A side scan sonar image position correction and positioning method

Abstract

This invention relates to a method for position correction and localization of side-scan sonar images, comprising: acquiring raw side-scan sonar data from first and second survey lines; performing seabed line tracking, radiation distortion correction, and slant range correction on the data; performing preliminary geocoding and overlaying to generate a preliminary mosaic map; measuring the lateral displacement difference and estimating the average slant range; calculating the initial heading offset; re-geocoding; verifying and locking the final heading offset through several common feature points; and generating a corrected geocoded mosaic map and underwater target positioning coordinates. The side-scan sonar image position correction and localization method designed using this invention can solve the problems of low efficiency, difficulty in stably ensuring simultaneous alignment of target features at different positions, and insufficient reliability of underwater target positioning coordinates in complex hydrological environments where there are differences between the actual motion state of the towed body and the recorded state by the survey vessel.

Description

Technical Field

[0001] This invention relates to the fields of geocoding, image position correction, and underwater target localization of side-scan sonar images in marine surveying and underwater target detection, specifically to a method for side-scan sonar image position correction and localization. Background Technology

[0002] Side-scan sonar is commonly used in marine engineering construction, underwater structure detection, seabed topography surveys, and underwater target identification. It generates sonar images by continuously collecting seabed echo data along a survey line, and then geocodes these images using shipborne positioning data, attitude data, and towed body position estimation data, thus establishing a correspondence between the sonar imagery and actual planar coordinates. In underwater structure positioning operations, to improve coverage integrity and reliability, adjacent or parallel survey lines are typically deployed for the same target area, and image results from different survey lines are cross-checked. This processing method places high demands on the accuracy of navigation data, attitude data, towed body position estimation, and image stitching consistency.

[0003] In complex hydrological environments, side-scan sonar towed bodies are easily affected by water currents, tides, changes in tow cable morphology, and vessel attitude disturbances, leading to discrepancies between the actual motion of the towed body and the state recorded by the survey vessel. Existing processing methods typically rely on the raw navigation attitude data for geocoding or on manual experience to adjust for image misalignment in post-processing. When the same underwater target exhibits systematic misalignment in images from different survey lines, manual processing is not only inefficient, but the correction scale, correction direction, and convergence criteria are also easily influenced by the operator's experience, making it difficult to reliably ensure the simultaneous alignment of target features at different locations. This, in turn, affects the reliability of underwater target positioning coordinates and the consistency of engineering results.

[0004] Therefore, existing technologies have shortcomings and need to be improved and developed. Summary of the Invention

[0005] The present invention provides a side-scan sonar image position correction and positioning method to solve the problem that in complex hydrological environments, manual processing of the actual motion state of the towed body and the recorded state of the survey vessel is not only inefficient, but also the correction scale, correction direction and convergence standard are easily affected by the operator's experience, making it difficult to stably ensure that the target features at different positions are aligned at the same time, thus affecting the reliability of underwater target positioning coordinates and the consistency of engineering results.

[0006] This invention provides a method for side-scan sonar image position correction and localization, including:

[0007] Acquire raw side-scan sonar data for the first and second survey lines, which are two survey lines formed by side-scan sonar collecting data on the same underwater target along different tracks.

[0008] The side-scan sonar data of the first and second survey lines are subjected to seabed line tracking, radiation distortion correction and slant range correction.

[0009] Preliminary geocoding was performed using the original navigation data and attitude data, and then the data was overlaid to generate a preliminary mosaic map.

[0010] In the preliminary mosaic map, common feature points of the same underwater target are identified, and the lateral displacement difference of the common feature points in the two survey line images is measured and the average slope distance is estimated. The lateral displacement difference is the displacement difference measured along the lateral direction of the survey line in the geocoded mosaic map coordinate system. The average slope is the average slope distance obtained by averaging the slope distances of the common feature points corresponding to the first survey line and the second survey line, respectively.

[0011] The initial heading offset is obtained based on the lateral displacement difference and the average slant distance.

[0012] Apply a reverse heading offset to at least one of the first survey line and the second survey line and regeocode it;

[0013] The alignment results are verified by several common feature points. When the remaining lateral displacement difference of several common feature points is less than the preset threshold, or when the root mean square error of the total offset vector of several common feature points reaches the minimum, the final heading offset is locked, and the corrected geocoded mosaic map and underwater target positioning coordinates are generated.

[0014] Furthermore, the common feature points originate from the port side sonar image data of the first survey line and the starboard side sonar image data of the second survey line when corresponding to the same underwater target, or from the starboard side sonar image data of the first survey line and the port side sonar image data of the second survey line when corresponding to the same underwater target.

[0015] The seabed line tracking is used to determine the seabed echo boundary, the radiation distortion correction is used to correct the grayscale difference of the sonar image, and the slant range correction is used to convert the sonar slant range observation into a planar distance observation.

[0016] The obtained side-scan sonar image data, corrected for seabed line, grayscale, and slant range, is used to identify common feature points of the same underwater target.

[0017] Furthermore, the preliminary geocoding includes determining the reference location of the side-scan sonar observation data based on the tow point coordinates, wherein the tow point coordinates are used as the reference point for subsequent tow body coordinate estimation or geocoding of the sonar observation data, and the tow point coordinates are obtained according to the following formula:

[0018] ;

[0019] In the formula, For geographic coordinate system, For the ship's hull, a fixed coordinate system is established. For drag point object identification, For shipborne satellite positioning antennas, These are the planar coordinates of the drag point in the geographic coordinate system. The coordinates of the shipborne satellite positioning antenna in the geographic coordinate system are given. Let be the planar coordinates of the shipborne satellite positioning antenna in the ship's fixed coordinate system. Let be the planar coordinates of the towing point in the fixed coordinate system of the ship's hull. This is the rotation matrix corresponding to the heading angle. This is the rotation matrix corresponding to the pitch angle. This is the rotation matrix corresponding to the roll angle. For heading angle, For pitch angle, The roll angle, and These are the two orthogonal coordinate components of the planar coordinate system.

