Underwater acoustic-magnetic multi-modal fusion processing method

Abstract

The invention discloses an underwater acoustic-magnetic multi-modal fusion processing method, which comprises the following steps: carrying out preprocessing and time-space alignment on collected sonar and magnetic data, and extracting a morphological feature vector and a magnetic feature vector of a pipeline; based on a pipeline route trend, calculating an optimal matching path of a sonar contour and a magnetic force abnormal zone through dynamic time warping, establishing acoustic-magnetic mapping, and fusing features to generate a joint vector; finally, the type and state of the pipeline are judged according to the combined characteristics, and the problem that in a near-shore port and an industrial area water area, due to the interference of construction waste and the influence of a pipeline coating layer, a single sonar is difficult to accurately detect the pollution discharge pipeline and the cable is solved. According to the method, the complementarity of acoustic and magnetic characteristics is utilized, the accuracy of pipeline detection in a complex scene is improved, and reliable data support is provided for operation and maintenance of near-shore water area pipelines.

Description

Technical Field

[0001] The present invention relates to the technical field of similar devices using other waves, in particular to acoustic waves and magnetism, and specifically to an underwater acoustic-magnetic multimodal fusion processing method. Background Art

[0002] Currently, with the rapid development of fields such as marine resource development, marine engineering construction, and marine monitoring, the demand for underwater target detection and monitoring is growing. In many underwater detection scenarios, such as submarine pipeline inspection, underwater facility status assessment, and underwater obstacle detection, accurate target information acquisition is crucial. Acoustic wave detection technology, especially various types of sonar, has become the core means of underwater target detection due to its advantage in underwater propagation. For example, side-scan sonar and multi-beam echo sounders can efficiently survey the seabed topography and the status of exposed pipeline sections. Synthetic aperture sonar can provide high-resolution images to identify pipeline surface details. Shallow subsurface profilers can detect relevant information about shallow buried pipelines. Acoustic positioning systems provide positioning for detection platforms and targets.

[0003] Existing technologies that rely solely on acoustic wave detection, especially technical solutions using a single type of sonar, have obvious limitations. In terms of buried pipeline detection, acoustic waves are limited by the type and frequency of the substrate, and their penetration depth for deeply buried pipelines is insufficient, and the resolution drops sharply with depth. Strong clutter and multiple reflections generated by geological structures such as rocks and gravel in complex substrates can mask pipeline signals or create false targets. When a pipeline has a special coating that causes its acoustic properties to be close to those of the surrounding sediments, an acoustic "stealth" effect occurs, and the acoustic wave reflection / scattering signal is weak and difficult to identify. In addition, sound waves cannot penetrate metal pipe walls, cannot detect the internal state of the pipeline, cannot distinguish between pipeline materials, and cannot evaluate the ferromagnetism of metal pipelines.

[0004] Therefore, it is necessary to improve the underwater acoustic and magnetic multimodal fusion processing method in the existing technology to solve the above problems. Summary of the Invention

[0005] The present invention overcomes the shortcomings of the existing technology and provides an underwater acoustic and magnetic multimodal fusion processing method, which aims to solve the problem in the existing technology that in the waters of near-shore ports and industrial areas, strong acoustic clutter is caused by construction waste, resulting in single sonar detection being difficult to identify sewage pipelines due to acoustic stealth formed by the coating layer, and the material and corrosion status of cables cannot be judged by sound waves.

[0006] To achieve the above-mentioned purpose, the technical solution adopted by the present invention is: an underwater acoustic and magnetic multimodal fusion processing method, comprising: S1. Preprocessing underwater pipeline sonar data and magnetic data; S2, extracting the pipeline's morphological feature vector from the sonar data, and extracting the pipeline's magnetic feature vector from the magnetic data; S3. Based on the pipeline route, the matching degree between the sonar profile and the magnetic anomaly zone is calculated. The optimal path is calculated through dynamic time warping to establish an acoustic-magnetic mapping. The morphological feature vector and the magnetic feature vector are then fused to generate a joint feature vector. The step of calculating the matching degree in step S3 includes: Extract pipeline strike angle from sonar data as sonar profile strike, and extract magnetic anomaly strike angle from magnetic data; Calculate the angle difference between two strike angles ,in, is the sonar profile strike angle, is the strike angle of the magnetic anomaly belt, when and When they are completely consistent, M=1; when they are perpendicular, M=0; when the matching degree corresponding to the difference value is ≥0.8, they are determined to be the same pipeline feature; S4. Determine the pipeline type and pipeline status based on the joint feature vector; wherein the pipeline type is a sewage pipeline and a cable.

[0007] In a preferred embodiment of the present invention, the pre-processing of the sonar data in step S1 includes: The non-local mean filtering algorithm is used to suppress the strong reflection clutter generated by construction waste by calculating the neighborhood pixel similarity weights, and retain the continuous contour signal of the pipeline.

[0008] In a preferred embodiment of the present invention, the pre-processing of magnetic data in step S1 includes: The geomagnetic model is used to calculate the theoretical geomagnetic field strength, and this value is subtracted from the original magnetic data to obtain the residual magnetic field data; The residual magnetic field data were filtered with a sliding window to remove the high-frequency magnetic field fluctuations caused by the anchoring of the ship.

[0009] In a preferred embodiment of the present invention, the sonar data includes: side scan sonar echo intensity data and shallow layer profiler reflection coefficient data; the magnetic data includes: magnetic field total intensity data and magnetic field gradient data; Preprocessing also includes time and space alignment. The specific steps are: achieving time synchronization between the sonar device and the magnetic device through hardware trigger pulses, combining GNSS positioning data and INS attitude data to convert the range-angle coordinates of the side-scan sonar into the WGS84 geographic coordinate system, and unifying the magnetic sensor coordinates into the same geographic coordinate system through cable position calibration.

[0010] In a preferred embodiment of the present invention, in step S2, the morphological feature vector includes pipeline length, pipeline width, pipeline strike angle, and contour continuity; the magnetic feature vector includes magnetic anomaly peak, magnetic anomaly range, magnetic field gradient maximum value, and magnetic anomaly symmetry.

