Detection method and device for rapidly constructing dike breach underwater topography based on unmanned aerial vehicle carrying

Abstract

The invention relates to the technical field of water conservancy surveying and mapping, in particular to a detection method and device for rapidly constructing an underwater terrain of a dike breach based on unmanned aerial vehicle carrying. Aiming at the defects of underwater terrain detection at the current dike breach, the method comprises the following steps of: acquiring water flow speeds, water flow depths, infrared spectrum images and visible light images of a plurality of point clouds by unmanned aerial vehicle carried equipment; processing the water flow depth according to a related algorithm, and processing the infrared spectrum image by using an image processing algorithm to obtain a breach obstacle map; and processing the visible light image by adopting an image processing algorithm to obtain the edge information of the breach, and finally generating a three-dimensional terrain model of the breach in the same coordinate system according to the edge information of the breach, the corrected water flow depth and the breach obstacle map. According to the method, the underwater terrain at the dike breach can be rapidly, accurately and safely detected, and the method is of great significance in improving the plugging engineering efficiency, optimizing the plugging scheme and ensuring the safety and effectiveness of plugging work.

Description

Technical Field

[0001] The present invention relates to the technical field of water conservancy surveying and mapping, and in particular to a method and device for detecting underwater terrain of a dike breach based on a rapid construction of an unmanned aerial vehicle. Background Art

[0002] On July 5, 2024, Tuanzhou Dike in Huarong County suffered a serious dike breach disaster. The width of the breach was about 220 meters, and more than 92% of the area in the dike was flooded, resulting in serious economic losses. After the dike breached, it is crucial to quickly and accurately assess the underwater topography of the breach area to formulate effective emergency response measures and subsequent repair plans. Currently, survey ships equipped with sonar are mostly used to measure the underwater topography. However, the water flow at the breach is extremely fast and carries a large amount of silt, floating objects, gravel and other sediments. When measuring using a survey ship equipped with sonar, and due to the presence of a large amount of sediment, gravel and floating objects, the survey ship will have a large floatation at the breach. Affected by the impact of sand and floating objects and the impact of water flow, the stability of the hull is not high, which in turn affects the measurement accuracy. The survey ship cannot carry out safe, stable, timely and effective detection.

[0003] Based on this, a method is needed that can overcome the defects of survey ships, is not affected by mud, floating objects, gravel and other sediments, and can conduct underwater topographic measurements quickly, smoothly and safely, so as to provide important data support for emergency response capabilities and disaster management efficiency in dealing with breaches. Summary of the invention

[0004] In view of the deficiencies in the prior art, the present invention provides a method and device for detecting underwater topography of a dike breach based on a drone, by which the underwater topography at the breach is measured by a drone, so that the drone is not affected by mud, floating objects, gravel and other sediments, and underwater topography can be measured quickly, smoothly and safely.

[0005] The present invention discloses a method for detecting underwater terrain of a dike breach based on a rapid construction of an unmanned aerial vehicle through an embodiment, comprising the following steps:

[0006] S1. The laser mapping radar carried by the surveying and mapping drone obtains the water flow velocity and water flow depth of multiple point clouds at the breach, the thermal infrared camera obtains the infrared spectrum image of the breach, and the visible light camera obtains the visible light image of the breach;

[0007] S2. Based on the water velocity and water depth of each point cloud at the breach, a water velocity correction algorithm and / or a wave and turbulence compensation algorithm are used to correct the water depth of each point cloud at the breach to obtain the corrected water depth at each point cloud; based on the infrared spectrum image, an image processing algorithm is used to identify obstacles at the breach to obtain a breach obstacle map; based on the visible light image, an image processing algorithm is used to obtain edge information at the breach;

[0008] S3. Generate a three-dimensional terrain model at the breach in the same coordinate system according to the breach edge information, the corrected water flow depth and the breach obstacle map.

[0009] Furthermore, when necessary, steps S1 to S3 may be repeated continuously to obtain a real-time three-dimensional terrain model at the breach.

[0010] Furthermore, it also includes: determining the flight altitude of the surveying and mapping UAV according to the width of the breach to ensure that the surveying radar, thermal infrared camera and visible light camera carried by the UAV can cover the entire breach area; adjusting the laser excitation intensity of the laser mapping radar to ensure that the laser penetration depth is greater than the water depth.

