A method for detecting hidden dangers of submarine water intake structures based on combined acoustic and optical detection

Abstract

The present invention relates to the field of submarine structure detection, and in particular to a method for detecting hidden dangers of submarine water intake structures based on combined acoustic and optical detection. This method uses multi-source data acquisition and intelligent fusion technology, including centimeter-level high-resolution sonar sounding, acoustic imaging of water layer structure, ROV optical camera observation, and laser calibration, to comprehensively obtain information on scouring, cracks, and blockages of submarine grids. Adaptive enhancement, threshold segmentation, and spatial interpolation algorithms are used to address the problems of traditional sonar imaging blind spots, blurred target boundaries, and difficulty in quantifying the thickness of biological attachments, thereby achieving high-precision, multi-dimensional detection of submarine structural damage, blockages, and scouring patterns. The overall solution improves the integrity and accuracy of detection, enhances the ability to monitor submarine environmental safety, and provides reliable technical support for the operation and maintenance of submarine facilities.

Description

Technical Field

[0001] The present invention relates to the technical field of submarine structure detection, and in particular to a method for detecting hidden dangers of submarine water intake structures based on combined acoustic and optical detection. Background Art

[0002] Currently, the detection of submarine structures primarily relies on acoustic imaging technology, particularly devices like side-scan sonar and multibeam sonar, which analyze echo signals to construct images of submarine structures. These methods, with their limited penetration and imaging range, have been widely used for tasks such as submarine structure identification and pipeline inspection. However, traditional sonar imaging has significant limitations. Due to the complex and ever-changing submarine environment, structures often exhibit obstructions, sudden changes, or grooves, preventing acoustic waves from effectively covering the target area. This creates numerous blind spots, resulting in missing image information and incomplete structural identification.

[0003] In addition, sonar images have limited resolution, poor detail expression capabilities, and blurred target edges, making it difficult to meet high-precision identification requirements. At the same time, acoustic signals are sensitive to the environment and are easily affected by changes in water temperature, salinity, and background noise, resulting in poor image stability and a low signal-to-noise ratio, further affecting the accuracy of structural positioning and dimensional measurement. More importantly, traditional detection methods in existing technologies have difficulty in simultaneously assessing the scour volume and the degree of blockage due to their single data dimension: although multi-beam sounding can identify macroscopic terrain changes, it cannot penetrate the grid gaps to detect internal siltation; although sonar images can locate structural damage, they are affected by sediment obstruction and have insufficient local resolution; and ROV optical detection is prone to image blur under strong current conditions, making it difficult to quantify the thickness of biological attachments.

[0004] In summary, when facing complex seabed structures and high-precision detection needs, existing acoustic detection technology still has problems such as detection blind spots, low image quality, and insufficient recognition accuracy. There is an urgent need for an improved method that integrates multi-source information and improves detection integrity and accuracy. Summary of the Invention

[0005] In view of the above-mentioned prior art, the present invention provides a method for detecting hidden dangers of submarine water intake structures based on combined acoustic and optical detection, which mainly solves the technical problems existing in the above-mentioned background technology.

[0006] To achieve the above-mentioned purpose, the technical solution of the embodiment of the present invention is implemented as follows:

[0007] A method for detecting hidden dangers of submarine water intake structures based on combined acoustic and optical detection, the method comprising the following steps: step S1: using an acoustic multi-beam sounding system, performing high-density scanning of the seabed around the water intake head through a broadband sonar array to obtain centimeter-level resolution water surface layer topography data, and calculating its overall scouring volume based on the centimeter-level resolution water surface layer topography data; step S2: using a dual-frequency imaging sonar high-frequency channel to identify millimeter-level cracks on the grid surface of the water layer structure and the entanglement of fishing nets, and dynamically adjusting its signal strength through an adaptive time-varying gain compensation algorithm to suppress image blurring caused by changes in water turbidity, and obtaining millimeter-level resolution acoustic image data of the grid surface; step S3: using an ROV optical camera to conduct close observation of the seabed layer, and obtaining the optical image data of the grid surface. Image data, and laser calibration of the optical image data to obtain laser calibration data; step S4: fusing the millimeter-level resolution acoustic image data of the grille surface with the optical image data of the grille surface, and applying adaptive histogram equalization to enhance the target area signal of the fused data, and then extracting the blocked area of the grille surface through the Otsu threshold segmentation algorithm; step S5: fusing the optical image data of the grille surface with the laser calibration data, and performing spatial interpolation compensation on the blind area of the grille backflow surface, extracting the geometric constraints of the blind area boundary, and combining the spatial correlation variation function model of the acoustic data under the geometric constraints of the blind area boundary with the Kriging interpolation algorithm to predict the elevation value of the blind area, and mapping the elevation value to the blind area, and restoring the scouring morphology of the blind area of the grille backflow surface.