[0011] In a preferred embodiment of the present invention, the specific steps of dynamic time warping in step S3 include: Extract the acoustic centerline coordinate sequence and the magnetic anomaly trajectory coordinate sequence, and construct the spatial distance matrix between the two sequences; Solve the optimal path with the minimum cumulative distance through dynamic programming and establish a point-to-point mapping relationship; When the cumulative distance of the optimal path is ≤ 0.5 m, the morphological eigenvector and the magnetic eigenvector are bound.

[0012] In a preferred embodiment of the present invention, the specific steps of feature fusion in step S3 include: Normalize the morphological and magnetic eigenvectors and map each characteristic parameter to the [0,1] interval; Based on the acoustic-magnetic mapping relationship, three types of interaction features, material correlation factor, spatial consistency factor and coupling coefficient, are calculated; Combined with dynamic compression, a 6-dimensional joint feature vector containing anti-interference factors, scale-invariant features and scale ratio is constructed.

[0013] In a preferred embodiment of the present invention, the material correlation factor is constructed by fusing the magnetic anomaly peak, the maximum value of the magnetic field gradient and the pipeline width; the spatial consistency factor is calculated based on the spatial deviation of the dynamic time warping mapping point; and the coupling coefficient reflects the coupling relationship between the magnetic anomaly range and the acoustic magnetic strike deviation.

[0014] In a preferred embodiment of the present invention, the specific steps of determining the pipeline type in step S4 include: When the magnetic anomaly peak value is greater than 500nT and the magnetic anomaly symmetry is greater than 0.7, it is determined to be a cable; When the magnetic anomaly peak value is less than 200nT, the contour continuity is greater than 0.8, and the coupling coefficient is less than 0.3, it is determined to be a sewage pipeline containing a coating layer; The rest of the cases are judged as standard sewage pipelines.

[0015] The present invention solves the defects existing in the background technology and has the following beneficial effects: (1) The present invention proposes an underwater acoustic-magnetic multimodal fusion processing method to solve the problem that in the waters of near-shore ports and industrial areas, a single sonar is difficult to accurately detect sewage pipelines and cables due to interference from construction waste and the influence of pipeline coatings. First, the sonar and magnetic data of the underwater pipeline are pre-processed to remove construction waste clutter and ship anchoring interference; then the morphological feature vectors and magnetic feature vectors of the pipeline are extracted respectively; then, based on the route direction of the pipeline, an acoustic-magnetic mapping is established by calculating the matching degree and dynamic time warping, and a joint vector is generated by fusion; finally, the type and status of the sewage pipeline and cable are judged based on the joint features, and the complementarity of acoustic-magnetic features is used to improve the detection accuracy in complex scenarios.

[0016] (2) The present invention calculates the matching degree between the sonar profile direction and the magnetic anomaly belt direction, and combines the dynamic time warping (DTW) algorithm to establish an acoustic-magnetic mapping to achieve feature fusion; the matching degree of the direction ensures that the sonar and magnetic features belong to the same pipeline from a macro perspective, and DTW adapts to the local bending or offset of the pipeline through flexible alignment, accurately binding the morphological feature vector and the magnetic feature vector; it can effectively avoid feature dislocation caused by irregular pipeline shape or environmental interference; the feature fusion of the existing technology does not consider the accuracy of spatial correlation, and it is easy to mistakenly fuse the features of different targets, resulting in distorted results; the precise mapping of the present invention enables the joint feature vector to truly reflect the comprehensive characteristics of the pipeline.

[0017] (3) The present invention generates the features of a joint vector containing interactive features through feature fusion. After normalizing the morphological and magnetic features, the interactive features such as material correlation factor and spatial consistency factor are calculated to construct a multi-dimensional joint vector. The interactive features explore the intrinsic relationship between morphological and magnetic features. For example, the material correlation factor combines magnetic anomalies with pipeline width to reflect material characteristics, adapting to the differences in pipeline materials in nearshore scenarios, the ferromagnetism of cables and the non-metallic properties and coating of sewage pipes. The joint vector can reflect both the external morphology of the pipeline and its material properties and spatial consistency. Compared with the fusion method of the existing technology that simply splices features and does not consider the correlation between features, the joint vector information of this method is more comprehensive. A further effect is that the ability to distinguish between sewage pipelines containing coatings and cables and the accuracy of pipeline status judgment are significantly improved, providing a reliable basis for the operation and maintenance of pipelines in nearshore waters.

[0018] (4) This invention addresses the strong reflection clutter generated by construction waste in nearshore waters. Sonar data is filtered using non-local means to preserve the continuous outline of pipelines. To address the high-frequency magnetic field fluctuations caused by anchored ships, magnetic data is calculated using a geomagnetic model and filtered using a sliding window to extract the residual magnetic field. This method accurately locates the main interference sources in the scene, suppresses clutter, and removes fluctuations, effectively preserving the effective signal of the pipeline. This makes the pipeline outline clearer in the sonar data and the magnetic anomaly signals more prominent in the magnetic data. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments described in the present invention. Those skilled in the art can also derive other drawings based on these drawings without inventive efforts. Figure 1 is an overall flow chart of a preferred embodiment of the present invention; Figure 2 is a pre-processing flow chart of a preferred embodiment of the present invention; Figure 3 It is a feature extraction flow chart of a preferred embodiment of the present invention. DETAILED DESCRIPTION

[0020] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0021] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the scope of protection of the present invention is not limited to the specific embodiments disclosed below.

[0022] Application Overview: The application scenarios of this application focus on the waters of near-shore ports and industrial areas, which are greatly disturbed by ship anchoring and contain a large amount of construction waste in the bottom, such as concrete blocks and steel bars. This application is aimed at submarine pipelines, especially sewage pipelines and cables in the above scenarios. Conventional technology has long relied on acoustic wave detection. Because of its long underwater propagation distance and ability to directly obtain target morphological information, it has become the mainstream means of underwater detection. This path dependence has led to the neglect of its inherent defects, which are: single acoustic wave detection technology in this scenario will generate strong clutter and multiple reflections due to the presence of a large amount of construction waste in the bottom, interfering with the identification of pipeline signals; at the same time, the coating layer formed on the sewage pipeline due to long-term use makes its acoustic characteristics close to the surrounding environment, resulting in weak acoustic wave reflection signals, and the metal material of the cable cannot be accurately identified and evaluated by acoustic waves.

