Fire extinguishing aiming method and system for unmanned aerial vehicle

Abstract

The invention discloses an unmanned aerial vehicle fire extinguishing aiming method and system. The method comprises the steps that an unmanned aerial vehicle carries a laser radar and an infrared thermal imager; performing dimension reduction on the laser point cloud to generate a two-dimensional depth map, calculating the discrete degree of each laser point, fitting the discrete trend direction, and moving the unmanned aerial vehicle along the discrete trend direction until the unmanned aerial vehicle reaches above the ignition point; the infrared thermal imager generates an infrared thermal image, calculates a heat ellipse, calculates a discrete ellipse according to the two-dimensional depth map, obtains a self-adaptive fire range according to the heat ellipse, the discrete ellipse and a range threshold value, and calculates a primary fire area, the unmanned aerial vehicle aims at the self-adaptive fire range, and extinguishes fire by adopting an anti-diffusion mechanism; and calculating a secondary fire area, updating a range threshold, and traversing the infrared thermal map. A plurality of sensors are adopted to realize fire extinguishing aiming, an anti-diffusion mechanism is adopted to extinguish fire after a self-adaptive fire range is obtained, fire diffusion can be effectively prevented, and high economic benefits and social benefits are achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicle fire extinguishing, and in particular to an unmanned aerial vehicle fire extinguishing aiming method and system. Background Art

[0002] With the continuous development of production and life in modern society, various types of fires have occurred frequently, and fires always bring huge economic losses to human society. In order to eliminate the challenges brought by fires in a timely manner, huge manpower and material resources need to be invested to reduce losses and save lives.

[0003] As a product of modern society, drones have played an important role in various human production activities. In order to reduce fire hazards and achieve better fire extinguishing aiming function, although the traditional drone fire aiming algorithm can aim at the fire location through the drone's sensor information, for outdoor scenes such as forests and grasslands, a single sensor is easily affected by environmental interference in outdoor environments, resulting in low accuracy of fire extinguishing aiming. In addition, due to the large area of ​​outdoor scenes, fires often spread rapidly, and a single sensor cannot accurately identify the fire extinguishing range, which can easily cause the fire to spread. There is a risk of re-ignition after a fire is extinguished once, and data support is needed for subsequent fire extinguishing. For a single sensor, once the sensor fails, it will seriously affect the accuracy of fire extinguishing aiming, and there is a lack of precise control over the fire extinguishing range. Therefore, it needs to be improved. Summary of the invention

[0004] The present invention provides a UAV fire extinguishing aiming method and system to solve the technical problems of the current fire extinguishing aiming sensor signal being single, the fire extinguishing range being inaccurate, and the risk of re-ignition.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0006] In one aspect, the present invention provides a method for aiming a fire extinguishing drone, the method comprising:

[0007] S1: The drone is equipped with a laser radar and an infrared thermal imager. The laser radar obtains the depth value of each laser point in the laser point cloud, and the infrared thermal imager obtains the temperature value of each pixel point.

[0008] S2: Dimensionality reduction of laser point cloud generates a two-dimensional depth map, calculates the discrete degree of each laser point, and fits the discrete trend direction;

[0009] S3: The UAV moves along the discrete trend direction until the variance of the discrete degree reaches the maximum and reaches above the fire point;

[0010] S4: Adjust the position of the drone, generate an infrared thermal image with the infrared thermal imager, expand the infrared thermal image by filtering the median method, calculate the thermal ellipse, calculate the discrete ellipse from the two-dimensional depth map, obtain the adaptive fire range from the thermal ellipse, the discrete ellipse and the range threshold, and calculate the fire area once;

[0011] S5: The drone targets the adaptive fire range and uses a non-spreading mechanism to extinguish the fire;

[0012] S6: Calculate the secondary fire area, update the range threshold, traverse the infrared heat map, calculate the number of pixels in the infrared heat map that exceed the fire threshold, and provide data for subsequent fire extinguishing.

