dimensional coordinates of obstacles in the area. Finally, the inspection robot's target path to the fire point is determined and navigation is performed based on the inspection robot's current position, the fire point, and the three-dimensional coordinates of the obstacles. This solves the problem of the difficulty of effectively applying traditional vision or single-sensor technologies in fire areas in smoke, low light conditions, or complex scenes, as well as the low accuracy of fire point positioning.

[0047] The experimental results of this application are as follows: Tables 1 and 2 show the comparison results of the fire source error determined by a single sensor, such as a single lidar, a single binocular camera, and a single thermal imaging sensor, and the fire source error determined by this application. It can be seen that the error values ​​of this application for fire sources of different sizes and in different scenarios are significantly smaller than those of a single sensor.

[0048] Fire source type Single visual positioning error (m) Multi-sensor fusion positioning error (m) Small fire sources 0.25 0.05 Medium fire source 0.30 0.08 Large fire source 0.40 0.10 Table 1 method Single sensor (thermal imaging) Single sensor (lidar) The present invention (multi-sensor fusion) Average positioning error (m) 0.30 0.25 0.05 Error in extreme environment (m) >0.50 >0.40 0.10 Table 2 Table 3 shows the ignition point location test under special conditions. The performance comparison between a single sensor and the present invention is shown below. It can be seen that the accuracy of the present application in various characteristic environments is significantly higher than that of a single sensor.

[0049] environment Accuracy of a single sensor The accuracy of the present invention Low light 65% 92% Smoke interference 50% 88% High temperature reflection interference 45% 85% Table 3 like Figure 2 As shown, the present application also provides a fire point positioning and navigation device, which is applied to a fire treatment center. The fire treatment center includes a binocular camera, a sensor, a laser radar, and an inspection robot. The sensor includes a thermal imaging sensor and an environmental detection sensor. The device includes: a fire point coordinate determination unit 201, configured to acquire fire source image data of a target area using a pre-calibrated binocular camera and the thermal imaging sensor, determine the image coordinates of the fire point in the fire source image data, and determine the three-dimensional coordinates of the fire point based on the parameter information of the binocular camera and the image coordinates, wherein the three-dimensional coordinates of the fire point are the coordinates of the fire point in three-dimensional space, and the parameter information of the binocular camera includes a camera intrinsic parameter matrix and a camera extrinsic parameter matrix; an obstacle coordinate determination unit 202, configured to determine point cloud data of the fire point and obstacles in the target area using the laser radar to obtain point cloud coordinates of the fire point and the obstacles, and then determine the three-dimensional coordinates of the thermal imaging center point of the fire point using the thermal imaging sensor; A coordinate fusion unit 203 is configured to determine weight values ​​corresponding to the point cloud coordinates, three-dimensional coordinates, and three-dimensional coordinates of the fire point and the thermal imaging center point through the environmental detection sensor, and determine fused coordinates of the three-dimensional coordinates of the fire point, the point cloud coordinates, and the three-dimensional coordinates of the thermal imaging center point according to the weight values; The path determination unit 204 is used to determine the target path from the inspection robot to the fire point based on the current position of the inspection robot, the fused coordinates of the fire point and the point cloud coordinates of the obstacle, and navigate the inspection robot according to the target path.

[0050] In another embodiment, before navigating the inspection robot according to the target path, the device further includes: The path optimization unit 205 is used to collect historical images and video data of fire scenes, and train the historical images and video data using a preset target detection algorithm and model to obtain a fire source recognition model; The fire intensity prediction result of the fire point spreading trend is predicted by the fire source recognition model and the fire source image data, and the target path is optimized by the fire intensity prediction result.

[0051] In another embodiment, determining the target path from the inspection robot to the fire point in the path determination unit includes: By using a preset algorithm, the shortest path is determined according to the current position of the inspection robot, the three-dimensional coordinates of the fire point and the three-dimensional coordinates of the obstacle, and the shortest path is determined as the target path.

[0052] In another embodiment, after the path determination unit determines the target path from the inspection robot to the fire point, the device further includes: The path updating unit 206 is used to update the current position of the inspection robot, the three-dimensional coordinates of the fire point and the three-dimensional coordinates of the obstacle in real time, recalculate the second path from the inspection robot to the fire point through the updated data, and determine the second path as the target path.

[0053] In another embodiment, the thermal imaging sensor in the fire point coordinate determination unit is a FLIR thermal imager.

[0054] The above embodiment of the present invention provides a fire point positioning and navigation method, and based on this method provides a fire point positioning and navigation device. Through the above method and device, the problems of low fire point positioning accuracy and poor environmental adaptability in existing fire detection and positioning technologies are solved.

[0055] This embodiment also discloses a computer device, such as Figure 3 As shown, the computer device includes a processor and a memory, wherein the memory stores at least one instruction, and the at least one instruction is loaded and executed by the processor to implement any of the above-mentioned fire point positioning and navigation methods.

[0056] In addition, in the implementation of the fire point positioning and navigation device in the above example, the logical division of each program module is only an example. In actual application, the above functions can be assigned to different program modules as needed, for example, for the configuration requirements of the corresponding hardware or the convenience of software implementation. That is, the internal structure of the fire point positioning and navigation device is divided into different program modules to complete all or part of the functions described above.