[0011] Furthermore, the laser mapping radar carried by the UAV obtains the water flow velocity and water flow depth of multiple point clouds at the breach, including: releasing tracer floating objects at a distance of at least 0.7 times the breach width from the breach, identifying the migration path and spatial position changes of the tracer floating objects, and obtaining the water level difference of each point cloud at the breach and the water flow velocity of each point cloud at the breach; the water flow depth is obtained according to the flight altitude of the mapping UAV from the highest water surface at the breach, the water level difference between each point cloud at the breach and the highest water surface, and the distance between the mapping UAV and the bottom of the water corresponding to each point cloud at the breach measured by the laser radar.

[0012] Furthermore, based on the water flow velocity and water flow depth of each point cloud at the breach, a water flow velocity correction algorithm is used to correct the water flow depth of each point cloud at the breach, and the corrected water flow depth at each point cloud includes:

[0013] The correction factor k is introduced to correct the water flow depth at each point cloud to compensate for the influence of water flow on the arrival time of the laser pulse. The calculation method is as follows:

[0014] d′=d·k

[0015]

[0016] Where d′ is the corrected water depth, d is the initially measured water depth, and t d -Delay caused by water velocity, v is the water velocity.

[0017] Furthermore, based on the water flow velocity and water flow depth of each point cloud at the breach, the wave and turbulence compensation algorithm is used to correct the water flow depth of each point cloud at the breach, and the corrected water flow depth at each point cloud includes:

[0018] The wave and turbulence intensity is estimated by the water velocity and water characteristics, and the compensation factor c is calculated according to the wave and turbulence intensity. The wave and turbulence compensation algorithm is used to filter out abnormal data caused by waves and turbulence. The calculation method is as follows:

[0019] d′=d±δ·k

[0020]

[0021] Among them, d′ is the corrected water depth, d is the initially measured water depth, δ is the displacement caused by waves and turbulence, and h is the average height of waves and turbulence.

[0022] Furthermore, if the water flow depth of each point cloud at the breach is corrected based on the water flow velocity and water flow depth of each point cloud at the breach, and the water flow velocity correction algorithm and the wave and turbulence compensation algorithm are used at the same time, the corrected water flow depth at each point cloud is obtained by the following method:

[0023] First, use the water flow velocity correction algorithm to correct the water flow depth at each point cloud to obtain the corrected water flow depth at each point cloud, which can be expressed as:

[0024] The correction factor k is introduced to correct the water flow depth at each point cloud to compensate for the influence of water flow on the arrival time of the laser pulse. The calculation method is as follows:

[0025] d′=d·k

[0026]

[0027] Where d′ is the corrected water depth, d is the initially measured water depth, and t d -Delay caused by water velocity, v is the water velocity;

[0028] Then the wave and turbulence compensation algorithm is used to compensate the corrected water flow depth to obtain the corrected water flow depth at each point cloud. Specifically, it can be expressed as:

[0029] The wave and turbulence intensity is estimated by the water velocity and water characteristics, and the compensation factor c is calculated according to the wave and turbulence intensity. The wave and turbulence compensation algorithm is used to filter out abnormal data caused by waves and turbulence. The calculation method is as follows:

[0030] d″=d′±δ·k

[0031]

[0032] Where d″ is the corrected water depth, d′ is the corrected water depth, δ is the displacement caused by waves and turbulence, and h is the average height of waves and turbulence.

[0033] Furthermore, based on the infrared spectrum image, the image processing algorithm is used to identify the obstacles at the breach, and the breach obstacle map obtained includes:

[0034] Based on multispectral image analysis, the spectral features of obstacles are extracted from infrared spectral images, and then the obstacles are detected using image processing algorithms. Then, the obstacles are classified using machine learning algorithms to obtain a breach obstacle map.

[0035] Among them, the method of extracting the spectral characteristics of obstacles from infrared spectrum images based on multispectral image analysis is as follows:

[0036] F(i,j)=[I1(i,j),I2(i,j),I3(i,j)...I4(i,j)]

[0037] The method of detecting obstacles using image processing algorithms is as follows:

[0038]

[0039] Where n is the number of bands of the infrared spectral image, and the pixel value of each band is I ki (i, j) (ki=1,2,…,n), the spectral feature vector F(i,j) of each pixel, B is the binary image of the obstacle; T B is the threshold value.