[0023] Conventional thinking is limited to optimizing acoustic wave detection technology, attempting to reduce the impact of clutter by enhancing signals and improving algorithms, but it is still impossible to get rid of the inherent defects of acoustic waves in complex bottoms and special coatings.

[0024] This application breaks away from the norm and, based on an in-depth analysis of specific scenarios, finds that: the ferromagnetism of metal pipelines such as cables is not affected by construction waste and coatings, and can be used as a stable detection mark; the propagation of the magnetic field in this scenario is not disturbed by construction waste, and can stably capture the target signal; the external morphological information obtained by sound waves and the material information obtained by magnetism are correlated in the route of the pipeline. Sonar data is good at capturing the external morphology of pipelines, but cannot reflect the material and internal state; magnetic data is good at reflecting the material properties of pipelines, but cannot provide morphological details. Based on this, the acoustic wave and magnetic detection are integrated to form a complementary rather than a simple superposition. By establishing a feature correlation model between the two, accurate verification of sewage pipelines and cables in this scenario can be achieved.

[0025] Exemplary methods: like Figure 1 As shown, an underwater acoustic and magnetic multimodal fusion processing method includes the following steps: S1. Preprocessing underwater pipeline sonar data and magnetic data; S2, extracting the pipeline's morphological feature vector from the sonar data, and extracting the pipeline's magnetic feature vector from the magnetic data; S3. Based on the pipeline route, the matching degree between the sonar profile and the magnetic anomaly zone is calculated. The optimal path is calculated through dynamic time warping to establish an acoustic-magnetic mapping. The morphological feature vector and the magnetic feature vector are fused to generate a joint feature vector. S4. Determine the pipeline type and pipeline status based on the joint feature vector, wherein the pipeline type is a sewage pipeline and a cable.

[0026] Among them, sonar data refers to the data generated by emitting sound waves into the water through sonar equipment, receiving reflected or scattered echo signals and processing them. It can reflect the geometric shape and position information of underwater targets, has high spatial resolution but is easily affected by environmental interference.

[0027] Specifically, side-scan sonar and shallow-sediment profilers are used to obtain the data. The side-scan sonar emits sound waves to both sides, receives echoes reflected by seabed targets, and generates two-dimensional image data; the shallow-sediment profiler emits low-frequency sound waves underwater, penetrates the surface sediments of the seabed, receives reflected signals from different stratum interfaces, and forms profile data.

[0028] In the context of nearshore ports and industrial waters in this application, sonar data includes: side-scan sonar echo intensity data and shallow subsurface profiler reflection coefficient data. Side-scan sonar echo intensity data reveals the external contours and surface features of sewage pipelines and cables, as well as the distribution of surrounding construction waste, identifying the general shape and location of the target; shallow subsurface profiler reflection coefficient data reflects the buried depth of the pipeline and the interface characteristics with the surrounding sediments, providing a basis for determining the pipeline's burial state.

[0029] The side-scan sonar equipment is fixed on the sides of the survey vessel or on an underwater robot to ensure that the direction of sound wave emission covers the surrounding areas of port terminals, the extension areas of industrial sewage outlets, and waters where known pipelines may pass; high-frequency sound waves of 200-500kHz are selected to improve resolution and avoid excessive scattering of low-frequency sound waves in shallow waters; the emission beam angle is controlled at 10-20° to reduce sidelobe interference.

[0030] When sound waves hit underwater targets, including sewage pipes, cables, and construction waste, they are reflected and the echo signals are captured by the receiving transducer. After pre-processing, they are converted into electrical signals. The echo intensity data is presented as pixel values ​​in a two-dimensional image. The shallow layer profiler is installed in the center of the bottom of the survey ship, with the transmitting transducer facing directly downward. It targets nearshore seabed sediments, mainly silt and fine sand, with a thickness of usually 1-5 meters. Low-frequency sound waves of 500-2000Hz are selected to ensure penetration depth; The device travels along a survey line perpendicular to the coastline, emitting low-frequency sound waves that penetrate the seafloor surface. These waves are reflected at interfaces between different media, including the seafloor surface, pipeline walls, and sediment layers. A receiving transducer captures the reflected signal and processes it to calculate the reflection coefficient. This reflection coefficient data is presented as a profile.

[0031] Magnetic data refers to data obtained by measuring underwater magnetic field strength and changes using magnetic detection equipment. It reflects the ferromagnetic characteristics of the target, is unaffected by water and sediment, and has strong penetrating power. This application uses fluxgate magnetometers and fiber-optic magnetic sensors to obtain this data. Fluxgate magnetometers determine magnetic field strength by measuring the induced electromotive force generated by the magnetic field on a sensitive element; fiber-optic magnetic sensors use the magneto-optical effect to measure magnetic fields by detecting changes in the propagation characteristics of light in an optical fiber.

[0032] In this application scenario, magnetic data includes total magnetic field intensity data collected by a fluxgate magnetometer and magnetic field gradient data collected by a fiber optic magnetic sensor. Total magnetic field intensity data can be used to identify ferromagnetic cables, as the cable's metal armor sheath can cause abnormal magnetic field variations. Magnetic field gradient data can more accurately locate pipelines, distinguishing them from magnetic anomalies caused by surrounding construction waste, and reducing interference.