[0013] Furthermore, the drone is equipped with a laser radar and an infrared thermal imager, wherein the laser radar obtains the depth value of each laser point in the laser point cloud, and the infrared thermal imager obtains the temperature value of each pixel point, including:

[0014] The drone is equipped with a laser radar and an infrared thermal imager, both of which are directly below the drone;

[0015] The laser radar is a 64-beam 3D laser radar with a frequency of 10HZ and an angular resolution of 0.2°. The laser radar obtains the depth value of each laser point in the laser point cloud. The specific representation is as follows:

[0016] D={d1,d2,...,d n}

[0017] Where D represents the set of laser point depths of a frame of laser point cloud, d1, d2, ..., d n represents the depth value of each laser point, and n represents the number of laser points;

[0018] The infrared thermal imager obtains the invisible infrared energy emitted by the object and uses different colors to represent the different temperature values ​​T of the measured object. The specific representation is as follows:

[0019] T={t1,t2,...,t m}

[0020] Where T represents the set of temperature values ​​of pixels collected by a frame of infrared thermal imager, t1, t2, ..., t m represents the temperature value of each pixel, and m represents the number of pixels;

[0021] It should be noted that the use of lidar and infrared thermal imager can avoid the overall failure of the method caused by the failure of a single sensor, thereby improving the robust performance of the method.

[0022] Furthermore, the laser point cloud is reduced in dimension to generate a two-dimensional depth map, the discrete degree of each laser point is calculated, and the discrete trend direction is fitted, including:

[0023] When the drone is launched to stable operation and heads to the fire point, the average depth d0 of each laser point in a fire-free environment can be obtained;

[0024] Dimensionality reduction processing of laser point cloud to generate two-dimensional depth map;

[0025] The discreteness of each laser point is calculated from the depth of each laser point in the two-dimensional depth map. The specific formula is as follows:

[0026] S={d1-d0,d2-d0,...,d n -d0}

[0027] Among them, S represents the set of discrete degrees of each laser point;

[0028] Find the coordinates of the first laser point (x1, y1) and the coordinates of the laser point with the largest degree of discreteness (x max ,y max );

[0029] Fitting discrete trend direction, the specific calculation formula is as follows:

[0030]

[0031] Among them, y represents the y-axis coordinate of the discrete trend direction, and x represents the x-axis coordinate of the discrete trend direction.

[0032] Further, the UAV moves along the discrete trend direction until the variance of the discrete degree reaches the maximum and reaches above the fire point, including:

[0033] After obtaining the discrete trend direction, it shows that due to the influence of smoke, the depth of the laser radar laser point is discrete. The greater the degree of dispersion of the laser point, the closer it is to the fire point. The UAV moves along the discrete trend direction and continuously searches for the fire point.

[0034] Calculate the mean of the discrete degree. The specific calculation formula is as follows:

[0035]

[0036] in, Represents the mean of the discrete degree, i represents an integer;

[0037] Calculate the variance of the discrete degree. The specific calculation formula is shown as follows:

[0038]

[0039] Among them, σ represents the variance of the degree of dispersion;

[0040] As the drone moves, when the variance of the discrete degree reaches the maximum, it indicates that the drone has reached the top of the fire point and found the fire extinguishing area;

[0041] It should be noted that the traditional method of finding the fire point is often based on the temperature of the pixel point, but there are many interferences in the actual field environment. Therefore, the efficiency and accuracy of the fire area search can be improved by calculating the discreteness.

[0042] Further, the position of the drone is adjusted, the infrared thermal imager generates an infrared thermal image, the infrared thermal image is expanded by filtering the median method, the thermal ellipse is calculated, the discrete ellipse is calculated from the two-dimensional depth map, the adaptive fire range is obtained from the thermal ellipse, the discrete ellipse and the range threshold, and the fire area is calculated once, including:

[0043] Adjust the position of the drone. To better extinguish the fire, lower the drone's altitude until it reaches the critical safety altitude h;

[0044] The infrared thermal imager generates an infrared heat map. Each pixel in the infrared heat map represents a temperature value. The infrared heat map is filtered using the expanded median method, and a 2*2 pixel neighborhood is selected as the filtering window. The specific calculation formula is shown in the following formula:

[0045] g(t)=10media{t1,t2,t3,t4}

[0046] Among them, g(t) represents the infrared thermal image after the expanded median filtering, t1, t2, t3, t4 represent the temperature values ​​of the pixels in the 2*2 pixel neighborhood;

[0047] In order to accurately extinguish the fire and aim at the fire location, the thermal ellipse is calculated from the infrared thermal image after the expanded median method filtering. First, the coordinates of the center point of the infrared thermal image are obtained. Where H represents the infrared heat map, and the coordinates of random points near the center of the infrared heat map are obtained.