[0008] As a preferred solution of the present invention, the step S1 specifically includes the following steps: step S101: using the CUBE algorithm to pre-process the water surface topography data to remove abnormal values caused by multipath effects or suspended matter interference; step S102: based on the baseline water surface topography data of the unwashed area, using the least squares method to fit the seabed surface , its mathematical expression is: in, are polynomial coefficients, express Coordinate direction, express Coordinate direction; Step S103: Use the Delaunay triangulation algorithm to convert the discrete points of the water surface layer terrain data into a continuous triangular mesh and calculate the scour pit volume of each surface element It can be expressed as: in, 、 、 Represents the depth of the scour pit at each vertex of the triangle, Represents the projected area of the triangular surface element; the overall scour pit volume is the sum of the volumes of all surface elements.

[0009] As a preferred solution of the present invention, the step S2 of dynamically adjusting the signal strength by the adaptive time-varying gain compensation algorithm specifically includes: the intensity of the sound wave decays with the propagation distance: in, represents the receiving intensity of the sound wave at a distance r, Indicates the water attenuation coefficient of the received signal, Indicates the distance between the sonar and the target, Indicates the initial intensity of the emitted sound source, represents the exponential function; according to and , dynamically adjust the receiving gain: in, Indicates distance The receiving real-time gain at represents the reference gain, represents the intensity attenuation caused by geometric diffusion, Indicates compensation by water attenuation coefficient Caused by energy loss.

[0010] As a preferred embodiment of the present invention, step S3 further specifically includes quantifying the attachment thickness of the biological community on the grid surface based on the optical image data of the grid surface combined with the laser calibration data: Assuming that the pixel height of the microscopic topography image contour of the biological community on the grid surface in the vertical direction is , the actual thickness for: in, Represents the conversion coefficient from pixel to actual distance; the laser calibration data specifically uses standard deviation to quantify the thickness of the biome attachment: in, represents the thickness standard deviation of all measurement points, represents the total number of all measurement points, Indicates the The actual thickness of each measuring point, Represents the average thickness of all measurement points.

[0011] As a preferred solution of the present invention, the specific process of applying adaptive histogram equalization to enhance the target area signal of the fused data in step S4 includes: step S401: generating a feature map from the fused data, and then dividing the feature map into To avoid discontinuity between blocks, for each sub-block The boundary pixels are interpolated using the CDF of the four adjacent blocks, and then for each sub-block Equalize each sub-block independently. Calculate the grayscale histogram Wherein B is the gray level; Step S402: for each sub-block , the normalized histogram generates the probability distribution: , and calculate the cumulative distribution function: ,Then the original grayscale value is mapped to the equalized value, thereby enhancing the target area.

[0012] As a preferred solution of the present invention, the specific process of extracting the blocked area of the grid surface by the Otsu threshold segmentation algorithm in step S4 includes: dividing the image grayscale of the enhanced target area into two categories: blocked and non-blocked, and finding the optimal threshold value by the following formula between-class variance maximum: in, represents the grayscale threshold, , Indicates the pixel ratio of occlusion in the enhanced target area, Indicates the proportion of non-blocked pixels in the enhanced target area, Indicates the average grayscale of the blocked pixel ratio, Indicates the average grayscale of the non-blocked pixel ratio; then according to the threshold Automatically re-segment blocked and non-blocked areas.