[0033] The fluxgate magnetometer is suspended 50-100 meters behind the survey vessel by a towed cable, at a depth of 3-5 meters underwater, with a sampling frequency of 1-10Hz. It measures the total magnetic field intensity at its location in real time. The total magnetic field intensity data is a continuous numerical sequence, including: background magnetic field data, magnetic anomaly data caused by the cable, and magnetic field data in the sewage pipeline area. The fiber-optic magnetic sensor adopts an array design and is fixed to the robotic arm of the underwater robot. The laser light source uses a 632.8nm helium-neon laser, and the light detector sampling frequency is 10-50Hz to capture subtle changes in the magnetic field. The sensor array synchronously measures the magnetic field intensity at different spatial points. The magnetic field gradient is calculated by calculating the ratio of the difference in magnetic field intensity between adjacent sensor units to the spacing between them. The magnetic field gradient data is a sequence of spatial gradient values, including the gradient peak at the cable location, the gradient data in the construction waste area, and the gradient data in the background area.

[0034] like Figure 2 As shown in FIG, the preprocessing includes denoising of sonar data and magnetic data.

[0035] Among them, for sonar data, the non-local mean filtering algorithm is used for denoising, and the window size is set to 0.7×0.7m. Since the construction waste in the scene is mostly 0.5-1m, this window can effectively distinguish isolated clutter from continuous pipeline contours. For the strong reflected clutter generated by construction waste in near-shore ports and industrial waters, the similarity weights of neighboring pixels are calculated to suppress isolated high-value noise points and retain the continuous contour signal of the pipeline.

[0036] The core calculation formula is as follows: , in, Indicates the coordinates of the side scan sonar echo intensity data The original pixel value of , which reflects the sound wave reflection intensity at the corresponding location in the nearshore port and industrial area waters; Indicates that the coordinates after filtering are The pixel value of , that is, the pixel value of the sonar data after denoising; Indicates The neighborhood window centered on is a window containing multiple pixels. , used to select and Pixels with similar characteristics; Represents pixel points Pixel The weight is determined by calculating the similarity of the local area where the two pixels are located. The higher the similarity, the greater the weight, thereby suppressing the isolated high-value noise points generated by construction waste and retaining the continuous contour signal of the pipeline.

[0037] Because the strong reflection of construction waste will mask the pipeline signal and affect the accuracy of subsequent feature extraction, after processing, it can effectively remove clutter interference, make the pipeline outline clearer, and improve the signal-to-noise ratio.

[0038] Among them, for magnetic data, the IGRF geomagnetic model is first called to calculate the theoretical geomagnetic field strength of the detection area, and the theoretical geomagnetic field strength is subtracted from the original magnetic data to obtain the residual magnetic field data ΔT to eliminate the influence of the earth's main magnetic field; then based on the ship's navigation trajectory and electromagnetic interference model, the ΔT data is subjected to sliding window filtering to remove the high-frequency magnetic field fluctuations caused by anchored ships and retain the low-frequency pipeline magnetic anomaly signals.

[0039] In magnetic data preprocessing, the formula for calculating the residual magnetic field data is as follows: , in, Indicates that the coordinates of the nearshore port and industrial area waters are The residual magnetic field data at the time t can highlight the local magnetic anomalies caused by sewage pipelines and cables, especially metal armored sheathed cables, and eliminate the masking effect of the Earth's main magnetic field. Indicates the fluxgate magnetometer at coordinate , the original total magnetic field intensity data collected at time t, which includes multiple components such as the Earth's main magnetic field, pipeline magnetic anomalies, and ship anchoring interference; Indicates the coordinates calculated by calling the IGRF geomagnetic model. , the theoretical geomagnetic field intensity at time t, which reflects the benchmark intensity of the Earth's main magnetic field in the nearshore detection area.

[0040] Since the Earth's main magnetic field and ship interference can mask the magnetic anomalies of the pipeline, the above processing can highlight the magnetic characteristics of the pipeline, lay the foundation for subsequent magnetic anomaly extraction, and improve the reliability of the data.

[0041] Spatiotemporal alignment uses hardware triggering to synchronize the time of sonar and magnetic data, ensuring a timestamp error of ≤1ms. Combining GNSS and INS positioning data, the relative distance coordinates of the sonar and the sensor coordinates of the magnetic field are unified into the same geographic coordinate system, keeping them spatially consistent.

[0042] To ensure strict matching of sonar data and magnetic data in the time dimension, a multi-device hardware synchronization trigger solution is adopted: a high-precision clock module is integrated into the main control system, and the acquisition and control units of the side-scan sonar, shallow subsurface profiler, fluxgate magnetometer and fiber optic magnetic sensor are connected respectively through synchronization trigger lines.

[0043] When the system starts, the main control module pre-processes the underwater pipeline sonar data and magnetic data every 1ms and sends a synchronization pulse signal. After receiving the pulse, each device immediately starts data collection and embeds the timestamp of the pulse into the header information of the collected data; for the array collection of fiber optic magnetic sensors, the synchronization signal is additionally transmitted through the optical fiber to avoid electrical signal transmission delay.

[0044] The core of spatial alignment is to convert the relative coordinate system of sonar data and the sensor coordinate system of magnetic data into the WGS84 geographic coordinate system; The raw data from side-scan sonar and subsurface profilers are expressed in distance-angle relative coordinates. Combining the positioning data from the shipborne GNSS and the attitude data from the INS, the relative distance of the sonar is converted into geographic coordinates using the coordinate transformation formula: The lateral distance is converted into east-west offset through the heading angle; The longitudinal distance is converted into north-south offset by the accumulated distance of the GNSS trajectory; The depth data from the shallow subsurface profiler is combined with the ship's draft and tide level data to convert it into absolute depth relative to sea level. Ultimately, each pixel of the side-scan sonar and each profile sampling point of the shallow subsurface profiler are assigned unique coordinates, including: longitude, latitude, and depth.

[0045] The streamer position of the fluxgate magnetometer is calibrated by: Based on the tow cable length, ship heading angle, and underwater depth, the offset coordinates of the sensor relative to the ship's GNSS antenna are calculated, and the real-time geographic coordinates of the ship are superimposed to obtain the absolute coordinates of the magnetometer. The coordinates of the fiber-optic magnetic sensor are directly linked to the underwater robot's positioning system: the robot's onboard GNSS or ultra-short baseline positioning system provides its own geographic coordinates, which are combined with the joint angle sensor data of the robotic arm to calculate the spatial coordinates of each sensor unit.