[0020] Furthermore, in the initial geocoding, under the condition that the tow body's orientation is approximated by the survey vessel's heading, where this condition corresponds to the tow cable's horizontal projection direction being approximated by the survey vessel's heading angle, the tow body coordinates are obtained according to the following formula:

[0021] ;

[0022] In the formula, For towing sonar tow body. The coordinates of the towed sonar towed body in the geographic coordinate system are given. These are the planar coordinates of the drag point in the geographic coordinate system. This refers to the length of the tow cable. The angle of inclination of the tow cable relative to the vertical direction. Relative to geographic coordinate system The heading angle of the surveying vessel measured along the axial direction. and These are the two orthogonal coordinate components of the planar coordinate system.

[0023] Furthermore, the lateral position error caused by the heading angle error is characterized by the following formula:

[0024] In the formula, The positional error is in the direction perpendicular to the survey line's heading. This refers to the slant range of a single survey line in a side-scan sonar. This refers to the heading angle error;

[0025] When the system-recorded heading deviates from the actual heading, the heading compensation amount applied in post-processing is determined according to the following formula:

[0026] In the formula, This is the heading compensation amount applied during post-processing. The negative sign indicates that the direction of the heading compensation is opposite to the direction of the deviation, so that the system records the deviation of the heading relative to the actual heading.

[0027] Furthermore, the common feature points are corresponding image features that can be identified in both the first and second survey line images of the same underwater target;

[0028] The common feature points include feature points located at the far end of the underwater target, where the far end is the region closer to the edge of the side-scan sonar range relative to the near end region of the corresponding survey line; the initial heading offset is estimated according to the following formula:

[0029] ;

[0030] In the formula, This is the initial heading offset. The difference in lateral displacement of the same common feature point in the first and second survey line images. The average slant distance of this common feature point relative to the position of the surveying vessel. It is an arcsine function, a constant. This represents the bilateral misalignment contribution term formed by the first and second survey lines for the same common feature point.

[0031] Furthermore, applying a reverse heading offset to at least one of the first survey line and the second survey line includes:

[0032] The initial heading offset is used as the first heading correction parameter;

[0033] When the misalignment of one of the first or second survey lines is greater than that of the other survey line, a reverse heading offset is applied to the survey line with the greater misalignment; wherein, the degree of misalignment is determined based on the remaining lateral displacement difference of common feature points, the root mean square error of the total offset vector, or the deviation from the coordinates of the check point;

[0034] When both the first survey line and the second survey line have systematic offsets related to the corresponding navigation attitude data, a heading offset amount opposite to the direction of their respective heading errors is applied to the first survey line and the second survey line respectively.

[0035] Furthermore, re-geocoding includes:

[0036] After each heading offset is applied, the geocoding results of the first and second survey lines are regenerated based on the original navigation data, attitude data, and the applied heading offset, and the mosaic map is updated; wherein, the input parameters for each heading offset include the original navigation data, attitude data, the heading offset of the first survey line, and the heading offset of the second survey line.

[0037] The remaining lateral displacement difference of common feature points is remeasured in the updated mosaic.

[0038] The heading offset is adjusted based on the remaining lateral displacement difference and geocoding is repeated. The heading offset corresponding to the smallest remaining lateral displacement difference is determined as the candidate final heading offset. When there are several common feature points, the heading offset corresponding to the smallest sum of squares of the remaining lateral displacement differences of the several common feature points or the smallest root mean square error of the total offset vector is determined as the candidate final heading offset.

[0039] Furthermore, the alignment results are verified through multiple common feature points, including:

[0040] Under the influence of the candidate final heading offset, check the alignment of multiple common feature points at different positions in the updated mosaic map;

[0041] When several common feature points distributed in different parts of an underwater target or different regions of a mosaic form corresponding alignment results with the candidate final heading offset, the candidate final heading offset is determined to have global validity and is locked as the final heading offset; where alignment means that the remaining lateral displacement difference is less than a preset threshold, or the root mean square error of the total offset vector is less than a preset threshold.

[0042] When at least one common feature point does not form a corresponding alignment result, continue to adjust the heading offset to avoid using the local alignment result as the final result.

[0043] Furthermore, the generation of the corrected geocoded mosaic map and underwater target location coordinates includes:

[0044] The first and second survey lines are re-geocoded using the final heading offset, and the final heading offset, the corrected geocoded mosaic map, the underwater target plane position coordinates, and the accuracy assessment results are output.

[0045] Among them, the planar position coordinates of the underwater target are the geographic coordinates corresponding to the geometric center of the underwater target, the preset positioning point, or the confirmed common feature points;

[0046] The accuracy assessment results include the deviation before and after the checkpoint correction, the eastward offset of the feature point, the northward offset of the feature point, and the root mean square error of the total offset vector.

[0047] The underwater target is any one of the following: a large cylindrical barrel, a pile foundation, a pipeline, a shipwreck, or an artificial reef, and has an identifiable geometric shape.

[0048] Beneficial effects:

[0049] As can be seen from the above technical solutions, the present invention provides a method for side-scan sonar image position correction and localization, which has the following beneficial effects:

[0050] 1. It can transform the problem of survey line misalignment into a calculable heading offset correction problem.