[0013] As a preferred solution of the present invention, the step S5 specifically includes: Step S501: Based on the spatial correlation variation function model of acoustic data To fit the spatial correlation of acoustic data: in, is the lag distance, Indicates the value of gold nuggets, represents the sill value, Represents the distance scale of spatial autocorrelation; Step S502: Calculate the interpolation weight of the known point to be interpolated according to the variogram model: in, is the Lagrange multiplier, Indicates the The known point and The distance between known points, Indicates the The known points and the positions of the points to be inserted The distance between Indicates the The known point and The spatial correlation variation value between known points, Indicates the The known points and the positions of the points to be inserted The spatial correlation variation between Indicates the number of known points in the area adjacent to the blind spot, Indicates the Known points Correct insertion point position Interpolation weight; Step S503: Use Kriging interpolation to predict the blind area elevation value: in, Indicates the location of the point to be inserted The blind spot elevation value, Indicates the elevation value of a known point in the area adjacent to the blind spot.

[0014] Step S504: The predicted insertion point position in step S503 is The blind area elevation value is mapped to the blind area of the grille backflow surface.

[0015] The beneficial effects of the present invention are as follows: this method uses high-resolution multi-beam sonar to measure underwater topography and accurately calculate the volume of scour pits; dual-frequency imaging sonar is used to identify structural cracks and entangled fishing nets to achieve millimeter-level detail detection; and ROV optical imaging and laser calibration are combined to obtain clear and high-quality optical images. Through multi-source data fusion, adaptive enhancement and threshold segmentation to extract blocked areas, the problems of traditional sonar blind spots, low resolution and target blur are solved. At the same time, combined with the Kriging spatial interpolation algorithm, the missing elevation in the blind spots is compensated and the structural scour appearance is restored. This method significantly improves the integrity, resolution and accuracy of submarine structure detection, improves blind spot coverage, effectively identifies hidden dangers, and improves the safety of submarine facilities and the scientific nature of operation and maintenance. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A flowchart of the steps of this method is provided;

[0017] Figure 2 Imaging sonar images for acoustic multi-beam systems;

[0018] Figure 3 Acoustic image of damaged grille and accumulation at the water intake head;

[0019] Figure 4 This is an optical image of the water intake head grille;

[0020] Figure 5 This is a comparison chart before and after enhancement of imaging sonar images;

[0021] Figure 6 Take close-up grid images for ROV optical camera. DETAILED DESCRIPTION

[0022] The technical solution of the present invention is further elaborated in detail below in conjunction with the drawings and specific embodiments of the specification. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which the present invention belongs. The terms used in the specification of the present invention herein are only for the purpose of describing specific embodiments and are not intended to limit the present invention. In the following description, reference is made to "some embodiments", which describes a subset of all possible embodiments, but it should be understood that "some embodiments" can be the same subset or different subsets of all possible embodiments, and can be combined with each other without conflict.

[0023] In the following description, numerous specific details are provided to provide a more thorough understanding of the present invention. However, it will be apparent to those skilled in the art that the present invention may be practiced without one or more of these details. In other instances, certain technical features well known in the art are not described to avoid confusion with the present invention.

[0024] It should be understood that the present invention can be implemented in different forms and should not be interpreted as being limited to the embodiments proposed herein. On the contrary, providing these embodiments will make the disclosure thorough and complete, and will fully convey the scope of the present invention to those skilled in the art. And the purpose of the terms used herein is only to describe specific embodiments and is not intended to limit the present invention. When used herein, the singular forms "one", "an" and "said / the" are also intended to include plural forms, unless the context clearly indicates another way. It should also be understood that the terms "comprising" and / or "comprising" when used in this specification determine the presence of the features, integers, steps, operations, elements and / or parts, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, parts and / or groups. When used herein, the term "and / or" includes any and all combinations of the relevant listed items.

[0025] It should also be noted that when an element is referred to as being "fixed to" another element, it may be directly on the other element or there may be an intermediate element. When an element is referred to as being "connected to" another element, it may be directly connected to the other element or there may be an intermediate element. The terms "vertical," "horizontal," "inner," "outer," "left," "right," and similar expressions used herein are for illustrative purposes only and do not represent the only implementation methods.