[0046] This is done because sonar and magnetic data come from different acquisition devices. Temporal and spatial differences can lead to data mismatches, impacting fusion results. Spatiotemporal alignment ensures the temporal and spatial correspondence between the two types of data, providing an accurate data foundation for subsequent feature fusion and correlation analysis, and improving the accuracy of the fusion process.

[0047] After preprocessing and spatiotemporal alignment of the sonar and magnetic data in step S1, noise interference is eliminated and temporal and spatial consistency is ensured, laying a solid foundation for subsequent feature extraction. Next, step S2 extracts feature vectors reflecting pipeline characteristics from the processed sonar and magnetic data.

[0048] like Figure 3 As shown, in step S2, the morphological feature vector of the pipeline is extracted from the sonar data, and the magnetic feature vector of the pipeline is extracted from the magnetic data.

[0049] Among them, the morphological feature vector refers to a set of quantitative parameters extracted from sonar data that can reflect the external geometric shape and spatial distribution characteristics of the pipeline. These parameters can effectively distinguish the pipeline from the surrounding environment and reflect the pipeline's own morphological properties.

[0050] Extract the pipeline's morphological feature vector from the sonar data. Specifically: The pre-processed and spatiotemporally aligned side-scan sonar echo intensity data is binarized, and a threshold is set to separate the pipeline contour area from the background area to obtain a binary contour image of the pipeline.

[0051] Based on the binary contour image, the length characteristics of the pipeline are extracted. That is, the total number of pixels of the continuous contour of the pipeline in the image is multiplied by the spatial resolution of the sonar image (m / pixel). The pipeline length in meters is obtained. This length can reflect the extension range of the pipeline in the waters of near-shore ports and industrial areas.

[0052] The width feature of the pipeline is extracted. The maximum pixel span perpendicular to the pipeline direction in the contour image is calculated and then multiplied by the spatial resolution to obtain the pipeline width. This can be used to distinguish sewage pipelines and cables with different diameters.

[0053] The strike angle feature of the pipeline is extracted, and the straight line segments in the contour image are detected using Hough transform. The angle between the straight line segment and the horizontal direction is calculated. This angle reflects the extension direction of the pipeline in the detection area.

[0054] Extract the continuity features of pipeline contours and count the proportion of continuous pixels in the contour image, the number of continuous pixels / total number of pixels. The closer the value is to 1, the more complete the pipeline contour is and the less obstructed by construction waste. 3 Containing 3-5 concrete blocks, the contour continuity value of the complete pipeline is usually greater than 0.6. If C is less than 0.4, it indicates that it is severely obscured and needs to be verified in combination with magnetic data.

[0055] The length (L), width (W), strike angle (θ), and contour continuity (C) extracted above constitute the pipeline morphological feature vector, that is, the morphological feature vector = .

[0056] Among them, the magnetic characteristic vector refers to a set of quantitative parameters extracted from magnetic data that can reflect the ferromagnetic properties of the pipeline and the abnormal distribution of the magnetic field. It can be used to identify the material characteristics of the pipeline and accurately locate it.

[0057] Extract the pipeline's magnetic feature vector from the magnetic data. Specifically: A sliding window search is performed on the pre-processed and time-space aligned residual magnetic field data with a window size of 2m×2m to identify magnetic anomaly areas with ΔT values ​​> 300nT. These areas correspond to ferromagnetic pipelines, especially cables.

[0058] The magnetic anomaly peak feature is extracted, that is, the maximum value of ΔT in the magnetic anomaly area. The larger the value, the stronger the ferromagnetism of the pipeline, and the more likely it is a cable.

[0059] Extract the magnetic anomaly range characteristics, that is, calculate the area of ​​the magnetic anomaly region, the number of pixels × the square of the spatial resolution. This area is related to the length and diameter of the pipeline, and assists in determining the scale of the pipeline.

[0060] Extract the maximum magnetic field gradient feature. From the magnetic field gradient data collected by the fiber optic magnetic sensor, extract the maximum value of the gradient in the magnetic anomaly area. This value can reflect the magnetic field change rate at the edge of the pipeline and help to accurately locate the pipeline boundary. The symmetry characteristics of the magnetic anomaly are extracted, and the degree of symmetry of the ΔT value in the magnetic anomaly area about the central axis of the area is calculated. The sum of the ΔT differences of the symmetrical pixels / the total number of pixels is calculated. The magnetic anomaly of the cable usually has a high symmetry, while the magnetic anomaly of the sewage pipeline has a low symmetry. The magnetic anomaly peak value (P), magnetic anomaly range (A), magnetic field gradient maximum value (G), and magnetic anomaly symmetry (S) extracted above constitute the magnetic characteristic vector of the pipeline, that is, the magnetic characteristic vector = .

[0061] In this step, by extracting morphological and magnetic eigenvectors, the raw sonar and magnetic data are converted into quantitative parameters with clear physical meaning. This achieves both data dimensionality reduction and the preservation of the pipeline's key characteristics. Morphological eigenvectors describe the pipeline from a geometric perspective, while magnetic eigenvectors reflect its ferromagnetic properties. These provide effective input parameters for subsequent feature fusion and pipeline identification, improving feature distinguishability and model recognition accuracy.

[0062] Step S3: Calculate the matching degree between the sonar profile direction and the magnetic anomaly belt direction based on the pipeline direction, establish an acoustic-magnetic mapping, and fuse the morphological feature vector with the magnetic feature vector to generate a joint feature vector.

[0063] Among them, the pipeline direction refers to the direction in which the pipeline extends in the waters of near-shore ports and industrial areas. It is reflected by the sonar contour direction and the magnetic anomaly belt direction, and is an important link between sonar data and magnetic data.

[0064] The sonar contour direction is the pipeline extension direction extracted from the pipeline binary contour image of the sonar data, and the magnetic anomaly belt direction is the pipeline extension direction extracted from the magnetic anomaly area of ​​the magnetic data.