[0048] Then calculate the geometric moment. The specific calculation formula is shown as follows:

[0049]

[0050] Among them, F i,j represents geometric moment;

[0051] Calculate the minor axis of the thermal ellipse from the geometric moment The specific calculation formula is shown as follows:

[0052]

[0053] Calculate the major axis of the thermal ellipse from the geometric moment The specific calculation formula is shown as follows:

[0054]

[0055] The discrete ellipse is calculated from the two-dimensional depth map, where the coordinates of the center point of the two-dimensional depth map are Discrete ellipse minor axis Discrete ellipse major axis Where L represents a two-dimensional depth map;

[0056] The heat ellipse, discrete ellipse and range threshold α are used to obtain the adaptive fire range.

[0057]

[0058] The coordinates of the center point of the adaptive fire range are as follows:

[0059]

[0060] Calculate the primary fire area C1, where the primary fire area represents the fire area before extinguishing;

[0061] It should be further explained that by calculating the center point of the adaptive fire range, the fusion of the heat ellipse and the discrete ellipse can be achieved, and when a large deviation occurs in one ellipse, the other ellipse can be corrected, providing the error correction capability of the method.

[0062] Furthermore, the drone targets the adaptive fire range and uses a non-diffusion mechanism to extinguish the fire, including:

[0063] The drone targets the adaptive fire range and starts extinguishing the fire;

[0064] In order to prevent the fire from spreading further, an anti-diffusion mechanism is used to extinguish the fire. Specifically, the adaptive fire range determines the length of the center point, the short axis and the long axis. First, aim at the long axis and extinguish the long axis path, then aim at the short axis and extinguish the short axis path, then aim at the center point and extinguish the center point and its vicinity, and finally extinguish the other fire locations within the adaptive fire range in a clockwise order.

[0065] Furthermore, the calculation of the secondary fire area, updating the range threshold, traversing the infrared heat map, and calculating the number of pixels in the infrared heat map that exceed the fire threshold to provide data for subsequent fire extinguishing include:

[0066] Calculate the secondary fire area C2, which refers to the fire area after one targeted fire extinguishing;

[0067] The range threshold is updated by the primary fire area and the secondary fire area. The specific calculation formula is as follows:

[0068]

[0069] Among them, α represents the range threshold updated by the first fire area and the second fire area;

[0070] The initial ignition threshold temperature is 55;

[0071] Traverse the infrared heat map. If the temperature of the infrared heat map pixels exceeds the ignition threshold and the number exceeds 10, secondary fire extinguishing is required. Otherwise, the fire extinguishing is successful.

[0072] On the other hand, the present invention also provides a UAV fire extinguishing aiming system, the UAV fire extinguishing aiming system comprising:

[0073] Signal module: The drone is equipped with a laser radar and an infrared thermal imager. The laser radar obtains the depth value of each laser point in the laser point cloud, and the infrared thermal imager obtains the temperature value of each pixel point.

[0074] Fire location search module: The laser point cloud is reduced in dimension to generate a two-dimensional depth map, the discrete degree of each laser point is calculated, and the discrete trend direction is fitted. The drone moves along the discrete trend direction until the variance of the discrete degree reaches the maximum and reaches above the fire point;

[0075] Fire extinguishing aiming module: adjust the position of the drone, generate an infrared thermal image with the infrared thermal imager, expand the infrared thermal image with the median method, calculate the thermal ellipse, calculate the discrete ellipse from the two-dimensional depth map, obtain the adaptive fire range from the thermal ellipse, the discrete ellipse and the range threshold, and calculate the fire area once. The drone aims at the adaptive fire range and uses the anti-diffusion mechanism to extinguish the fire.

[0076] Threshold update module: calculates the secondary fire area, updates the range threshold, traverses the infrared heat map, calculates the number of pixels in the infrared heat map that exceed the fire threshold, and provides data for subsequent fire extinguishing.

[0077] The beneficial effects brought about by the technical solution provided by the present invention include at least:

[0078] 1. The present invention uses multiple sensors to search for the location of the fire point, providing more accurate data support for fire extinguishing aiming. According to the characteristics of the fire, the laser radar is first used to find the central fire point position for the discrete degree of smoke to obtain a discrete ellipse, and then the infrared thermal imager is used to measure the temperature to obtain a thermal ellipse, and the adaptive fire range is jointly determined. Compared with the fire extinguishing aiming method of a single sensor, this solution avoids the risk of detection and aiming failure caused by a single sensor, and provides multi-dimensional data for fire extinguishing aiming. In addition, compared with the solution of using a camera to achieve fire extinguishing aiming, this solution can avoid the influence of light on aiming accuracy. Not only is the solution feasible during the day, but it can also play a role at night, thereby further increasing the use scenarios of the solution and improving the accuracy of fire extinguishing aiming;