[0051] This application does not rely solely on manual image dragging or experience-based judgment for image registration. Instead, it first establishes the misalignment relationship between survey lines using common feature points of the same underwater target in the first and second survey lines. Then, it establishes a geometric relationship between the lateral displacement difference of the common feature points and the average slant range, thereby estimating the initial heading offset. Since the impact of heading error on the lateral position of distant targets on side-scan sonar is related to the slant range, quantifying the misalignment through common feature points at the distant targets can better reflect the direction and magnitude of the heading error. This processing logic transforms the phenomenon of image stitching misalignment, which was originally manifested as misalignment, into a heading compensation parameter that can be input into the geocoding process. This allows subsequent correction to move beyond empirical movement at the image level and return to the spatial positioning stage of sonar observation data for correction, which is beneficial for improving the repeatability and engineering feasibility of the processing.

[0052] 2. It can improve the spatial consistency of the same target in geocoding mosaic maps.

[0053] Misalignment in side-scan sonar images is often not a uniform translation, but a systematic spatial deviation influenced by the heading angle, towed body position, and slant range. While ordinary image stitching or local translation might align target features at one location, other areas remain misaligned. This application re-geocodes the original navigation data, attitude data, and heading offset, incorporating correction into the coordinate calculation process of each sonar sampling point. This ensures that image coordinate updates match spatial attitude changes along the survey line. Further verification of the final heading offset using multiple common feature points reduces misjudgments caused by accidental alignment of local features, resulting in a more consistent spatial representation of the corrected mosaic in terms of target boundaries, geometric contours, and strip edges. This improves the interpretability of underwater structure imagery.

[0054] 3. It can reduce the impact of complex hydrological environments on the positioning results of towed sonar.

[0055] The positioning accuracy of towed side-scan sonar is affected by the tow cable length, tow cable inclination angle, hull attitude, heading angle, and water current. In strong currents or tidal conditions, the assumption of approximating the tow body's bearing based on the survey vessel's heading can easily introduce systematic biases, leading to discrepancies between the calculated tow body position and the actual position. This application, after initial geocoding, does not directly accept the mosaic results generated from the original navigation attitude data. Instead, it uses the spatial misalignment of common feature points to infer the heading offset and applies reverse compensation during post-processing. This method can correct image misalignment caused by heading error propagation without requiring re-acquisition of data, thereby reducing the impact of tow body attitude instability, tow cable drift, and complex water currents on the final positioning results. This allows side-scan sonar data to still produce positioning results that meet engineering application requirements even under adverse hydrological conditions.

[0056] 4. It can form a closed-loop processing flow from initial estimation to iterative verification.

[0057] The processing flow of this application does not simply provide a one-time parameter correction. Instead, after estimating the initial heading offset, it forms a closed loop through steps such as applying a reverse heading offset, re-geocoding, measuring the remaining lateral displacement difference, adjusting the heading offset, and verifying with multiple feature points. This closed-loop process provides feedback for parameter correction, allowing for further evaluation of the adequacy of the correction based on the updated mosaic. When multiple common feature points meet the alignment conditions, the candidate heading offset is then locked as the final heading offset, reducing the impact of estimation errors, single-point recognition errors, and local image texture interference on the final result. Thus, the entire processing flow retains the guiding role of the geometric model for the initial parameters while improving the robustness of the final result through iterative verification.

[0058] 5. It can simultaneously output image results, positioning results, and accuracy evaluation results.

[0059] After correcting the heading offset, this application not only generates a corrected geocoded mosaic map but also outputs the underwater target's planar position coordinates and accuracy assessment results. The image results are used for target identification and engineering interpretation, the positioning coordinates are used for construction layout, structure verification, and subsequent spatial analysis, and the accuracy assessment results are used to evaluate the deviations before and after correction, the eastward offset, the northward offset, and the root mean square error of the total offset vector. Because the output covers three levels—"image alignment," "coordinate reliability," and "error reduction"—it can generate verifiable and traceable engineering data processing results, facilitating the transformation of side-scan sonar image correction from simple image processing into a complete technical process for precise underwater target positioning.

[0060] It should be understood that all combinations of the foregoing concepts and the additional concepts described in more detail below can be considered part of the inventive subject matter of this disclosure, provided that such concepts do not contradict each other.

[0061] The foregoing and other aspects, embodiments, and features of the teachings of the present invention will be more fully understood from the following description in conjunction with the accompanying drawings. Other additional aspects of the invention, such as features and / or beneficial effects of exemplary embodiments, will become apparent from the following description or may be learned through practice of specific embodiments according to the teachings of the present invention. Attached Figure Description

[0062] The accompanying drawings are not drawn to scale. In the drawings, each identical or nearly identical component shown in the various figures may be denoted by the same reference numeral. For clarity, not every component is labeled in each figure. Embodiments of various aspects of the invention will now be described by way of example and with reference to the accompanying drawings, wherein:

[0063] Figure 1 This is a flowchart illustrating a side-scan sonar image position correction and localization method according to an embodiment of this application.

[0064] Figure 2 This is a comparison table of absolute positioning accuracy using a side-scan sonar image position correction and positioning method in the embodiments of this application.

[0065] Figure 3 This is an image matching accuracy table for a side-scan sonar image position correction and positioning method used in the embodiments of this application.

[0066] Figure 4 This is a schematic diagram of an electronic device according to an embodiment of this application. Detailed Implementation

[0067] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the described embodiments of the present invention without creative effort are within the scope of protection of the present invention. Unless otherwise defined, the technical or scientific terms used herein should have the ordinary meaning understood by those skilled in the art to which this invention pertains.

[0068] The terms "first," "second," and similar words used in the specification and claims of this patent application do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Similarly, unless the context clearly indicates otherwise, the singular forms of "an," "a," or "the," etc., do not indicate a quantity limitation, but rather indicate the presence of at least one. Terms such as "comprising" or "including" mean that the element or object preceding "comprising" encompasses the features, integrals, steps, operations, elements, and / or components listed following "comprising" or "including," and do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components, and / or collections thereof. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; these relative positional relationships may change accordingly when the absolute position of the described object changes.