[0026] In order to fully understand the present invention, a detailed structure will be provided in the following description to illustrate the technical solution proposed by the present invention. Optional embodiments of the present invention are described in detail below. However, in addition to these detailed descriptions, the present invention may also have other implementations.

[0027] Example 1

[0028] Please refer to the attached Figure 1 The present application provides a method for detecting hidden dangers of submarine water intake structures based on combined acoustic and optical detection, the method comprising the following steps: Step S1: using an acoustic multi-beam bathymetry system to perform high-density scanning of the seabed around the water intake head through a broadband sonar array to obtain centimeter-level resolution water surface topography data, and calculating its overall scour volume based on the centimeter-level resolution water surface topography data;

[0029] Specifically, the multi-beam bathymetric system is used as the basis for macro-topographic detection, and a broadband sonar array (200-400kHz) is used to perform high-density scanning of the seabed around the water head to generate a digital elevation model (DEM) with centimeter-level resolution. In order to address the problem that sound wave propagation is affected by water temperature and salinity fluctuations, the AML SV+ sensor is used to correct the sound velocity profile in real time, and combined with the GNSS / INS combined navigation technology, it ensures that the plane positioning accuracy reaches ±5cm and the elevation error is less than ±2cm. The data at this stage provides a reliable spatial benchmark for the macro-distribution of scour pits. Specifically, the experiment uses the Reson T50 multi-beam bathymetric system (working frequency 200-400kHz, maximum depth 300m), and its strip width and water depth The relationship is:

[0030]

[0031] Among them, the transducer opening angle θ=150°;

[0032] Integrated GNSS / INS navigation achieves a horizontal positioning accuracy of ±5cm and an elevation accuracy of ±2cm. The lateral coverage width is set at four times the water depth to ensure complete coverage of a 20m × 20m area around the water intake. To eliminate acoustic propagation errors, real-time sound velocity profile data (AML SV+ sensor) is collected and dynamically corrected for tidal levels. A 0.1m × 0.1m grid digital elevation model is generated after removing outliers (confidence level > 95%) using the CUBE filtering algorithm. Figure 2 The results show that a scour depression approximately 15 meters long and 10 meters wide exists northeast of the water intake, with an average depth of 4.2 meters and a maximum scour depth of 6.5 meters. Step S2: Use the high-frequency channel of the dual-frequency imaging sonar to identify millimeter-level cracks on the surface of the grid and fishing net entanglements in the water layer structure. Dynamically adjust its signal strength using an adaptive time-varying gain compensation algorithm to suppress image blur caused by changes in water turbidity, and obtain millimeter-level resolution acoustic image data of the grid surface.

[0033] Specifically, to compensate for the limitations of water surface topography data, see Figure 3A dual-frequency imaging sonar (EdgeTech 4200MP) was introduced for water layer structure detection. The high-frequency 1.2MHz channel focuses on the precise identification of millimeter-level cracks on the grille surface and fishing net entanglements. An adaptive time-varying gain (TVG) compensation algorithm dynamically adjusts signal strength to suppress image blur caused by changes in water turbidity. The low-frequency 750kHz channel penetrates surface sediments to detect the thickness of silt deposits within the box culvert, with a maximum penetration depth of 20 meters. Through spatiotemporal registration of acoustic image data with topographic data, the scour pit edge morphology and the location of grille damage can be precisely correlated.

[0034] The technical parameters of the multi-beam bathymetry system are shown in Table 1:

[0035] Table 1 Technical parameters of multi-beam system

[0036]

[0037] Step S3: Using an ROV optical camera to perform close observation of the seabed, obtain optical image data of the grid surface, and perform laser calibration on the optical image data to obtain laser calibration data;