[0079] 2. When aiming at firefighting, in order to eliminate fire hazards in a timely manner, it is necessary to find the best firefighting aiming position. This solution uses laser radar and infrared thermal imager to jointly determine the adaptive fire range, which can lock the area with the most serious fire. In order to avoid further expansion of the fire, the fire is extinguished through the anti-diffusion mechanism. Compared with the firefighting method that only aims at the center of the fire, this solution can effectively prevent the spread of fire, reduce the economic losses caused by the fire and improve the efficiency of firefighting, thereby further reducing the hardware cost of the firefighting drone, and has better economic and social benefits;

[0080] 3. This scheme sets up a fire extinguishing detection mechanism and a range threshold updating mechanism. Due to the influence of factors such as wind direction, wind speed and fire environment, there is a risk of fire rekindling after only one fire extinguishing. If left unattended, all previous efforts will be wasted. This scheme sets up a fire extinguishing detection mechanism. By measuring the primary fire area and the secondary fire area, it can be determined whether the current fire is spreading. To ensure safety, the range threshold is updated, the area of ​​the adaptive fire range is expanded, and secondary targeted fire extinguishing is carried out. Compared with the scheme of only conducting one targeted fire extinguishing, this scheme is safer and more scientific, and can eliminate fire hazards in a timely manner and reduce the potential risks caused by fire hazards. BRIEF DESCRIPTION OF THE DRAWINGS

[0081] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0082] Figure 1 A schematic diagram of the overall execution flow of a drone fire extinguishing aiming method provided by an embodiment of the present invention;

[0083] Figure 2 A schematic diagram of a UAV fire-fighting aiming system provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0084] In order to make the objectives, technical solutions and advantages of the present invention more clear, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings.

[0085] Example 1

[0086] This embodiment provides a method for aiming a drone to extinguish a fire. The method can be implemented by an electronic device as follows: Figure 1 Specifically, the method of this embodiment includes the following steps:

[0087] S1: The drone is equipped with a laser radar and an infrared thermal imager. The laser radar obtains the depth value of each laser point in the laser point cloud, and the infrared thermal imager obtains the temperature value of each pixel point.

[0088] Specifically, in this embodiment, the drone is equipped with a laser radar and an infrared thermal imager, wherein the laser radar obtains the depth value of each laser point in the laser point cloud, and the infrared thermal imager obtains the temperature value of each pixel point, including:

[0089] The drone is equipped with a laser radar and an infrared thermal imager, both of which are directly below the drone;

[0090] The laser radar is a 64-beam 3D laser radar with a frequency of 10HZ and an angular resolution of 0.2°. The laser radar obtains the depth value of each laser point in the laser point cloud. The specific representation is as follows:

[0091] D={d1,d2,...,d n}

[0092] Where D represents the set of laser point depths of a frame of laser point cloud, d1, d2, ..., d n represents the depth value of each laser point, and n represents the number of laser points;

[0093] The infrared thermal imager obtains the invisible infrared energy emitted by the object and uses different colors to represent the different temperature values ​​T of the measured object. The specific representation is as follows:

[0094] T={t1,t2,...,t m}

[0095] Where T represents the set of temperature values ​​of pixels collected by a frame of infrared thermal imager, t1, t2, ..., t m represents the temperature value of each pixel, and m represents the number of pixels;

[0096] It should be further explained that in this solution, in order to improve the accuracy of fire-fighting aiming, the drone uses an industrial drone and is equipped with a vehicle-mounted level laser radar.

[0097] S2: Dimensionality reduction of laser point cloud generates a two-dimensional depth map, calculates the discrete degree of each laser point, and fits the discrete trend direction;

[0098] Specifically, in this embodiment, the laser point cloud is reduced in dimension to generate a two-dimensional depth map, the discrete degree of each laser point is calculated, and the discrete trend direction is fitted, including:

[0099] When the drone is launched to stable operation and heads to the fire point, the average depth d0 of each laser point in a fire-free environment can be obtained;

[0100] Dimensionality reduction processing of laser point cloud to generate two-dimensional depth map;

[0101] The discreteness of each laser point is calculated from the depth of each laser point in the two-dimensional depth map. The specific formula is as follows:

[0102] S={d1-d0,d2-d0,...,d n -d0}

[0103] Among them, S represents the set of discrete degrees of each laser point;

[0104] Find the coordinates of the first laser point (x1, y1) and the coordinates of the laser point with the largest degree of discreteness (x max ,y max );

[0105] Fitting discrete trend direction, the specific calculation formula is as follows:

[0106]

[0107] Among them, y represents the y-axis coordinate of the discrete trend direction, and x represents the x-axis coordinate of the discrete trend direction.