[0065] The matching degree is a quantitative indicator used to measure the consistency between the sonar profile direction and the magnetic anomaly belt direction. The higher the value, the closer the correlation between the two.

[0066] Acousto-magnetic mapping refers to establishing the correspondence between pipeline features in sonar data and pipeline features in magnetic data, providing a basis for feature fusion.

[0067] The joint eigenvector is a comprehensive eigenvector obtained by fusing the morphological eigenvector and the magnetic eigenvector. It has the characteristics of both and can more comprehensively reflect the properties of the pipeline.

[0068] The matching degree between the sonar profile direction and the magnetic anomaly belt direction is calculated based on the pipeline direction. Specifically: Obtain the strike angle of the pipeline from the morphological feature vector , as a quantitative parameter of the sonar profile direction; It is the angle between the straight line segment extracted from the pipeline binary contour image by Hough transform and the horizontal direction; Perform Hough transform on the magnetic anomaly area in the magnetic data, detect the straight line segment of the magnetic anomaly belt, and calculate its angle with the horizontal direction , as a quantitative parameter of the magnetic anomaly belt trend; Calculate trend matching ,in, is the sonar profile strike angle, is the strike angle of the magnetic anomaly belt, when and When they are completely aligned, M=1, and when they are perpendicular, M=0; the above is rigid alignment.

[0069] In the nearshore port and industrial waters scenario, if M ≥ 0.8, it is preliminarily determined that the sonar profile and the magnetic anomaly zone correspond to the characteristics of the same pipeline. This is because the pipeline direction in this scenario is relatively stable, and a certain measurement error is allowed.

[0070] Furthermore, the optimal path mapping is solved by matching the acoustic centerline with the magnetic anomaly trajectory through dynamic time warping; Extracting the coordinate sequence of the acoustic centerline from sonar data ,in, For the first The plane coordinates of the sampling points reflect the center position of the pipeline in the sonar field of view; Extracting coordinate sequences of magnetic anomaly tracks from magnetic data ,in, For the The plane coordinates of the peak points of the magnetic anomaly correspond to the position where the pipeline has the strongest ferromagnetism.

[0071] Calculate the Euclidean distance between each point pair in the acoustic centerline and the magnetic anomaly trajectory, and construct a distance matrix , where the element ,express and Distance in space.

[0072] The distance matrix quantifies the degree of spatial misalignment of all pairs of points on the two trajectories. Based on the idea of ​​dynamic programming, the optimal path from the upper left corner to the lower right corner of the matrix is ​​found to minimize the cumulative distance of the path.

[0073] In nearshore port and industrial water scenarios, pipelines may be locally bent due to construction deviations or sedimentation. DTW's flexible alignment allows for non-uniform trajectories in time and space, making it more adaptable to complex environments than rigid alignment.

[0074] A one-to-one correspondence between the acoustic centerline sampling points and the magnetic anomaly trajectory sampling points is established based on the optimal path. This mapping can correct the spatial misalignment of the two trajectories and accurately correlate the sonar morphological characteristics with the magnetic characteristics in space.

[0075] The establishment of the acoustic-magnetic mapping was based on the matching degree (M≥0.8) and the optimal path; Based on the calculated strike matching degree M≥0.8, combined with DTW optimal path mapping, we further screened the truly correlated sonar and magnetic features: If the cumulative distance of the optimal path is ≤ 0.5m, the acoustic center line and the magnetic anomaly trajectory are determined to correspond to the same pipeline, and the morphological feature vector With the magnetic eigenvector Map bindings by optimal path; If the cumulative distance is greater than 0.5m, it is necessary to recheck whether it is a false magnetic anomaly caused by construction waste or a pipeline outline misidentified by the sonar to eliminate the interference.

[0076] The morphological feature vector and the magnetic feature vector are fused to generate a joint feature vector. The process is as follows: The morphological and magnetic eigenvectors are normalized, and each characteristic parameter is mapped to the interval [0,1] to eliminate the dimension difference.

[0077] Traditional feature fusion methods, such as simple weighting or splicing, fail to fully exploit the inherent physical connections between acoustic and magnetic data and are susceptible to interference from environmental noise. This method leverages the complementary nature of sonar data, which accurately depicts pipeline geometry, and magnetic data, which effectively reflects pipeline material properties. By employing interactive feature modeling and feature-level deep correlation strategies, this method aims to deeply fuse morphological and geometric information with material physical properties, achieving more accurate and robust feature representation.

[0078] The joint feature vector is formed by dynamically compressing the interactive features and derived features, including: material correlation factor, spatial consistency factor, coupling coefficient, anti-interference factor, scale-invariant feature and scale ratio, and six-dimensional features.

[0079] In terms of feature interaction, a variety of factors and coefficients are constructed for different situations to achieve accurate identification of pipelines; specifically, they include: material correlation factor, spatial consistency factor, and coupling coefficient; The material correlation factor integrates magnetic anomaly intensity, edge sharpness, and pipeline width to form a composite index, revealing the correlation between ferromagnetic materials and tubular structures, and reflecting the difference between cable profile continuity and sewage pipeline profile continuity. The spatial consistency factor quantifies the matching degree of the acoustic and magnetic trajectories through the spatial deviation of the DTW mapping points, and measures the consistency of the spatial distribution of the acoustic and magnetic features. The closer the value is to 1, the higher the possibility that the sonar morphology and the magnetic anomaly come from the same pipeline; For the contradictory acoustic and magnetic characteristics of the coated sewage pipe, that is, the sonar outline is clear but the magnetic signal is weak, a coupling coefficient is constructed to reveal the coupling relationship between the magnetic anomaly range and the acoustic direction deviation, and to suppress the interference of broken or blocked pipelines.