[0038] For details, see Figure 4 To address the limitations of acoustic methods in quantifying microscopic defects, an ROV equipped with an optical camera was deployed to the seabed for close-up observation. In strong currents of 1.5 m / s, the ROV's optical camera, using an inertially stabilized gimbal to suppress motion blur, captured the microscopic morphology of the biomes on the grating surface. Laser calibration data was then used to quantify the attachment thickness (to an accuracy of ±0.2 mm). The fusion of optical and acoustic imaging data further eliminated the problem of missed detection of weakly reflective targets such as biofilms. Step S4: Fuse the millimeter-level resolution acoustic image data of the grille surface with the optical image data of the grille surface, apply adaptive histogram equalization to enhance the target area signal of the fused data, and then extract the blocked area of the grille surface through the Otsu threshold segmentation algorithm; Step S5: Fuse the optical image data of the grille surface with the laser calibration data, perform spatial interpolation compensation on the blind area of the grille backflow surface, extract the geometric constraints of the blind area boundary, and use the spatial correlation variation function model of the acoustic data under the geometric constraints of the blind area boundary, combined with the Kriging interpolation algorithm, to predict the elevation value of the blind area, and map the elevation value to the blind area to restore the scouring morphology of the blind area of the grille backflow surface.

[0039] As a preferred solution of the present invention, the step S1 specifically includes the following steps: step S101: using the CUBE algorithm to pre-process the water surface topography data to remove abnormal values caused by multipath effects or suspended matter interference; step S102: based on the baseline water surface topography data of the unwashed area, using the least squares method to fit the seabed surface , its mathematical expression is: in, are polynomial coefficients, express Coordinate direction, express Coordinate direction;

[0040] Specifically, based on the benchmark terrain data of the unscoured area, a quadratic surface model was constructed, and its fitting residual was controlled within ±3 cm. The elevation difference between the measured terrain and the benchmark surface was converted into a continuous triangular mesh through the Delaunay triangulation algorithm. The final cumulative scour volume was 625.8 m³, accounting for 18% of the water head design volume. In order to verify the robustness of the algorithm, the present invention conducted 1000 Monte Carlo simulations, and superimposed Gaussian noise with a mean of 0 and a standard deviation of 5 cm on the original data. The results showed that the scour volume error was stable at 3.2%-4.7% (confidence level 95%). The edge of the scour pit on the northeast side was sampled by ROV optical photography. The deviation between the measured depth and the calculated value was only 4.3%, indicating the reliability of this method under complex terrain. Specifically, the polynomial coefficients were determined by optimizing the regional terrain characteristics, and the fitting residual was controlled within ±3 cm. Measured terrain surface Elevation difference from datum That is the spatial distribution of the scour depth. In order to accurately calculate the volume of the scour pit, the Delaunay triangulation algorithm is used to convert the discrete point cloud into a continuous triangular mesh. The algorithm generates an optimal triangulation network under complex boundary conditions by maximizing the minimum internal angle criterion, effectively avoiding the integration error caused by narrow triangles. Step S103: Use the Delaunay triangulation algorithm to convert the discrete points of the water surface layer terrain data into a continuous triangular mesh and calculate the scour pit volume of each face element. It can be expressed as: in, 、 、 Represents the depth of the scour pit at each vertex of the triangle, Represents the projected area of the triangular surface element; the overall scour pit volume is the sum of the volumes of all surface elements.

[0041] As a preferred solution of the present invention, the step S2 of dynamically adjusting the signal strength by the adaptive time-varying gain compensation algorithm specifically includes: the intensity of the sound wave decays with the propagation distance: in, Indicates the distance of sound waves The receiving strength at Indicates the water attenuation coefficient of the received signal, Indicates the distance between the sonar and the target, Indicates the initial intensity of the emitted sound source, represents the exponential function; according to and , dynamically adjust the receiving gain: in, Indicates distance The receiving real-time gain at represents the reference gain, represents the intensity attenuation caused by geometric diffusion, Indicates compensation by water attenuation coefficient Caused by energy loss.

[0042] As a preferred embodiment of the present invention, step S3 further specifically includes quantifying the attachment thickness of the biological community on the grid surface based on the optical image data of the grid surface combined with the laser calibration data: Assuming that the pixel height of the microscopic topography image contour of the biological community on the grid surface in the vertical direction is , the actual thickness for: in, Represents the conversion coefficient from pixel to actual distance; the laser calibration data specifically uses standard deviation to quantify the thickness of the biome attachment: in, represents the thickness standard deviation of all measurement points, represents the total number of all measurement points, Indicates the The actual thickness of each measuring point, Represents the average thickness of all measurement points.