[0108] S3: The UAV moves along the discrete trend direction until the variance of the discrete degree reaches the maximum and reaches above the fire point;

[0109] Specifically, in this embodiment, the UAV moves along the discrete trend direction until the variance of the discrete degree reaches the maximum and reaches above the fire point, including:

[0110] After obtaining the discrete trend direction, it shows that due to the influence of smoke, the depth of the laser radar laser point is discrete. The greater the degree of dispersion of the laser point, the closer it is to the fire point. The UAV moves along the discrete trend direction and continuously searches for the fire point.

[0111] Calculate the mean of the discrete degree. The specific calculation formula is as follows:

[0112]

[0113] in, Represents the mean of the discrete degree, i represents an integer;

[0114] Calculate the variance of the discrete degree. The specific calculation formula is shown as follows:

[0115]

[0116] Among them, σ represents the variance of the degree of dispersion;

[0117] As the drone moves, when the variance of the discrete degree reaches the maximum, it indicates that the drone has reached the top of the fire point and found the fire extinguishing area.

[0118] S4: Adjust the position of the drone, generate an infrared thermal image with the infrared thermal imager, expand the infrared thermal image by filtering the median method, calculate the thermal ellipse, calculate the discrete ellipse from the two-dimensional depth map, obtain the adaptive fire range from the thermal ellipse, the discrete ellipse and the range threshold, and calculate the fire area once;

[0119] Specifically, in this embodiment, the position of the drone is adjusted, the infrared thermal imager generates an infrared thermal image, the infrared thermal image is expanded by filtering the median method, the thermal ellipse is calculated, the discrete ellipse is calculated from the two-dimensional depth map, the adaptive fire range is obtained from the thermal ellipse, the discrete ellipse and the range threshold, and the fire area is calculated once, including:

[0120] Adjust the position of the drone. To better extinguish the fire, lower the drone's altitude until it reaches the critical safety altitude h;

[0121] The infrared thermal imager generates an infrared heat map. Each pixel in the infrared heat map represents a temperature value. The infrared heat map is filtered using the expanded median method, and a 2*2 pixel neighborhood is selected as the filtering window. The specific calculation formula is shown in the following formula:

[0122] g(t)=10media{t1,t2,t3,t4}

[0123] Among them, g(t) represents the infrared thermal image after the expanded median filtering, t1, t2, t3, t4 represent the temperature values ​​of the pixels in the 2*2 pixel neighborhood;

[0124] In order to accurately extinguish the fire and aim at the fire location, the thermal ellipse is calculated from the infrared thermal image after the expanded median method filtering. First, the coordinates of the center point of the infrared thermal image are obtained. Where H represents the infrared heat map, and the coordinates of random points near the center of the infrared heat map are obtained.

[0125] Then calculate the geometric moment. The specific calculation formula is shown as follows:

[0126]

[0127] Among them, F i,j represents geometric moment;

[0128] Calculate the minor axis of the thermal ellipse from the geometric moment The specific calculation formula is shown as follows:

[0129]

[0130] Calculate the major axis of the thermal ellipse from the geometric moment The specific calculation formula is shown as follows:

[0131]

[0132] The discrete ellipse is calculated from the two-dimensional depth map, where the coordinates of the center point of the two-dimensional depth map are Discrete ellipse minor axis Discrete ellipse major axis Where L represents the two-dimensional depth map, and the specific calculation formula is as follows:

[0133] First, get the coordinates of the center point of the two-dimensional depth map Get the coordinates of random points near the center of the 2D depth map

[0134] Then calculate the geometric moment. The specific calculation formula is shown as follows:

[0135]

[0136] Among them, F i,j represents geometric moment;

[0137] Calculate the minor axis of a discrete ellipse from geometric moments The specific calculation formula is shown as follows:

[0138]

[0139] Calculating the major axis of a discrete ellipse from geometric moments The specific calculation formula is shown as follows:

[0140]

[0141] The heat ellipse, discrete ellipse and range threshold α are used to obtain the adaptive fire range.

[0142]

[0143] The coordinates of the center point of the adaptive fire range are as follows:

[0144]

[0145] Calculate the primary fire area C1, where the primary fire area represents the fire area before extinguishing the fire.