[0080] Specifically, feature interactions include: Sonar alone cannot identify the material of a sewage pipe covered by a coating, and magnetism alone cannot distinguish between rebar from construction waste and real cables. The material correlation factor combines magnetic anomaly intensity (P), edge sharpness (G), and pipeline width (W) to construct a composite index for material identification. Material correlation factor: Among them, the molecule Multiplying the two can highlight the magnetic characteristics and structural integrity of pipelines; cables have high magnetic peaks and high gradients due to metal armor, while steel fragments have scattered magnetic gradients; It is used to weaken the influence of pipe diameter differences on magnetic signals and make the features scale invariant. The nonlinear mapping of the contour continuity reveals the strong correlation between ferromagnetic materials and tubular structures. , the sewage pipeline usually weakens the magnetic signal due to the coating .

[0081] Construction debris can cause sonar profiles to break, and magnetic data may drift due to mooring interference. The spatial consistency factor quantifies the degree of matching between the acoustic and magnetic tracks by using the spatial deviation of the DTW mapping points: Spatial consistency factor: ,in, Adapt constants to the scene; The consistency of the spatial distribution of acoustic and magnetic characteristics is measured by calculating the sum of the spatial deviations between the sonar profile and the magnetic anomaly band using the DTW mapping points; Used for normalization processing to eliminate the influence of the number of mapping points N. The closer this factor value is to 1, the more likely the sonar morphology and magnetic anomaly are from the same pipeline, and the stronger the ability to resist environmental interference such as construction waste.

[0082] The acoustic and magnetic properties of the cladding sewage pipe are contradictory: the sonar outline is clear (high C value) but the magnetic signal is weak (low P value). The coupling coefficient reveals the coupling relationship between the magnetic anomaly range (A) and the acoustic strike deviation (Δθ): Coupling coefficient: ,in, is the difference between the strike angle of the sonar profile and the strike angle of the magnetic anomaly belt, To avoid the minimum value with denominator equal to 0, It reflects the coupling relationship between the magnetic anomaly range and the acoustomagnetic strike deviation. The larger the deviation, the higher the value. Consider the effective length of the pipeline to suppress interference caused by broken or blocked pipelines; the cable has a high consistency in magnetic anomalies and direction due to its uniform metal material.

[0083] The above interactive features and derived features are dynamically compressed. The dynamic compression features include: anti-interference factor, scale-invariant feature and scale ratio, and a 6-dimensional joint feature vector is constructed. ;in, is the anti-interference factor, element product The real pipeline requires the coexistence of magnetic anomalies and acoustic contours. represents element-wise multiplication; The local deformation of the pipeline will cause the pipe width (W) and magnetic gradient (G) to change, but the proportional relationship between the two is stable. The angle between the direction of magnetic field gradient change and the width of the tubular structure is calculated using a four-quadrant inverse tangent function, which eliminates the influence of pipeline size differences such as different pipe diameters and focuses on the proportional relationship between the sharpness of the magnetic anomaly edge and the pipe width; For scale ratio, it distinguishes short-distance strong magnetic anomalies from long-distance linear targets and characterizes the magnetic anomaly intensity per unit length of pipeline.

[0084] In this step, DTW path alignment compensates for the limitations of simple direction matching. Flexible matching adapts to pipeline trajectory deformation in complex nearshore environments, reducing the spatial correlation error of acoustic and magnetic features. Combined with matching screening and optimal path mapping, false associations with interfering targets such as construction waste are further eliminated, providing a more reliable basis for feature fusion. Ultimately, the joint feature vector possesses both morphological integrity and material specificity, providing more comprehensive feature support for pipeline identification.

[0085] In step S4, the pipeline type and pipeline status are determined based on the joint feature vector, wherein the pipeline type is a sewage pipeline and a cable.

[0086] Among them, pipeline type determination refers to distinguishing whether the detected pipeline is a sewage pipeline or a cable. These two types of pipelines have significant differences in material, function, and detection requirements: Sewage pipelines are mostly made of non-metallic materials, such as concrete and plastic, and are mainly used to transport sewage. They may have a biological coating on the outside; cables are mostly made of metal materials, such as copper core with steel armor sheath, and are used to transmit electricity or signals. They have strong ferromagnetism.

[0087] The pipeline status refers to the integrity of the pipeline. For sewage pipelines, the main focus is on the integrity of its external covering layer; for cables, the main focus is on the corrosion of its metal sheath and the continuity of the overall structure.

[0088] Determine the pipeline type based on the joint feature vector, specifically: Different types of pipelines, such as sewage pipelines and cables, have different morphological and magnetic features. By combining the features in the feature vector, the pipeline types can be distinguished.

[0089] In nearshore ports and industrial areas, the strong ferromagnetic properties of cable metal armor often result in higher magnetic anomaly peaks and a symmetrical distribution. Sewage pipelines, on the other hand, are often made of non-metallic materials, resulting in lower magnetic anomaly peaks and less symmetric distribution. Based on these characteristics, setting thresholds allows for differentiation between pipeline types. Cable identification: When and When it is determined to be a cable; or 、 and When , it is determined to be a cable; Sewage line identification: Sewage pipe with coating: when meeting 、 and When it is detected, it is determined to be a sewage pipe with a coating layer; standard sewage pipe: when it does not meet the cable identification conditions and does not meet the characteristics of a sewage pipe with a coating layer, it is determined to be a standard sewage pipe.

[0090] The pipeline status is determined based on the joint feature vector. Specifically: The integrity of a pipeline is an important indicator for evaluating whether it can operate normally. By combining the contour continuity and magnetic anomalies of the pipeline, the integrity of the pipeline can be comprehensively judged.

[0091] Integrity Assessment Model ,in, It is the benchmark magnetic anomaly value for similar pipelines; Furthermore, for pipelines identified as cables, the focus is on the maximum magnetic field gradient and contour continuity. The maximum magnetic field gradient reflects the edge changes of the cable's metal sheath. If the maximum magnetic field gradient shows a sudden change, it indicates that the sheath may be corroded or damaged at that location. Contour continuity < 0.8 indicates that the cable may have a local break or be severely obscured by construction waste, affecting its normal function. For pipelines identified as sewage pipelines, focus on changes in contour continuity and width characteristics: if contour continuity is less than 0.7, it indicates that the external covering layer of the sewage pipeline may be damaged or detached; if the width characteristic increases significantly in a certain section, it may be due to local expansion of the covering layer or deformation of the pipeline.