[0040] Furthermore, generating a three-dimensional terrain model at the breach in the same coordinate system according to the breach edge information, the corrected water flow depth and the breach obstacle map includes:

[0041] Eliminate outliers in the corrected water flow depth at each point cloud, and then use the least squares method or iterative closest point algorithm to optimize the positions of all point clouds based on the precise coordinates provided by the Beidou base station, and use linear interpolation to convert the positions of each point cloud into a regular grid in the same coordinate system;

[0042] The corrected water flow depth of each point cloud excluding outliers is converted into the bottom point cloud height at the breach, the bottom point cloud height is filled into a regular grid, and the height of all bottom positions is determined according to the nearest neighbor method to obtain a three-dimensional height map;

[0043] The breach edge information and the breach obstacle map are added to the three-dimensional height map to obtain a three-dimensional terrain model at the breach.

[0044] Furthermore, the calculation method used to remove the outliers in the corrected water flow depth at each point cloud is as follows:

[0045]

[0046] Among them, d i To correct for water flow depth.

[0047] In order to achieve the above-mentioned purpose, the present invention also includes a detection device for rapidly constructing underwater terrain of dike breach based on a drone, which includes: a surveying drone, a tracer floating object, a first correction module, a second image processing module and a three-dimensional terrain generation module;

[0048] The surveying drone is used to carry surveying laser radar, thermal infrared camera and visible light camera, and is used to obtain the water flow velocity and water flow depth of multiple point clouds at the breach, infrared spectrum image and visible light image of the breach;

[0049] Tracer floaters are used to display their migration paths and spatial position changes, and to obtain the water level difference and water flow velocity of each point cloud at the breach.

[0050] The first correction module is used to correct the water flow depth of each point cloud at the breach based on the water flow velocity and water flow depth of each point cloud at the breach, using a water flow velocity correction algorithm and / or a wave and turbulence compensation algorithm to obtain a corrected water flow depth at each point cloud;

[0051] The second image processing module is used to identify obstacles at the breach by using an image processing algorithm based on the infrared spectrum image to obtain a breach obstacle map; it is also used to obtain edge information at the breach by using an image processing algorithm based on the visible light image;

[0052] The three-dimensional terrain generation module is used to generate a three-dimensional terrain model at the breach in the same coordinate system based on the breach edge information, the corrected water flow depth and the breach obstacle map.

[0053] The technical principle of the present invention is: first, the water flow velocity and water flow depth of multiple point clouds at the breach are obtained by a surveying and mapping radar, the infrared spectrum image of the breach is obtained by a thermal infrared camera, and the visible light image of the breach is obtained by a visible light camera. Then, the water flow depth is processed according to a relevant algorithm, and based on the infrared spectrum image, an image processing algorithm is used to identify obstacles at the breach to obtain a breach obstacle map; based on the visible light image, an image processing algorithm is used to obtain edge information at the breach, and finally, a three-dimensional terrain model of the breach is generated in the same coordinate system according to the breach edge information, the corrected water flow depth and the breach obstacle map. If necessary, the above method can be repeated continuously to obtain a real-time three-dimensional terrain model of the breach.

[0054] Compared with the prior art, the present invention has the following beneficial effects: the water flow at the dike breach has the characteristics of fast flow velocity, high sand content, and complex distribution patterns of water level field and flow velocity field. In view of the shortcomings of the current underwater terrain detection at the dike breach, the present invention proposes a detection device and a detection method for quickly constructing the underwater terrain of the dike breach based on an unmanned aerial vehicle. This method can realize the rapid, accurate and safe detection of the underwater terrain at the dike breach, which is of great significance to improving the efficiency of the plugging project, optimizing the plugging plan, and ensuring the safety and effectiveness of the plugging work. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] In order to more clearly illustrate the technical solution of the present invention and facilitate further understanding of the technical effects, technical features and purposes of the present invention, the present invention is described in detail below in conjunction with the accompanying drawings. The accompanying drawings constitute an essential part of the specification and are used together with the embodiments of the present invention to illustrate the technical solution of the present invention, but do not constitute a limitation to the present invention.

[0056] Figure 1 A flow chart of the method described in the present invention;

[0057] Figure 2 It is a detection flow chart of the present invention. DETAILED DESCRIPTION

[0058] The present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. Of course, the specific embodiments described below are only intended to explain the technical solutions of the present invention, rather than to limit the present invention. In addition, the parts described in the embodiments or the accompanying drawings are only examples of the relevant parts of the present invention, rather than the entirety of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art should naturally belong to the protection scope of the present invention.