[0146] S5: The drone targets the adaptive fire range and uses a non-spreading mechanism to extinguish the fire;

[0147] Specifically, in this embodiment, the drone targets the adaptive fire range and uses a non-diffusion mechanism to extinguish the fire, including:

[0148] The drone targets the adaptive fire range and starts extinguishing the fire;

[0149] In order to prevent the fire from spreading further, an anti-diffusion mechanism is used to extinguish the fire. Specifically, the adaptive fire range determines the length of the center point, the short axis and the long axis. First, aim at the long axis and extinguish the long axis path, then aim at the short axis and extinguish the short axis path, then aim at the center point and extinguish the center point and its vicinity, and finally extinguish the other fire locations within the adaptive fire range in a clockwise order.

[0150] S6: Calculate the secondary fire area, update the range threshold, traverse the infrared heat map, calculate the number of pixels in the infrared heat map that exceed the fire threshold, and provide data for subsequent fire extinguishing;

[0151] Specifically, in this embodiment, the secondary fire area is calculated, the range threshold is updated, and the infrared heat map is traversed. If the temperature of the infrared heat map pixel exceeds the fire threshold and the number exceeds 10, secondary fire extinguishing is required. Otherwise, the fire extinguishing is successful, including:

[0152] Calculate the secondary fire area C2, which refers to the fire area after one targeted fire extinguishing;

[0153] The range threshold is updated by the primary fire area and the secondary fire area. The specific calculation formula is as follows:

[0154]

[0155] Among them, α represents the range threshold updated by the first fire area and the second fire area;

[0156] The initial ignition threshold temperature is 55;

[0157] Traverse the infrared heat map. If the temperature of the infrared heat map pixels exceeds the ignition threshold and the number exceeds 10, secondary fire extinguishing is required. Otherwise, the fire extinguishing is successful.

[0158] In summary, the drone fire extinguishing aiming method of this embodiment has the following main features:

[0159] The present invention uses multiple sensors to search for the location of the fire point, providing more accurate data support for fire extinguishing aiming. The discrete ellipse obtained by the laser radar and the heat ellipse obtained by the infrared thermal imager jointly determine the adaptive fire range, and then obtain the precise range of fire extinguishing aiming, locking the area with the most serious fire; to avoid further expansion of the fire, the fire is extinguished through an anti-diffusion mechanism. Compared with the fire extinguishing method that only aims at the center of the fire, this scheme can effectively prevent the spread of fire, reduce the economic losses caused by the fire and improve the efficiency of fire extinguishing; a fire extinguishing detection mechanism and a range threshold update mechanism are set to eliminate fire hazards in a timely manner, reduce the potential risks caused by fire hazards, and better protect production and life safety.

[0160] Example 2

[0161] This embodiment provides a UAV fire extinguishing aiming system. Figure 2 As shown, it includes the following modules:

[0162] Signal module: The drone is equipped with a laser radar and an infrared thermal imager. The laser radar obtains the depth value of each laser point in the laser point cloud, and the infrared thermal imager obtains the temperature value of each pixel point.

[0163] Fire location search module: The laser point cloud is reduced in dimension to generate a two-dimensional depth map, the discrete degree of each laser point is calculated, and the discrete trend direction is fitted. The drone moves along the discrete trend direction until the variance of the discrete degree reaches the maximum and reaches above the fire point;

[0164] Fire extinguishing aiming module: adjust the position of the drone, generate an infrared thermal image with the infrared thermal imager, expand the infrared thermal image with the median method, calculate the thermal ellipse, calculate the discrete ellipse from the two-dimensional depth map, obtain the adaptive fire range from the thermal ellipse, the discrete ellipse and the range threshold, and calculate the fire area once. The drone aims at the adaptive fire range and uses the anti-diffusion mechanism to extinguish the fire.

[0165] Threshold update module: calculates the secondary fire area, updates the range threshold, traverses the infrared heat map, calculates the number of pixels in the infrared heat map that exceed the fire threshold, and provides data for subsequent fire extinguishing.

[0166] The unmanned aerial vehicle fire extinguishing aiming system of the present embodiment corresponds to the unmanned aerial vehicle fire extinguishing aiming method of the above-mentioned embodiment; the functions implemented by each functional module in the unmanned aerial vehicle fire extinguishing aiming system of the present embodiment correspond one-to-one to each process step in the unmanned aerial vehicle fire extinguishing aiming method of the above-mentioned embodiment; therefore, they will not be repeated here.