[0069] In marine engineering construction, side-scan sonar technology is a crucial tool for seabed detection and target identification. However, in complex hydrological environments such as Yangshan Deep-Water Port, images acquired by side-scan sonar often suffer from severe positional misalignment and geometric distortion due to multiple factors including water flow, tides, and towed vehicle attitude. This results in significant discrepancies between the actual location of structures like large underwater cylinders and their displayed location in the images, severely impacting the construction quality and efficiency of breakwater laying projects. Traditional manual correction methods suffer from low processing efficiency, high subjectivity, and difficulty in guaranteeing accuracy.

[0070] When conducting high-precision seabed surveys using side-scan sonar, especially for structures with well-defined geometric shapes such as pile foundations, pipelines, and shipwrecks, parallel "back-and-forth" survey lines are standard operating procedures to achieve full coverage and cross-verify data. However, in the later data processing, after geocoding the sonar images, it is often observed that the images of the same structure on the two survey lines (i.e., the port side data of one survey line and the starboard side data of another survey line) cannot be precisely matched spatially, exhibiting systematic misalignments of several meters or even tens of meters.

[0071] This misalignment is not random noise, but a systematic error in the sensor integration system. Its core root cause can be attributed to the high coupling and error propagation of the three major systems of positioning, attitude determination, and timing. Among them, the error of the heading angle is the most sensitive and common factor causing this type of misalignment.

[0072] Side-scan sonar systems operate via towed cables, and their precise positioning is a prerequisite for accurate image geocoding. Towed body positioning aims to determine the dynamic position of the towed equipment (such as a side-scan sonar fish tower) to provide precise coordinates for its observation data. However, in the complex hydrological environment of Yangshan Port, positioning faces significant challenges, and the methods employed have limitations:

[0073] The estimation method calculates the position based on the towline length, inclination angle, and ship's attitude. This method is low-cost, but its core assumption—that the towed body's bearing coincides with the surveying ship's bearing—is highly invalid under the strong crosscurrents of Yangshan Port. The bending of the towline and the significant drift of the towed body introduce significant systematic errors between the estimated and actual positions.

[0074] Therefore, embodiments of the present invention provide a method for side-scan sonar image position correction and localization, referring to... Figure 1 ,include:

[0075] Step S102: Obtain the raw side-scan sonar data of the first and second survey lines. The first and second survey lines are two survey lines formed by the side-scan sonar collecting data on the same underwater target along different tracks.

[0076] Step S104: Perform seabed line tracking, radiation distortion correction, and slant range correction on the side-scan sonar data of the first and second survey lines.

[0077] Step S106: Perform preliminary geocoding using the original navigation data and attitude data respectively, and overlay them to generate a preliminary mosaic map.

[0078] Step S108: Identify common feature points of the same underwater target in the preliminary mosaic map, measure the lateral displacement difference of the common feature points in the two survey line images and estimate the average slope distance; wherein, the lateral displacement difference is the displacement difference measured along the lateral direction of the survey line in the geocoded mosaic map coordinate system; the average slope is the average slope distance obtained by averaging the slope distances of the common feature points corresponding to the first survey line and the second survey line respectively.

[0079] Step S110: Obtain the initial heading offset based on the lateral displacement difference and the average slant distance.

[0080] Step S112: Apply a reverse heading offset to at least one of the first survey line and the second survey line and re-geocode it.

[0081] Step S114: Verify the alignment result using several common feature points. When the remaining lateral displacement difference of several common feature points is less than a preset threshold, or when the root mean square error of the total offset vector of several common feature points reaches its minimum, lock the final heading offset and generate the corrected geocoded mosaic map and underwater target positioning coordinates.

[0082] First, the raw side-scan sonar data of the first and second survey lines are acquired, and the basic quality of the image data is improved through seabed line tracking, radiation distortion correction, and slant range correction. Then, preliminary geocoding is performed using the raw navigation data and attitude data, so that the images of the two survey lines are superimposed under the same spatial reference. Common feature points of the same underwater target are identified in the preliminary mosaic, and the initial heading offset is deduced based on the lateral displacement difference and average slant range of the common feature points. Then, a reverse heading offset is applied to at least one survey line and re-geocoded. Finally, the alignment results are checked by multiple common feature points to determine the final heading offset, and the corrected geocoded mosaic and the underwater target positioning coordinates are output.

[0083] This method uses common feature points of the same underwater target in images from two survey lines as the basis for retrieving heading errors. It establishes a basis for estimating heading offset using lateral displacement difference and average slant range, and then re-enters the geocoding process through reverse heading compensation. This approach differs from simply manually translating or locally registering mosaics at the image level. Its core lies in deriving geocoding parameter deviations from image misalignment phenomena and feeding these deviations back to the spatial coding stage of the original survey lines. Based on this, the heading deviations causing misalignment can be inferred from the image differences of the same target in two survey lines, providing a clear geometric basis for the correction process. Since side-scan sonar geocoding results are sensitive to heading angle errors, especially in distant areas with large slant ranges where lateral misalignment is more likely, measuring lateral displacement difference through common feature points can quickly determine the heading compensation direction and initial value. Recoding and multi-point verification avoid local alignment problems caused by single manual adjustments, ensuring good consistency in spatial location, target outline, and positioning coordinates in the final mosaic, making it suitable for underwater structure positioning and marine engineering mapping output.

[0084] In some embodiments, the common feature points are derived from the port side sonar image data of the first survey line and the starboard side sonar image data of the second survey line when corresponding to the same underwater target, or from the starboard side sonar image data of the first survey line and the port side sonar image data of the second survey line when corresponding to the same underwater target.