[0092] Based on the above characteristics, the pipeline status is divided into levels: Condition Level 1 (Intact): All characteristics are within normal range; Status Level 2 (Minor Abnormality): There are slight deviations in individual characteristics, but they do not affect overall function; Condition Level 3 (Moderately Abnormal): Multiple features are deviated, and there may be local damage; Status Level 4 (Severe Abnormality): Characteristics are severely deviated, posing a major safety hazard and requiring immediate repair.

[0093] The pipeline type and status are judged based on the joint feature vector, fully utilizing the complementarity of morphological and magnetic features: the material type is accurately distinguished through magnetic features, avoiding misjudgment caused by interference from the coating layer in single sonar data; through the comprehensive analysis of morphological and magnetic features, the pipeline status is comprehensively assessed, overcoming the limitation that single magnetic data cannot reflect changes in the external morphology of the pipeline.

[0094] The above description is based on the ideal embodiment of the present invention. Based on the above description, relevant personnel can make various changes and modifications without departing from the technical scope of this invention. The technical scope of this invention is not limited to the content of the specification and must be determined according to the scope of the claims.

Claims

1. A method for underwater acoustic and magnetic multimodal fusion processing, characterized in that: The following steps are involved: S1. Preprocessing underwater pipeline sonar data and magnetic data; S2, extracting the pipeline's morphological feature vector from the sonar data, and extracting the pipeline's magnetic feature vector from the magnetic data; S3. Based on the pipeline route, the matching degree between the sonar profile and the magnetic anomaly zone is calculated. The optimal path is calculated through dynamic time warping to establish an acoustic-magnetic mapping. The morphological feature vector and the magnetic feature vector are fused to generate a joint feature vector; The step of calculating the matching degree in step S3 includes: Extract pipeline strike angle from sonar data as sonar profile strike, and extract magnetic anomaly strike angle from magnetic data; Calculate the angle difference between two strike angles ,in, is the sonar profile strike angle, is the strike angle of the magnetic anomaly belt, when and When they are completely consistent, M=1; when they are perpendicular, M=0; when the matching degree corresponding to the difference value is ≥0.8, they are determined to be the same pipeline feature; S4. Determine the pipeline type and pipeline status based on the joint feature vector; wherein the pipeline type is a sewage pipeline and a cable.

2. The underwater acoustic and magnetic multimodal fusion processing method according to claim 1, characterized in that: The pre-processing of sonar data in step S1 includes: The non-local mean filtering algorithm is used to suppress the strong reflection clutter generated by construction waste by calculating the neighborhood pixel similarity weights, and retain the continuous contour signal of the pipeline.

3. The underwater acoustic and magnetic multimodal fusion processing method according to claim 1, characterized in that: The pre-processing of magnetic data in step S1 includes: The geomagnetic model is used to calculate the theoretical geomagnetic field strength, and this value is subtracted from the original magnetic data to obtain the residual magnetic field data; The residual magnetic field data were filtered with a sliding window to remove the high-frequency magnetic field fluctuations caused by the anchoring of the ship.

4. The underwater acoustic and magnetic multimodal fusion processing method according to claim 1, characterized in that: The sonar data includes: side scan sonar echo intensity data and shallow layer profiler reflection coefficient data; the magnetic data includes: magnetic field total intensity data and magnetic field gradient data; Preprocessing also includes time and space alignment. The specific steps are: achieving time synchronization between the sonar device and the magnetic device through hardware trigger pulses, combining GNSS positioning data and INS attitude data to convert the range-angle coordinates of the side-scan sonar into the WGS84 geographic coordinate system, and unifying the magnetic sensor coordinates into the same geographic coordinate system through cable position calibration.

5. The underwater acoustic and magnetic multimodal fusion processing method according to claim 1, characterized in that: In step S2, the morphological feature vector includes pipeline length, pipeline width, pipeline strike angle, and contour continuity; the magnetic feature vector includes magnetic anomaly peak, magnetic anomaly range, magnetic field gradient maximum, and magnetic anomaly symmetry.

6. The underwater acoustic and magnetic multimodal fusion processing method according to claim 1, characterized in that: The specific steps of dynamic time warping in step S3 include: Extract the acoustic centerline coordinate sequence and the magnetic anomaly trajectory coordinate sequence, and construct the spatial distance matrix between the two sequences; Solve the optimal path with the minimum cumulative distance through dynamic programming and establish a point-to-point mapping relationship; When the cumulative distance of the optimal path is ≤ 0.5 m, the morphological eigenvector and the magnetic eigenvector are bound.

7. The underwater acoustic and magnetic multimodal fusion processing method according to claim 1, characterized in that: The specific steps of feature fusion in step S3 include: Normalize the morphological and magnetic eigenvectors and map each characteristic parameter to the [0,1] interval; Based on the acoustic-magnetic mapping relationship, three types of interaction features, material correlation factor, spatial consistency factor and coupling coefficient, are calculated; Combined with dynamic compression, a 6-dimensional joint feature vector containing anti-interference factors, scale-invariant features and scale ratio is constructed.

8. The underwater acoustic and magnetic multimodal fusion processing method according to claim 7, characterized in that: The material correlation factor is constructed by integrating the magnetic anomaly peak, the maximum magnetic field gradient and the pipeline width; the spatial consistency factor is calculated based on the spatial deviation of the dynamic time warping mapping point; and the coupling coefficient reflects the coupling relationship between the magnetic anomaly range and the acoustomagnetic strike deviation.

9. The underwater acoustic and magnetic multimodal fusion processing method according to claim 1, characterized in that: The specific steps of determining the pipeline type in step S4 include: When the magnetic anomaly peak value is greater than 500nT and the magnetic anomaly symmetry is greater than 0.7, it is determined to be a cable; When the magnetic anomaly peak value is less than 200nT, the contour continuity is greater than 0.8, and the coupling coefficient is less than 0.3, it is determined to be a sewage pipeline containing a coating layer; The rest of the cases are judged as standard sewage pipelines.

Images