Claims

1. A method for aiming a drone to extinguish a fire, characterized in that: include: S1: The drone is equipped with a laser radar and an infrared thermal imager. The laser radar obtains the depth value of each laser point in the laser point cloud, and the infrared thermal imager obtains the temperature value of each pixel point. S2: Reduce the dimension of the laser point cloud to generate a two-dimensional depth map, calculate the discrete degree of each laser point, and fit the discrete trend direction; S3: The UAV moves along the discrete trend direction until the variance of the discrete degree reaches the maximum and reaches above the fire point; S4: Adjust the position of the drone, generate an infrared thermal image with the infrared thermal imager, expand the infrared thermal image by filtering the median method, calculate the thermal ellipse, calculate the discrete ellipse from the two-dimensional depth map, obtain the adaptive fire range from the thermal ellipse, the discrete ellipse and the range threshold, and calculate the fire area once; S5: The drone targets the adaptive fire range and uses a non-spreading mechanism to extinguish the fire; S6: Calculate the secondary fire area, update the range threshold, traverse the infrared heat map, calculate the number of pixels in the infrared heat map that exceed the fire threshold, and provide data for subsequent fire extinguishing.

2. The UAV fire extinguishing aiming method according to claim 1, characterized in that: The step S1 comprises: Both the lidar and thermal imaging cameras are directly below the drone; The laser radar is a 64-beam 3D laser radar with a frequency of 10HZ and an angular resolution of 0.2°. The laser radar obtains the depth value of each laser point in the laser point cloud, which is specifically expressed as follows: D={d1,d2,...,d n } Where D represents the set of laser point depths of a frame of laser point cloud, d1, d2, ..., d n represents the depth value of each laser point, and n represents the number of laser points; The infrared thermal imager collects invisible infrared light and uses different colors to represent different temperature values ​​T of the measured object. The specific representation method is as follows: T={t1,t2,...,t m } Where T represents the set of temperature values ​​of pixels collected by a frame of infrared thermal imager, t1, t2, ..., t m Represents the temperature value of each pixel, and m represents the number of pixels.

3. The UAV fire extinguishing aiming method according to claim 2, characterized in that: The step S2 comprises: The UAV is lifted off to stable operation, and when it goes to the fire point, the average depth d0 of each laser point in the fire-free environment is obtained; Reduce the dimension of the laser point cloud to generate a two-dimensional depth map; The discreteness of each laser point is calculated from the depth of each laser point in the two-dimensional depth map. The specific formula is as follows: S={d1-d0,d2-d0,...,d n -d0} Among them, S represents the set of discrete degrees of each laser point; Find the coordinates of the first laser point (x1, y1) and the coordinates of the laser point with the largest degree of discreteness (x max ,y max ); Fitting discrete trend direction, the specific calculation formula is as follows: Among them, y represents the y-axis coordinate of the discrete trend direction, and x represents the x-axis coordinate of the discrete trend direction.

4. The UAV fire extinguishing aiming method according to claim 3, characterized in that: The step S3 comprises: After obtaining the discrete trend direction, it shows that due to the influence of smoke, the depth of the laser radar laser point is discrete. The greater the degree of dispersion of the laser point, the closer it is to the fire point. The UAV moves along the discrete trend direction and continuously searches for the fire point. Calculate the mean of the discrete degree. The specific calculation formula is as follows: in, Represents the mean of the discrete degree, i represents a positive integer; Calculate the variance of the discrete degree. The specific calculation formula is shown as follows: Among them, σ represents the variance of the degree of dispersion; As the drone moves, when the variance of the discrete degree reaches the maximum, it indicates that the drone has reached the top of the fire point and found the fire extinguishing area.