[0085] Sea line tracking is used to determine the seabed echo boundary, radiation distortion correction is used to correct grayscale differences in sonar images, and slant range correction is used to convert sonar slant range observations into planar distance observations.

[0086] The obtained side-scan sonar image data, corrected for seabed line, grayscale, and slant range, is used to identify common feature points of the same underwater target.

[0087] For the same underwater target, the port side image of the first survey line and the starboard side image of the second survey line, or the starboard side image of the first survey line and the port side image of the second survey line, are used to find the same image features to ensure that the common feature points come from the relative observation results of the same target area. At the same time, the seabed echo boundary is determined by seabed line tracking, the gray-scale difference of sonar images is reduced by radiation distortion correction, and the slant range observation is converted into a planar distance observation by slant range correction, so that the target boundary and geometric features are more suitable for identification in the image.

[0088] Instead of generalizing common feature points to arbitrary image features, this approach limits them to the same underwater target feature in the side-scan sonar images of two survey lines relative to each other. Side-scan sonar images traveling back and forth, or port and starboard images of adjacent survey lines, differ in their observation directions. Directly comparing arbitrary image regions can easily introduce non-identical features. By limiting common feature points to corresponding side-scan images and performing seabed line, radiation, and slant range processing beforehand, interference from image quality differences, slant range geometric differences, and inconsistencies in seabed echo boundaries on feature recognition can be reduced.

[0089] The inputs for seabed line tracking are: the original side-scan sonar waterfall image, single-ping echo grayscale data, grayscale difference threshold, minimum water depth, smoothing window size, and towed depth; the output is: the pixel position sequence of each row / ping seabed line; the target is: the original waterfall image of a single strip of side-scan sonar. The inputs for radiation distortion correction are: towed track height data obtained from seabed line tracking, the original waterfall image before grayscale equalization, slant range-backscattering intensity (BS) sequence, beam incidence angle calculation parameters, k-means classification number, and z-score normalization parameters; the outputs are: incidence angle-BS sequence, unsupervised seabed classification image, average incidence angle-BS curves for various seabed types, and sonar waterfall image after grayscale equalization of horizontal and vertical tracks; the target is: side-scan sonar images with radiation distortion (uneven grayscale at near and far distances, grayscale deviation along the track). The inputs for slant range correction are: the sonar waterfall image after grayscale equalization, the seabed height of the towed fish provided by the seabed line, the slant range pixel width of a single Ping echo, the distance conversion geometric parameters, and the basic assumptions of sonar imaging; the outputs are: the corrected pixel coordinates of slant range converted to horizontal distance, the corrected sonar image with geometric distortion eliminated, no water column separation, and seamless stitching between the two sides; the target is: the side-scan sonar strip image that still has lateral tilt geometric distortion and water column area fragmentation after grayscale equalization; it corrects the positional distortion caused by slant range recording and eliminates the water column blank area.

[0090] The above design improves the reliability of common feature point identification. Since common feature points are fundamental for subsequent calculations of initial heading offsets and verification of alignment results, incorrect selection of these points can lead to deviations in the heading compensation direction or magnitude. By first performing seabed line tracking, radiation distortion correction, and slant range correction, the underwater target contours in the image are made more stable. Furthermore, by determining the features of corresponding targets relative to the side-view image, common feature points can be made closer to their actual physical locations in real space, thereby improving the accuracy of heading offset estimation and subsequent recoding correction.

[0091] In some embodiments, preliminary geocoding includes determining the reference location of the side-scan sonar observation data based on the tow point coordinates, wherein the tow point coordinates are used as the reference point for subsequent tow body coordinate extrapolation or geocoding of the sonar observation data, and the tow point coordinates are obtained according to the following formula:

[0092] .

[0093] In the formula, For geographic coordinate system, For the ship's hull, a fixed coordinate system is established. For drag point object identification, For shipborne satellite positioning antennas, These are the planar coordinates of the drag point in the geographic coordinate system. The coordinates of the shipborne satellite positioning antenna in the geographic coordinate system are given. Let be the planar coordinates of the shipborne satellite positioning antenna in the ship's fixed coordinate system. Let be the planar coordinates of the towing point in the fixed coordinate system of the ship's hull. This is the rotation matrix corresponding to the heading angle. This is the rotation matrix corresponding to the pitch angle. This is the rotation matrix corresponding to the roll angle. For heading angle, For pitch angle, The roll angle, and These are the two orthogonal coordinate components of the planar coordinate system.

[0094] By using the planar coordinates of the shipborne satellite positioning antenna in the geographic coordinate system, and the relative positions of the shipborne satellite positioning antenna and the tow point in the ship's fixed coordinate system, combined with the rotation matrices corresponding to the heading angle, pitch angle, and roll angle, the planar coordinates of the tow point in the geographic coordinate system are calculated. These tow point coordinates serve as the basic position in the towed sonar positioning chain and can be used for subsequent tow body position estimation and geocoding of side-scan sonar observation data, thereby establishing a spatial correspondence between ship positioning data, attitude data, and sonar data.

[0095] By linking the issue of misalignment in side-scan sonar images with the tow point repositioning calculation in the towed sonar positioning chain, this study clarifies that preliminary geocoding does not solely rely on shipborne positioning points, but rather determines the tow point's reference position through attitude transformation between the ship's fixed coordinate system and the geographic coordinate system. Since heading angle, pitch angle, and roll angle affect the spatial repositioning result of the tow point relative to the shipborne satellite positioning antenna, the tow point coordinate formula provides a parameter basis for subsequent analysis of heading angle error propagation, improving the accuracy and traceability of the preliminary geocoding reference position. The spatial position of towed sonar observation data is not the shipborne antenna position, but is related to the tow point, tow body, and tow cable status. By first calculating the tow point coordinates from the shipborne antenna position and the installation geometry in the ship's fixed coordinate system, the systematic errors caused by directly using the ship's position as a substitute for the sonar towed system reference point can be reduced. This calculation model also clearly demonstrates the influence of heading angle and attitude angle on the geocoding input position, providing an interpretable error propagation basis for subsequent reverse correction of heading offset based on image misalignment.