[0059] like Figure 1 , 2 As shown, the present invention proposes a method for detecting underwater terrain of a dike breach based on a rapid construction of an unmanned aerial vehicle through the following embodiments, comprising the following steps:

[0060] S1. The laser mapping radar carried by the surveying and mapping drone obtains the water flow velocity and water flow depth of multiple point clouds at the breach, the thermal infrared camera obtains the infrared spectrum image of the breach, and the visible light camera obtains the visible light image of the breach;

[0061] S2. Based on the water velocity and water depth of each point cloud at the breach, a water velocity correction algorithm and / or a wave and turbulence compensation algorithm are used to correct the water depth of each point cloud at the breach to obtain the corrected water depth at each point cloud; based on the infrared spectrum image, an image processing algorithm is used to identify obstacles at the breach to obtain a breach obstacle map; based on the visible light image, an image processing algorithm is used to obtain edge information at the breach;

[0062] S3. Generate a three-dimensional terrain model at the breach in the same coordinate system according to the breach edge information, the corrected water flow depth and the breach obstacle map.

[0063] In this embodiment, steps S1 to S3 may be repeated continuously to obtain a real-time three-dimensional terrain model of the breach.

[0064] In this embodiment, it also includes: determining the flight altitude of the surveying and mapping UAV according to the width of the breach to ensure that the surveying and mapping radar, thermal infrared camera and visible light camera carried by the UAV can cover the entire breach area; adjusting the laser excitation intensity of the laser surveying and mapping radar to ensure that the laser penetration depth D is greater than the water depth H. Wherein, it can be expressed as:

[0065]

[0066] Where W is the scanning range of the mapping radar, W j is the breach width, θ is the maximum scanning angle of the instrument; α is the absorption coefficient, which reflects the degree of laser absorption by water; I is the intensity of the laser at a certain point in the water; I o The initial intensity of the laser when it is incident.

[0067] In this embodiment, the laser mapping radar carried by the drone obtains the water flow velocity and water flow depth of multiple point clouds at the breach, including: releasing a tracer float at a distance of at least 0.7 times the width of the breach from the breach, identifying the migration path and spatial position change of the tracer float, and obtaining the water level difference of each point cloud at the breach and the water flow velocity of each point cloud at the breach; the water flow depth is obtained according to the flight altitude of the mapping drone from the highest water surface at the breach, the water level difference between each point cloud at the breach and the highest water surface, and the distance between the mapping drone and the bottom of the water corresponding to each point cloud at the breach measured by the laser radar.

[0068] In this embodiment, based on the water flow velocity and water flow depth of each point cloud at the breach, a water flow velocity correction algorithm is used to correct the water flow depth of each point cloud at the breach, and the corrected water flow depth at each point cloud includes:

[0069] The correction factor k is introduced to correct the water flow depth at each point cloud to compensate for the influence of water flow on the arrival time of the laser pulse. The calculation method is as follows:

[0070] d′=d·k

[0071]

[0072] Where d′ is the corrected water depth, d is the initially measured water depth, and t d -Delay caused by water velocity, v is the water velocity.

[0073] If the floating object is at position (x1, y1) at time t1 and at position (x1, y1) at time t2, the water velocity v of a point cloud is calculated as follows:

[0074]

[0075] If the highest position of the floating object on the water surface is Z1, and the position of a certain point is Z2, then the water level difference of a certain point cloud is: ΔZ = |Z1-Z2|.

[0076] In this embodiment, based on the water flow velocity and water flow depth of each point cloud at the breach, a wave and turbulence compensation algorithm is used to correct the water flow depth of each point cloud at the breach, and the corrected water flow depth at each point cloud includes:

[0077] The wave and turbulence intensity is estimated by the water velocity and water characteristics, and the compensation factor c is calculated according to the wave and turbulence intensity. The wave and turbulence compensation algorithm is used to filter out abnormal data caused by waves and turbulence. The calculation method is as follows:

[0078] d′=d±δ·k

[0079]

[0080] Among them, d′ is the corrected water depth, d is the initially measured water depth, δ is the displacement caused by waves and turbulence, and h is the average height of waves and turbulence.