5. The UAV fire extinguishing aiming method according to claim 4, characterized in that: Calculating the heat ellipse and the discrete ellipse in step S4 includes: The infrared thermal imager generates an infrared thermal image, and each pixel in the infrared thermal image represents a temperature value. The infrared thermal image is filtered using the expanded median method, and a 2*2 pixel neighborhood is selected as the filtering window. The specific calculation method is as follows: g(t)=10media{t1,t2,t3,t4} Among them, g(t) represents the infrared thermal image after the expanded median filtering, t1, t2, t3, t4 represent the temperature values ​​of the pixels in the 2*2 pixel neighborhood; In order to determine the precise fire area, the thermal ellipse is calculated from the infrared thermal image after the expanded median method filtering. First, the coordinates of the center point of the infrared thermal image are obtained. Where H represents the infrared heat map, and the coordinates of random points near the center of the infrared heat map are obtained. Calculate the geometric moment. The specific calculation method is as follows: Among them, F i,j represents geometric moment; Calculate the minor axis of the thermal ellipse from the geometric moment The specific calculation formula is shown as follows: Calculate the major axis of the thermal ellipse from the geometric moment The specific calculation formula is shown as follows: The discrete ellipse is calculated from the two-dimensional depth map, where the coordinates of the center point of the two-dimensional depth map are Discrete ellipse minor axis Discrete ellipse major axis Where L represents the two-dimensional depth map, and the specific calculation formula is as follows: First, get the coordinates of the center point of the two-dimensional depth map Get the coordinates of random points near the center of the 2D depth map Then calculate the geometric moment. The specific calculation formula is shown as follows: Among them, F i,j represents geometric moment; Calculate the minor axis of a discrete ellipse from geometric moments The specific calculation formula is shown as follows: Calculating the major axis of a discrete ellipse from geometric moments The specific calculation formula is shown as follows:

6. The UAV fire extinguishing aiming method according to claim 5, characterized in that: The calculation of a fire area includes: The adaptive fire range is calculated by the heat ellipse, the discrete ellipse and the range threshold α. The short axis of the adaptive fire range is The specific calculation method is as follows: The short axis of the adaptive fire range The specific calculation method is as follows: The coordinates of the center point of the adaptive fire range are as follows: Calculate the primary fire area C1, where the primary fire area represents the fire area before extinguishing the fire.

7. The UAV fire extinguishing aiming method according to claim 1, characterized in that: The drone targets the adaptive fire range and uses a non-spreading mechanism to extinguish the fire, including: The drone targets the adaptive fire range and starts extinguishing the fire; In order to prevent the fire from spreading further, an anti-diffusion mechanism is adopted to extinguish the fire. Specifically, the center point, the length of the short axis and the long axis are determined by the adaptive fire range. First, the long axis is aimed at to extinguish the fire on the long axis path, then the short axis is aimed at to extinguish the fire on the short axis path, then the center point is aimed at to extinguish the fire at the center point and its vicinity, and finally other fire locations within the adaptive fire range are extinguished in a clockwise order.

8. The UAV fire extinguishing aiming method according to claim 1, characterized in that: The calculation of the secondary fire area, updating the range threshold, traversing the infrared heat map, calculating the number of pixels in the infrared heat map that exceed the fire threshold, and providing data for subsequent fire extinguishing include: Calculate the secondary fire area C2, which refers to the fire area after one targeted fire extinguishing; The range threshold is updated by the primary fire area and the secondary fire area. The specific calculation formula is as follows: Among them, α' represents the range threshold updated by the first fire area and the second fire area.

9. The UAV fire extinguishing aiming method according to claim 8, characterized in that: The infrared heat map is traversed. If the temperature of the infrared heat map pixel exceeds the ignition threshold and the number exceeds 10, a secondary fire extinguishing is required. Otherwise, the fire extinguishing is successful, including: The initial ignition threshold temperature is 55; Traverse the infrared heat map. If the temperature of the infrared heat map pixels exceeds the ignition threshold and the number exceeds 10, secondary fire extinguishing is required. Otherwise, the fire extinguishing is successful.

10. A UAV fire extinguishing aiming system, characterized in that: include: Signal module: The drone is equipped with a laser radar and an infrared thermal imager. The laser radar obtains the depth value of each laser point in the laser point cloud, and the infrared thermal imager obtains the temperature value of each pixel point. Fire location search module: The laser point cloud is reduced in dimension to generate a two-dimensional depth map, the discrete degree of each laser point is calculated, and the discrete trend direction is fitted. The drone moves along the discrete trend direction until the variance of the discrete degree reaches the maximum and reaches above the fire point; Fire extinguishing aiming module: adjust the position of the drone, generate an infrared thermal image with the infrared thermal imager, expand the infrared thermal image with the median method, calculate the thermal ellipse, calculate the discrete ellipse from the two-dimensional depth map, obtain the adaptive fire range from the thermal ellipse, the discrete ellipse and the range threshold, and calculate the fire area once. The drone aims at the adaptive fire range and uses the anti-diffusion mechanism to extinguish the fire. Threshold update module: calculates the secondary fire area, updates the range threshold, traverses the infrared heat map, calculates the number of pixels in the infrared heat map that exceed the fire threshold, and provides data for subsequent fire extinguishing. Implement a drone fire extinguishing aiming method as described in any one of claims 1-9.

Images