[0096] In some embodiments, during initial geocoding, under the condition that the tow body's orientation is approximated by the survey vessel's heading, where this condition corresponds to the tow cable's horizontal projection direction being approximated by the survey vessel's heading angle, the tow body coordinates are obtained according to the following formula:

[0097] .

[0098] In the formula, For towing sonar tow body. The coordinates of the towed sonar towed body in the geographic coordinate system are given. These are the planar coordinates of the drag point in the geographic coordinate system. This refers to the length of the tow cable. The angle of inclination of the tow cable relative to the vertical direction. Relative to geographic coordinate system The heading angle of the surveying vessel measured along the axial direction. and These are the two orthogonal coordinate components of the planar coordinate system.

[0099] Using the planar coordinates of the tow point in the geographic coordinate system as a reference, and assuming the tow body's azimuth is approximated by the survey vessel's heading, the horizontal projected displacement of the towed sonar body relative to the tow point is calculated based on the tow cable length, tow cable inclination angle, and survey vessel heading angle. This horizontal projected displacement is then superimposed onto the tow point coordinates to obtain the planar coordinates of the towed sonar body in the geographic coordinate system. These coordinates can serve as an important location input for the initial geocoding of side-scan sonar observation data.

[0100] The estimation of the towed body's position is used as the foundation for subsequent image misalignment correction. This estimation relies on the premise that "the towed body's orientation is approximated by the survey vessel's course." This premise may become a source of systematic error in complex water flow environments. Therefore, this application not only limits the position estimation model used in the initial geocoding but also provides an explanation of the source of error for the subsequent back-calculation of the course offset through misalignment of common feature points.

[0101] The towed body position estimation can incorporate the influence of tow cable length, tow cable inclination angle and survey vessel heading angle on the towed body position into the geocoding process, making the position benchmark of towed sonar observation data closer to the actual acquisition position of the towed body. This provides an initial positioning calculation path and allows systematic errors in complex hydrological environments to be identified and corrected by subsequent heading compensation steps.

[0102] In some embodiments, the lateral position error caused by the heading angle error is characterized by the following formula:

[0103] In the formula, The positional error is in the direction perpendicular to the survey line's heading. This refers to the slant range of a single survey line in a side-scan sonar. This represents the heading angle error.

[0104] When the system-recorded heading deviates from the actual heading, the heading compensation amount applied in post-processing is determined according to the following formula:

[0105] In the formula, This is the heading compensation amount applied during post-processing. To record the deviation of the heading from the actual heading, the negative sign indicates that the direction of the heading compensation is opposite to the direction of the deviation.

[0106] Heading angle error causes lateral position errors related to slant range at side-scan sonar observation points. Furthermore, when the recorded heading deviates from the actual heading, a heading compensation in the opposite direction can be applied during post-processing. This application formulates the relationship between image misalignment, heading angle error, and compensation direction, linking spatial misalignment in side-scan sonar images to the geometric amplification effect of heading angle error, and explicitly defining the reverse compensation relationship as the post-processing heading correction rule. This explains why the same target exhibits lateral misalignment related to slant range in images from different survey lines, and also explains why distant targets are more suitable for estimating heading offsets.

[0107] Based on this design, the direction and necessity of heading compensation become interpretable. Without an error propagation model, image misalignment might be mistaken for ordinary stitching errors or random noise, and post-processing would have to rely solely on trial and error based on experience. This claim... This reveals the relationship between lateral position error and slant range and heading angle errors, and then... Determining the compensation direction provides a clear rule for applying the heading offset. This rule helps reduce erroneous heading adjustments, improves parameter iteration efficiency, and enhances the stability of the final positioning results.

[0108] In some embodiments, common feature points are corresponding image features of the same underwater target that can be identified in both the first and second survey line images.

[0109] Common feature points include those located at the far end of the underwater target, where "far end" refers to the region closer to the edge of the side-scan sonar range relative to the near end of the corresponding survey line; the initial heading offset is estimated using the following formula:

[0110] ;

[0111] In the formula, This is the initial heading offset. The difference in lateral displacement of the same common feature point in the first and second survey line images. The average slant distance of this common feature point relative to the position of the surveying vessel. It is an arcsine function, a constant. This represents the bilateral misalignment contribution term formed by the first and second survey lines to the same common feature point.

[0112] This application defines common feature points as corresponding image features of the same underwater target that can be identified in both the first and second survey line images. Furthermore, it utilizes feature points located at the far end of the underwater target to estimate the initial heading offset. Since the far-end feature points are close to the edge of the side-scan sonar range, they respond more strongly to lateral displacement caused by heading errors. Therefore, by measuring the difference in lateral displacement of these common feature points in the two survey line images and combining this with the average slant range, the initial heading offset can be calculated.

[0113] The lateral displacement difference of common feature points at the far end is directly used to estimate the initial heading offset, and an approximate relationship between the misalignment of the two survey lines and the heading error on one side is established through the bilateral misalignment contribution term. Unlike the processing method that relies solely on empirical input of fixed angle offsets, this application uses the misalignment data of the target image itself to generate the first-round correction parameters, which can improve the efficiency and accuracy of the initial heading offset estimation. Since the near-end region of the side-scan sonar is less affected by the heading angle error, its misalignment may be difficult to measure, while the lateral displacement of the far-end region is more likely to reflect the heading error. Therefore, selecting common feature points at the far end can amplify the observable error signal. Obtaining the initial heading correction parameters can reduce the number of manual trials and provide initial values ​​that are closer to the true correction direction for subsequent iterative recoding, thereby improving the convergence efficiency of the overall processing flow.