[0043] This process improves the visibility of the attachment layer edge through image preprocessing, and combined with the precise dimensional reference provided by laser calibration, it enables millimeter-level quantitative measurement of the thickness of the seafloor biome. This provides fundamental quantitative data for environmental monitoring, structural assessment, and blind spot repair.

[0044] As a preferred solution of the present invention, the specific process of applying adaptive histogram equalization to enhance the target area signal of the fused data in step S4 includes: step S401: generating a feature map from the fused data, and then dividing the feature map into To avoid discontinuity between blocks, for each sub-block The boundary pixels are interpolated using the CDF of the four adjacent blocks, and then for each sub-block Equalize each sub-block independently. Calculate the grayscale histogram Wherein B is the gray level; Step S402: for each sub-block , the normalized histogram generates the probability distribution: , and calculate the cumulative distribution function: ,Then the original grayscale value is mapped to the equalized value, thereby enhancing the target area.

[0045] As a preferred solution of the present invention, the specific process of extracting the blocked area of the grid surface by the Otsu threshold segmentation algorithm in step S4 includes: dividing the image grayscale of the enhanced target area into two categories: blocked and non-blocked, and finding the optimal threshold value by the following formula between-class variance maximum: in, represents the grayscale threshold, , Indicates the pixel ratio of occlusion in the enhanced target area, Indicates the proportion of non-blocked pixels in the enhanced target area, Indicates the average grayscale of the blocked pixel ratio, Indicates the average grayscale of the non-blocked pixel ratio; then according to the threshold Automatically re-segment blocked and non-blocked areas.

[0046] For details, see Figure 5 Adaptive histogram equalization was applied to 1.2MHz high-frequency sonar images to enhance contrast in target areas, followed by Otsu threshold segmentation to identify suspected blockage areas. Taking a 1m×1m weakly reflective area as an example, the enhanced image clearly shows the outline of a grid break, and the area was calculated to be 0.82m². Specifically, quantifying the grid blockage area requires combining acoustic image data with optical image data of the grid surface. The 1.2MHz high-frequency channel of the EdgeTech 4200MP sonar was used to acquire millimeter-level resolution acoustic images of the grid surface. To address the image contrast loss caused by water attenuation, adaptive histogram equalization was applied to enhance the target area signal, followed by Otsu threshold segmentation to identify suspected blockage areas, such as entangled fishing nets or sediment accumulation. Taking a 1m×1m weakly reflective area as an example, after acoustic image enhancement and segmentation, the grid break area was identified as 0.82m². Furthermore, ROV optical imagery showed a 73% fishing net coverage rate in the area, verifying the reliability of the acoustic area calculation.

[0047] See also Figure 6 ROV optical camera footage showed that fishing net coverage in the area reached 73%, with biofouling thickness ranging from 8-12mm. The calculated flow area loss rate was 32.1%. Of the eight groups of screens, three were found to be blocked beyond the threshold.

[0048] As a preferred solution of the present invention, the step S5 specifically includes: Step S501: Based on the spatial correlation variation function model of acoustic data To fit the spatial correlation of acoustic data: in, is the lag distance, Indicates the value of gold nuggets, represents the sill value, Represents the distance scale of spatial autocorrelation; Step S502: Calculate the interpolation weight of the known point to be interpolated according to the variogram model: in, is the Lagrange multiplier, Indicates the The known point and The distance between known points, Indicates the The known points and the positions of the points to be inserted The distance between Indicates the The known point and The spatial correlation variation value between known points, Indicates the The known points and the positions of the points to be inserted The spatial correlation variation between Indicates the number of known points in the area adjacent to the blind spot, Indicates the Known points Correct insertion point position Interpolation weight; Step S503: Use Kriging interpolation to predict the blind area elevation value: in, Indicates the location of the point to be inserted The blind spot elevation value, Indicates the elevation value of a known point in the area adjacent to the blind spot.