[0081] In this embodiment, if the water flow depth of each point cloud at the breach is corrected based on the water flow velocity and water flow depth of each point cloud at the breach, and the water flow velocity correction algorithm and the wave and turbulence compensation algorithm are used at the same time, the corrected water flow depth at each point cloud is obtained by the following method:

[0082] First, use the water flow velocity correction algorithm to correct the water flow depth at each point cloud to obtain the corrected water flow depth at each point cloud, which can be expressed as:

[0083] The correction factor k is introduced to correct the water flow depth at each point cloud to compensate for the influence of water flow on the arrival time of the laser pulse. The calculation method is as follows:

[0084] d′=d·k

[0085]

[0086] Where d′ is the corrected water depth, d is the initially measured water depth, and t d -Delay caused by water velocity, v is the water velocity;

[0087] Then the wave and turbulence compensation algorithm is used to compensate the corrected water flow depth to obtain the corrected water flow depth at each point cloud. Specifically, it can be expressed as:

[0088] The wave and turbulence intensity is estimated by the water velocity and water characteristics, and the compensation factor c is calculated according to the wave and turbulence intensity. The wave and turbulence compensation algorithm is used to filter out abnormal data caused by waves and turbulence. The calculation method is as follows:

[0089] d″=d′±δ·k

[0090]

[0091] Where d″ is the corrected water depth, d′ is the corrected water depth, δ is the displacement caused by waves and turbulence, and h is the average height of waves and turbulence.

[0092] In this embodiment, based on the infrared spectrum image, an image processing algorithm is used to identify obstacles at the breach, and the breach obstacle map obtained includes:

[0093] Based on multispectral image analysis, the spectral features of obstacles are extracted from infrared spectral images, and then the obstacles are detected using image processing algorithms. Then, the obstacles are classified using machine learning algorithms to obtain a breach obstacle map.

[0094] Among them, the method of extracting the spectral characteristics of obstacles from infrared spectrum images based on multispectral image analysis is as follows:

[0095] F(i,j)=[I1(i,j),I2(i,j),I3(i,j)...I4(i,j)]

[0096] The method of detecting obstacles using image processing algorithms is as follows:

[0097]

[0098] Where n is the number of bands of the infrared spectral image, and the pixel value of each band is I ki (i, j) (ki=1,2,…,n), the spectral feature vector F(i,j) of each pixel, B is the binary image of the obstacle; T B is the threshold value.

[0099] In this embodiment, generating a three-dimensional terrain model at the breach in the same coordinate system according to the breach edge information, the corrected water flow depth and the breach obstacle map includes:

[0100] Eliminate outliers in the corrected water flow depth at each point cloud, and then use the least squares method or iterative closest point algorithm to optimize the positions of all point clouds based on the precise coordinates provided by the Beidou base station, and use linear interpolation to convert the positions of each point cloud into a regular grid in the same coordinate system;

[0101] The corrected water flow depth of each point cloud excluding outliers is converted into the bottom point cloud height at the breach, the bottom point cloud height is filled into a regular grid, and the height of all bottom positions is determined according to the nearest neighbor method to obtain a three-dimensional height map;

[0102] The breach edge information and the breach obstacle map are added to the three-dimensional height map to obtain a three-dimensional terrain model at the breach.

[0103] Among them, the calculation method used to remove the outliers in the corrected water flow depth at each point cloud is as follows:

[0104]

[0105] Among them, d i To correct for water flow depth.

[0106] In order to achieve the above-mentioned purpose, the present invention also includes a detection device for rapidly constructing underwater terrain of dike breach based on a drone, which includes: a surveying drone, a tracer floating object, a first correction module, a second image processing module and a three-dimensional terrain generation module;

[0107] The surveying drone is used to carry surveying laser radar, thermal infrared camera and visible light camera, and is used to obtain the water flow velocity and water flow depth of multiple point clouds at the breach, infrared spectrum image and visible light image of the breach;

[0108] Tracer floaters are used to display their migration paths and spatial position changes, and to obtain the water level difference and water flow velocity of each point cloud at the breach.

[0109] The first correction module is used to correct the water flow depth of each point cloud at the breach based on the water flow velocity and water flow depth of each point cloud at the breach, using a water flow velocity correction algorithm and / or a wave and turbulence compensation algorithm to obtain a corrected water flow depth at each point cloud;

[0110] The second image processing module is used to identify obstacles at the breach by using an image processing algorithm based on the infrared spectrum image to obtain a breach obstacle map; it is also used to obtain edge information at the breach by using an image processing algorithm based on the visible light image;

[0111] The three-dimensional terrain generation module is used to generate a three-dimensional terrain model at the breach in the same coordinate system according to the breach edge information, the corrected water flow depth and the breach obstacle map.