[0114] In some embodiments, applying a reverse heading offset to at least one of the first survey line and the second survey line includes:

[0115] The initial heading offset is used as the first heading correction parameter.

[0116] When the misalignment of one survey line or the second survey line is greater than that of the other survey line, a reverse heading offset is applied to the survey line with the greater misalignment; wherein, the degree of misalignment is determined based on the remaining lateral displacement difference of common feature points, the root mean square error of the total offset vector, or the deviation from the coordinates of the check point.

[0117] When both the first and second survey lines have systematic offsets related to the corresponding navigation attitude data, a heading offset opposite to the direction of their respective heading errors is applied to the first and second survey lines respectively.

[0118] This application uses the initial heading offset as the first-round heading correction parameter and determines the correction target based on the degree of misalignment between the first and second survey lines. When the misalignment of one survey line is greater than that of the other, a reverse heading offset is applied to the survey line with the greater misalignment. When both survey lines have systematic deviations that need correction, heading offsets opposite to their respective heading errors are applied to each survey line. This design ensures that the heading offset is no longer fixed to a single survey line, but can be distributed according to the actual misalignment state of the two survey lines, thus providing a processing mechanism for selecting the correction target based on the degree of survey line misalignment. The misalignment of the two survey lines is not necessarily caused by a single survey line; it may be caused by the heading error of one survey line being dominant, or it may be caused by heading errors of different directions or magnitudes on both survey lines. Designing two paths, single-surface correction and separate correction for both survey lines, is beneficial for adapting to the distribution of systematic errors under different acquisition conditions.

[0119] This design improves the alignment between the heading compensation strategy and the actual error state of the survey lines. Applying the same compensation mechanically to two survey lines may introduce new deviations when one line is already close to the correct position; similarly, mechanically correcting only one survey line may fail to eliminate the compound misalignment formed when both lines have errors. This application determines the correction target and compensation direction based on the degree of misalignment, making heading compensation more targeted, reducing unnecessary secondary errors, and providing a more reasonable initial correction state for subsequent recoding and multi-feature point verification.

[0120] In some embodiments, re-geocoding includes:

[0121] After each heading offset is applied, the geocoding results of the first and second survey lines are regenerated based on the original navigation data, attitude data, and the applied heading offset, and the mosaic map is updated. The input parameters for each heading offset include the original navigation data, attitude data, the heading offset of the first survey line, and the heading offset of the second survey line.

[0122] The remaining lateral displacement difference of common feature points is remeasured in the updated mosaic.

[0123] The heading offset is adjusted based on the remaining lateral displacement difference and geocoding is repeated. The heading offset corresponding to the smallest remaining lateral displacement difference is determined as the candidate final heading offset. When there are several common feature points, the heading offset corresponding to the smallest sum of squares of the remaining lateral displacement differences of the several common feature points or the smallest root mean square error of the total offset vector is determined as the candidate final heading offset.

[0124] After each heading offset is applied, the geocoding results of the first and second survey lines are regenerated based on the original navigation data, attitude data, and the applied heading offset, and the mosaic map is updated. Then, the remaining lateral displacement difference of common feature points is remeasured in the updated mosaic map. The heading offset is then adjusted according to the remaining lateral displacement difference and geocoding is repeated until the candidate final heading offset corresponding to the smallest remaining lateral displacement difference is determined.

[0125] Based on the above design, this application extends the one-time estimation to a feedback iterative process. The heading offset correction is embedded in a closed-loop iterative process of "applying parameters, recoding, measuring residuals, and readjusting," rather than performing static correction at the image level. Since the spatial location of sonar images is determined by the geocoding process, each change in heading offset affects the spatial coordinates of each sampling point. Therefore, updating the mosaic by re-geocoding can reflect the true spatial effect of parameter changes.

[0126] This design reduces the impact of initial estimation errors on the final result. The initial heading offset, obtained from approximate geometric relationships, may be affected by feature point selection errors, slant range estimation errors, and local image distortion. Directly using it as the final parameter may not achieve optimal global alignment. By remeasuring the remaining lateral displacement difference after each recoding, the updated spatial results can be used to adjust the direction, gradually bringing the heading offset closer to the value that minimizes the residuals of multiple feature points. This iterative mechanism improves the reliability of the final heading offset and reduces the processing cost of repeated manual trial and error.

[0127] In some embodiments, verifying the alignment result using multiple common feature points includes:

[0128] Under the influence of the candidate final heading offset, examine the alignment of multiple common feature points at different locations in the updated mosaic.

[0129] When several common feature points distributed in different parts of an underwater target or different regions of a mosaic all form corresponding alignment results with the candidate final heading offset, the candidate final heading offset is determined to have global validity and is locked as the final heading offset; where alignment means that the remaining lateral displacement difference is less than a preset threshold, or the root mean square error of the total offset vector is less than a preset threshold.

[0130] When at least one common feature point does not form a corresponding alignment result, continue to adjust the heading offset to avoid using the local alignment result as the final result.

[0131] Under the influence of the candidate final heading offset, the alignment of multiple common feature points at different positions in the updated mosaic is checked; when multiple common feature points form corresponding alignment results with the candidate final heading offset, the candidate final heading offset is determined to be globally valid and locked as the final heading offset; when at least one common feature point does not form a corresponding alignment result, the heading offset is adjusted to avoid taking the local alignment result as the final result.

[0132] Instead of determining the final parameters based solely on the alignment of a single control point, a global verification of candidate heading offsets is performed using common feature points from multiple spatial locations. A single feature point in a side-scan sonar image can be affected by local reflection intensity, target occlusion, image stretching, or human error; single-point alignment does not prove that the entire mosaic has been reasonably corrected. Verification using multiple common feature points at different locations can identify local optima.