[0049] Step S504: The predicted insertion point position in step S503 is The blind spot elevation value is mapped to the blind spot of the backflow surface of the grille. Specifically, there are detection blind spots in the backflow surface of the grille and the inside of the box culvert in the acoustic multi-beam, resulting in the lack of scour terrain data. To this end, the present invention uses ROV optical imaging and laser calibration data to perform spatial interpolation compensation for the blind spot. High-resolution images are acquired by close-up shooting with ROV optical imaging, and the three-dimensional coordinates of the feature points are measured (accuracy ±0.2mm) in combination with the laser calibration module to extract the geometric constraints of the blind spot boundary; then, the spatial correlation variation function model of the acoustic data is combined with the Kriging interpolation algorithm to predict the elevation value of the blind spot area. For example, the backflow surface of the grille at a certain place is missing 30% of the data due to sound wave obstruction. After optical constraint interpolation, the elevation estimation error is reduced from ±15cm of traditional linear interpolation to ±9cm, effectively restoring the scour morphology of the area. Specifically, through this method, combined with the historical operation and maintenance data of the past five years and the regional geological characteristics (the bottom soil is silty clay, the median particle size , critical starting flow velocity 0.2m / s; maximum flow velocity difference between high and low tides 0.7m / s). This survey identified three typical hidden dangers. The specific results are shown in Table 2:

[0050] Table 2 Hazard points and repair suggestions

[0051]

[0052] This invention systematically solves the challenge of diagnosing the coupled damage of scouring and grid blockage in submarine water intake structures through multi-source data fusion technology. Engineering practice has demonstrated that this three-dimensional detection system, integrating multi-beam bathymetry, dual-frequency imaging sonar, and ROV optical detection, can effectively identify hidden defects that are difficult to detect using traditional single-source methods.

[0053] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present invention should be included in the scope of protection of the present invention. The scope of protection of the present invention should be based on the scope of protection of the claims.

Claims

1. A method for detecting hidden dangers of submarine water intake structures based on combined acoustic and optical detection, characterized in that: The method comprises the following steps: Step S1: Using an acoustic multi-beam bathymetry system, a broadband sonar array is used to perform high-density scanning of the seabed around the water intake head to obtain centimeter-level resolution water surface topography data. The overall scour pit volume is calculated based on the centimeter-level resolution water surface topography data. Step S2: Using the high-frequency channel of the dual-frequency imaging sonar, the millimeter-level cracks on the grille surface and the entanglement of fishing nets in the water layer structure are identified. The signal strength is dynamically adjusted using an adaptive time-varying gain compensation algorithm to suppress image blur caused by changes in water turbidity, thereby obtaining millimeter-level resolution acoustic image data of the grille surface. Step S3: Using an ROV optical camera to perform close observation of the seabed, obtain optical image data of the grid surface, and perform laser calibration on the optical image data to obtain laser calibration data; Step S4: Fusing the millimeter-level resolution acoustic image data of the grille surface with the optical image data of the grille surface, applying adaptive histogram equalization to enhance the target area signal of the fused data, and then extracting the blocked area of the grille surface using the Otsu threshold segmentation algorithm; Step S5: Fusing the optical image data of the grille surface with the laser calibration data, and performing spatial interpolation compensation on the blind area of the grille backflow surface, extracting the geometric constraints of the blind area boundary, and using the spatial correlation variogram model of the acoustic data under the geometric constraints of the blind area boundary, combined with the Kriging interpolation algorithm, predicting the elevation value of the blind area, and mapping the elevation value to the blind area, thereby restoring the scour morphology of the blind area of the grille backflow surface; Step S501: Spatial correlation variogram model based on acoustic data To fit the spatial correlation of acoustic data: in, is the lag distance, Indicates the value of gold nuggets, represents the sill value, The distance scale representing the effect of spatial autocorrelation; Step S502: Calculate the interpolation weights of the known points to the points to be interpolated based on the variance function model: in, is the Lagrange multiplier, Indicates the The known point and The distance between known points, Indicates the The known points and the positions of the points to be inserted The distance between Indicates the The known point and The spatial correlation variation value between known points, Indicates the The known points and the positions of the points to be inserted The spatial correlation variation between Indicates the number of known points in the area adjacent to the blind spot, Indicates the Known points Treat insertion point location The interpolation weight of Step S503: Use Kriging interpolation to predict the blind area elevation value: in, Indicates the location of the point to be inserted The blind spot elevation value, Indicates the elevation value of a known point in the adjacent area of the blind zone; Step S504: The predicted insertion point position in step S503 is The blind area elevation value is mapped to the blind area of the grille backflow surface.