[0112] It should be noted that the above embodiments are only for more clearly illustrating the technical solutions of the present invention. Those skilled in the art will understand that the implementation methods of the present invention are not limited to the above contents, and obvious changes, replacements or substitutions based on the above contents do not exceed the scope of the technical solutions of the present invention; other implementation methods will naturally fall within the scope of the present invention without departing from the concept of the present invention.

Claims

1. A method for detecting underwater terrain of dike breaches based on rapid construction of drones, characterized in that: The steps include: S1. The laser mapping radar carried by the surveying and mapping drone obtains the water flow velocity and water flow depth of multiple point clouds at the breach, the thermal infrared camera obtains the infrared spectrum image of the breach, and the visible light camera obtains the visible light image of the breach; S2. Based on the water velocity and water depth of each point cloud at the breach, a water velocity correction algorithm and / or a wave and turbulence compensation algorithm are used to correct the water depth of each point cloud at the breach to obtain the corrected water depth at each point cloud; based on the infrared spectrum image, an image processing algorithm is used to identify obstacles at the breach to obtain a breach obstacle map; based on the visible light image, an image processing algorithm is used to obtain edge information at the breach; S3. Generate a three-dimensional terrain model at the breach in the same coordinate system according to the breach edge information, the corrected water flow depth and the breach obstacle map.

2. The method for detecting underwater terrain of a dike breach based on a drone as claimed in claim 1, characterized in that: Also includes: Repeat steps S1-S3 continuously to obtain a real-time three-dimensional terrain model at the breach.

3. The method for detecting underwater topography of a dike breach based on a drone as claimed in claim 1 or 2, characterized in that: Also includes: Determine the flight altitude of the surveying drone based on the width of the breach to ensure that the surveying radar, thermal infrared camera and visible light camera carried by the drone can cover the entire breach area; Adjust the laser excitation intensity of the lidar to ensure that the laser penetration depth is greater than the water depth.

4. The method for detecting underwater terrain of a dike breach based on a drone as claimed in claim 3 is characterized in that: The laser mapping radar carried by the drone obtains the water flow velocity and water flow depth of multiple point clouds at the breach, including: Release the tracer floating objects at a distance of at least 0.7 times the width of the breach, identify the migration path and spatial position change of the tracer floating objects, and obtain the water level difference and water flow velocity of each point cloud at the breach; The water flow depth is obtained based on the flight altitude of the surveying UAV from the highest water surface at the breach, the water level difference between each point cloud at the breach and the highest water surface, and the distance between the surveying UAV and the water bottom corresponding to each point cloud at the breach measured by the lidar.

5. The method for detecting underwater terrain of a dike breach based on a drone as claimed in claim 1, characterized in that: Based on the water flow velocity and water flow depth of each point cloud at the breach, a water flow velocity correction algorithm is used to correct the water flow depth of each point cloud at the breach, and the corrected water flow depth at each point cloud is obtained, which includes: The correction factor k is introduced to correct the water flow depth at each point cloud to compensate for the influence of water flow on the arrival time of the laser pulse. The calculation method is as follows: d′=d·k Where d′ is the corrected water depth, d is the initially measured water depth, and t d -Delay caused by water velocity, v is the water velocity.

6. The method for detecting underwater terrain of a dike breach based on a drone as claimed in claim 1, characterized in that: Based on the water flow velocity and water flow depth of each point cloud at the breach, the wave and turbulence compensation algorithm is used to correct the water flow depth of each point cloud at the breach, and the corrected water flow depth at each point cloud is obtained, which includes: The wave and turbulence intensity is estimated by the water velocity and water characteristics, and the compensation factor c is calculated according to the wave and turbulence intensity. The wave and turbulence compensation algorithm is used to filter out abnormal data caused by waves and turbulence. The calculation method is as follows: d′=d±δ·k Among them, d′ is the corrected water depth, d is the initially measured water depth, δ is the displacement caused by waves and turbulence, and h is the average height of waves and turbulence.