[0057] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

Claims

1. A fire point positioning and navigation method, characterized in that: Applied to a fire handling center, the fire handling center includes a binocular camera, a sensor, a laser radar, and an inspection robot. The sensor includes a thermal imaging sensor and an environmental detection sensor. The method includes: Acquire fire source image data of a target area using a pre-calibrated binocular camera and the thermal imaging sensor, determine the image coordinates of the fire point in the fire source image data, and determine the three-dimensional coordinates of the fire point based on parameter information of the binocular camera and the image coordinates, wherein the three-dimensional coordinates of the fire point are the coordinates of the fire point in three-dimensional space, and the parameter information of the binocular camera includes a camera intrinsic parameter matrix and a camera extrinsic parameter matrix; Determine point cloud data of the fire point and obstacles in the target area using the laser radar to obtain point cloud coordinates of the fire point and the obstacles, and then determine the three-dimensional coordinates of the thermal imaging center point of the fire point using the thermal imaging sensor; Determine, by means of the environmental detection sensor, weight values ​​corresponding to the point cloud coordinates, the three-dimensional coordinates, and the three-dimensional coordinates of the thermal imaging center point of the fire point, and determine, based on the weight values, a fusion coordinate of the three-dimensional coordinates of the fire point, the point cloud coordinates, and the three-dimensional coordinates of the thermal imaging center point; A target path from the inspection robot to the fire point is determined according to the current position of the inspection robot, the fused coordinates of the fire point, and the point cloud coordinates of the obstacle, and the inspection robot is navigated according to the target path.

2. The method according to claim 1, characterized in that Before navigating the inspection robot according to the target path, the method further includes: Collect historical images and video data of fire scenes, and train the historical images and video data using a preset target detection algorithm and model to obtain a fire source recognition model; The fire intensity prediction result of the fire point spreading trend is predicted by the fire source recognition model and the fire source image data, and the target path is optimized by the fire intensity prediction result.

3. The method according to claim 1, characterized in that Determining the target path from the inspection robot to the fire point includes: By using a preset algorithm, the shortest path is determined according to the current position of the inspection robot, the three-dimensional coordinates of the fire point and the three-dimensional coordinates of the obstacle, and the shortest path is determined as the target path.

4. The method according to claim 1, wherein Before navigating the inspection robot according to the target path, the method further includes: The current position of the inspection robot, the three-dimensional coordinates of the fire point and the three-dimensional coordinates of the obstacle are updated in real time, the second path from the inspection robot to the fire point is recalculated using the updated data, and the second path is determined as the target path.

5. The method according to claim 1, wherein The thermal imaging sensor is a FLIR thermal imager.

6. A fire point positioning and navigation device, characterized in that: Applied to a fire treatment center, the fire treatment center includes a binocular camera, a sensor, a laser radar, and an inspection robot. The sensor includes a thermal imaging sensor and an environmental detection sensor. The device includes: a fire point coordinate determination unit, configured to acquire fire source image data of a target area using a pre-calibrated binocular camera and the thermal imaging sensor, determine the image coordinates of the fire point in the fire source image data, and determine the three-dimensional coordinates of the fire point based on parameter information of the binocular camera and the image coordinates, wherein the three-dimensional coordinates of the fire point are the coordinates of the fire point in three-dimensional space, and the parameter information of the binocular camera includes a camera intrinsic parameter matrix and a camera extrinsic parameter matrix; an obstacle coordinate determination unit, configured to determine point cloud data of the fire point and obstacles in the target area using the laser radar to obtain point cloud coordinates of the fire point and the obstacles, and then determine the three-dimensional coordinates of the thermal imaging center point of the fire point using the thermal imaging sensor; a coordinate fusion unit, configured to determine weight values ​​corresponding to the point cloud coordinates, three-dimensional coordinates, and three-dimensional coordinates of the fire point and the thermal imaging center point through the environmental detection sensor, and determine fused coordinates of the three-dimensional coordinates of the fire point, the point cloud coordinates, and the three-dimensional coordinates of the thermal imaging center point according to the weight values; A path determination unit is used to determine a target path from the inspection robot to the fire point based on the current position of the inspection robot, the fused coordinates of the fire point, and the point cloud coordinates of the obstacle, and to navigate the inspection robot according to the target path.

7. The device according to claim 6, characterized in that Before navigating the inspection robot according to the target path, the device further includes: A path optimization unit is used to collect historical images and video data of fire scenes, and train the historical images and video data using a preset target detection algorithm and model to obtain a fire source recognition model; The fire intensity prediction result of the fire point spreading trend is predicted by the fire source recognition model and the fire source image data, and the target path is optimized by the fire intensity prediction result.

8. The device according to claim 6, characterized in that Determining the target path from the inspection robot to the fire point in the path determination unit includes: By using a preset algorithm, the shortest path is determined according to the current position of the inspection robot, the three-dimensional coordinates of the fire point and the three-dimensional coordinates of the obstacle, and the shortest path is determined as the target path.

9. The device according to claim 6, characterized in that After the path determination unit determines the target path from the inspection robot to the fire point, the device further includes: A path updating unit is used to update the current position of the inspection robot, the three-dimensional coordinates of the fire point and the three-dimensional coordinates of the obstacle in real time, recalculate the second path from the inspection robot to the fire point through the updated data, and determine the second path as the target path.

10. The device according to claim 6, characterized in that The thermal imaging sensor in the fire point coordinate determination unit is a FLIR thermal imager.

Images