2. The method for detecting hidden dangers of submarine water intake structures based on combined acoustic and optical detection according to claim 1, characterized in that: The step S1 specifically includes the following steps: step S101: pre-processing the water surface topography data using the CUBE algorithm to remove abnormal values caused by multipath effects or suspended matter interference; step S102: fitting the seabed surface using the least squares method based on the baseline water surface topography data of the unwashed area , its mathematical expression is: in, are polynomial coefficients, express Coordinate direction, express Coordinate direction; Step S103: Use the Delaunay triangulation algorithm to convert the discrete points of the water surface terrain data into a continuous triangular mesh and calculate the volume of the scour pit of each surface element. It can be expressed as: ,in, 、 、 Represents the depth of the scour pit at each vertex of the triangle, It represents the projected area of the triangular surface element, and the overall scour pit volume is the sum of the volumes of all surface elements.

3. The method for detecting hidden dangers of submarine water intake structures based on combined acoustic and optical detection according to claim 2, characterized in that: The process of dynamically adjusting the signal strength by the adaptive time-varying gain compensation algorithm in step S2 specifically includes: the intensity of the sound wave decays with the propagation distance: in, represents the receiving intensity of the sound wave at a distance r, Indicates the water attenuation coefficient of the received signal, Indicates the distance between the sonar and the target, Indicates the initial intensity of the emitted sound source, represents the exponential function; according to and , dynamically adjust the receiving gain: in, Indicates distance The receiving real-time gain at represents the reference gain, represents the intensity attenuation caused by geometric diffusion, Indicates compensation by water attenuation coefficient Caused by energy loss.

4. The method for detecting hidden dangers of submarine water intake structures based on combined acoustic and optical detection according to claim 3 is characterized in that: The step S3 further specifically includes quantifying the attachment thickness of the biological community on the grid surface based on the optical image data of the grid surface combined with the laser calibration data: Assuming that the pixel height of the microscopic topography image contour of the biological community on the grid surface in the vertical direction is , the actual thickness for: in, Represents the conversion coefficient from pixel to actual distance; the laser calibration data specifically uses standard deviation to quantify the thickness of the biome attachment: in, represents the thickness standard deviation of all measurement points, represents the total number of all measurement points, Indicates the The actual thickness of each measuring point, Represents the average thickness of all measurement points.

5. The method for detecting hidden dangers of submarine water intake structures based on combined acoustic and optical detection according to claim 4 is characterized in that: The specific process of applying adaptive histogram equalization to enhance the target area signal of the fused data in step S4 includes: step S401: generating a feature map from the fused data, and then dividing the feature map into To avoid discontinuity between blocks, for each sub-block The boundary pixels are interpolated using the CDF of the four adjacent blocks, and then for each sub-block Equalize each sub-block independently. Calculate the grayscale histogram , where B is the gray level; Step S402: for each sub-block , the normalized histogram generates the probability distribution: , and calculate the cumulative distribution function: ,Then the original grayscale value is mapped to the equalized value, thereby enhancing the target area.

6. The method for detecting hidden dangers of submarine water intake structures based on combined acoustic and optical detection according to claim 5, characterized in that: The specific process of extracting the blocked area of the grid surface by the Otsu threshold segmentation algorithm in step S4 includes: dividing the image grayscale of the enhanced target area into two categories of blocked and non-blocked, and finding the optimal threshold by the following formula between-class variance maximum: in, represents the grayscale threshold, , Indicates the pixel ratio of occlusion in the enhanced target area, Indicates the proportion of non-blocked pixels in the enhanced target area, Indicates the average grayscale of the blocked pixel ratio, Indicates the average grayscale of the non-blocked pixel ratio; then according to the threshold Automatically re-segment blocked and non-blocked areas.

Images