7. The method for detecting underwater terrain of a dike breach based on a drone as claimed in claim 1, characterized in that: Based on the water flow velocity and water flow depth of each point cloud at the breach, a water flow velocity correction algorithm and / or a wave and turbulence compensation algorithm are used to correct the water flow depth of each point cloud at the breach, and the corrected water flow depth at each point cloud is obtained, which includes: First, use the water flow velocity correction algorithm to correct the water flow depth at each point cloud to obtain the corrected water flow depth at each point cloud, which can be expressed as: The correction factor k is introduced to correct the water flow depth at each point cloud to compensate for the influence of water flow on the arrival time of the laser pulse. The calculation method is as follows: d′=d·k Where d′ is the corrected water depth, d is the initially measured water depth, and t d -Delay caused by water velocity, v is the water velocity; Then the wave and turbulence compensation algorithm is used to compensate the corrected water flow depth to obtain the corrected water flow depth at each point cloud. Specifically, it can be expressed as: The wave and turbulence intensity is estimated by the water velocity and water characteristics, and the compensation factor c is calculated according to the wave and turbulence intensity. The wave and turbulence compensation algorithm is used to filter out abnormal data caused by waves and turbulence. The calculation method is as follows: d″=d′±δ·k Where d″ is the corrected water depth, d′ is the corrected water depth, δ is the displacement caused by waves and turbulence, and h is the average height of waves and turbulence.

8. The method for detecting underwater terrain of a dike breach based on a drone as claimed in claim 1, characterized in that: The method of identifying obstacles at the breach by using an image processing algorithm based on infrared spectrum images to obtain a breach obstacle map includes: Based on multispectral image analysis, the spectral features of obstacles are extracted from infrared spectral images, and then the obstacles are detected using image processing algorithms. Then, the obstacles are classified using machine learning algorithms to obtain a breach obstacle map. Among them, the method of extracting the spectral characteristics of obstacles from infrared spectrum images based on multispectral image analysis is as follows: F(i,j)=[I1(i,j),I2(i,j),I3(i,j)...I4(i,j)] The method of detecting obstacles using image processing algorithms is as follows: Where n is the number of bands of the infrared spectral image, and the pixel value of each band is I ki (i, j) (ki=1,2,…,n), the spectral feature vector F(i,j) of each pixel, B is the binary image of the obstacle; T B is the threshold value.

9. The method for detecting underwater terrain of a dike breach based on a drone as claimed in claim 1, characterized in that: The method of generating a three-dimensional terrain model at the breach in the same coordinate system according to the breach edge information, the corrected water flow depth and the breach obstacle map includes: Eliminate outliers in the corrected water flow depth at each point cloud, and then use the least squares method or iterative closest point algorithm to optimize the positions of all point clouds based on the precise coordinates provided by the Beidou base station, and use linear interpolation to convert the positions of each point cloud into a regular grid in the same coordinate system; The corrected water flow depth of each point cloud excluding outliers is converted into the bottom point cloud height at the breach, the bottom point cloud height is filled into a regular grid, and the height of all bottom positions is determined according to the nearest neighbor method to obtain a three-dimensional height map; The breach edge information and the breach obstacle map are added to the three-dimensional height map to obtain a three-dimensional terrain model at the breach; The calculation method used to remove the abnormal values ​​in the corrected water flow depth at each point cloud is as follows: Among them, d i To correct for water flow depth.

10. A detection device for rapidly constructing underwater terrain of dike breach based on drone, characterized in that: include: Mapping drone, tracer floating objects, first correction module, second image processing module and three-dimensional terrain generation module; The surveying and mapping drone is used to carry a surveying and mapping laser radar, a thermal infrared camera and a visible light camera, and is used to obtain the water flow velocity and water flow depth of multiple point clouds at the breach, and the infrared spectrum image and visible light image of the breach; The tracer float is used to display its migration path and spatial position change, and obtain the water level difference and water flow velocity of each point cloud at the breach; The first correction module is used to correct the water flow depth of each point cloud at the breach based on the water flow velocity and water flow depth of each point cloud at the breach, using a water flow velocity correction algorithm and / or a wave and turbulence compensation algorithm to obtain a corrected water flow depth at each point cloud; The second image processing module is used to identify obstacles at the breach using an image processing algorithm based on an infrared spectrum image to obtain a breach obstacle map; and is also used to obtain edge information at the breach using an image processing algorithm based on a visible light image; The three-dimensional terrain generation module is used to generate a three-dimensional terrain model at the breach in the same coordinate system according to the breach edge information, the corrected water flow depth and the breach obstacle map.

Images