[0133] This improves the global applicability of the final heading offset. If only alignment of a single target point is used as the termination condition, images near that target point may appear to match, while other areas may still be misaligned, especially in long survey lines or complex terrain areas where this risk is more pronounced. This application, through simultaneous verification of multiple common feature points, ensures that the final heading offset must simultaneously meet the alignment requirements of different locations, thereby reducing the impact of local image feature errors on the overall results. This helps improve the overall consistency of the corrected mosaic and enhances the reliability of underwater target positioning coordinates.

[0134] In some embodiments, generating the corrected geocoded mosaic and underwater target location coordinates includes:

[0135] The first and second survey lines are re-geocoded using the final course offset, and the final course offset, the corrected geocoded mosaic, the underwater target plane position coordinates, and the accuracy assessment results are output.

[0136] Among them, the planar position coordinates of the underwater target are the geographic coordinates corresponding to the geometric center of the underwater target, the preset positioning point, or the confirmed common feature points.

[0137] The accuracy assessment results include the deviation before and after the checkpoint correction, the eastward offset of the feature point, the northward offset of the feature point, and the root mean square error of the total offset vector.

[0138] The underwater target is any underwater target with an identifiable geometric shape, such as a large barrel, pile foundation, pipeline, shipwreck, or artificial reef.

[0139] The first and second survey lines are re-geocoded using the final heading offset, and the final heading offset, the corrected geocoded mosaic, the underwater target's planar position coordinates, and the accuracy assessment results are output. The accuracy assessment results include the deviation of the checkpoint before and after correction, the eastward offset of the feature point, the northward offset of the feature point, and the root mean square error of the total offset vector. The underwater target can be a large cylinder, pile foundation, pipeline, shipwreck, or artificial reef, or any other target with an identifiable geometric shape.

[0140] By unifying the output of image results after heading offset correction with positioning coordinates and accuracy evaluation indicators, this application goes beyond mere image visual correction and forms a result representation for underwater target positioning. The corrected geocoded mosaic is used for image interpretation, the underwater target's planar position coordinates are used for engineering positioning, and the accuracy evaluation results are used to demonstrate the changes in spatial deviation before and after correction.

[0141] Therefore, this application can transform the side-scan sonar image processing results into verifiable engineering mapping results. Simply outputting corrected images is insufficient to prove whether the positioning coordinates meet the usage requirements, while simply outputting coordinates is not conducive to judging whether the image stitching and target shape are correct. This application simultaneously outputs the heading offset, mosaic map, planar position coordinates, and accuracy evaluation results, establishing a correspondence between data processing parameters, image results, positioning results, and error evaluation. This result format is beneficial for subsequent construction verification, structure positioning, underwater target identification, and quality traceability, and can be extended to various types of underwater targets with identifiable geometric shapes.

[0142] The side-scan sonar image position correction and positioning method provided in this application was applied to a breakwater paving project, and quantitative analysis was carried out. By comparing the data before and after correction, a quantitative accuracy assessment was conducted.

[0143] First, positional accuracy analysis was performed. The following results were obtained: Figure 2 The absolute positioning accuracy comparison table shown shows that after applying heading compensation, the average positioning error of the target point decreased from 4.5 meters to 0.73 meters, and the absolute accuracy improved by more than 83%, successfully meeting the sub-meter level engineering accuracy requirements.

[0144] Secondly, the accuracy of image matching and stitching was analyzed. Results were obtained as follows: Figure 3 The image matching accuracy table shown indicates that, after correction, the average total offset of feature point pairs was reduced by 85.7%, with the improvement being particularly significant (88.1%) for the eastward offset, which is most sensitive to heading errors. This results in natural feature edges in the final base map mosaic, with no discernible misalignments or gaps.

[0145] Finally, parameter correction and efficiency analysis were performed. The survey lines were processed, and statistical analysis revealed that the optimal heading offset ΔH is mainly distributed between -0.8° and +1.2°, which is in high agreement with the theoretical estimate calculated using the formula ΔHestimate. This confirms the reliability of the theoretical model in guiding practice.

[0146] In terms of efficiency, it shows an advantage over the purely manual trial-and-error method:

[0147] Manual trial and error method: It takes about 2 hours to process a single survey line on average, and the process requires the operator to have certain experience and gradually converge through iteration.

[0148] The method in this application is based on feature points and iterative processing. The streamlined operation reduces the average processing time of a single survey line to 0.5 hours, improving efficiency by 300% and reducing reliance on personnel experience.

[0149] Another embodiment of the present invention provides a side-scan sonar image position correction and positioning device, comprising:

[0150] The acquisition module is used to acquire raw side-scan sonar data from the first and second survey lines. The first and second survey lines are two survey lines formed by the side-scan sonar collecting data from the same underwater target along different tracks.

[0151] The first processing module is used to perform seabed line tracking, radiation distortion correction, and slant range correction on the side-scan sonar data of the first and second survey lines.

[0152] The first generation module is used to perform preliminary geocoding using the original navigation data and attitude data, and then overlay them to generate a preliminary mosaic map.

[0153] The identification module is used to identify common feature points of the same underwater target in the preliminary mosaic map, measure the lateral displacement difference of the common feature points in the two survey line images and estimate the average slope distance; wherein, the lateral displacement difference is the displacement difference measured along the lateral direction of the survey line in the geocoded mosaic map coordinate system; the average slope is the average slope distance obtained by averaging the slope distances of the common feature points corresponding to the first survey line and the second survey line respectively.

[0154] The calculation module is used to obtain the initial heading offset based on the lateral displacement difference and the average slant distance.

[0155] The second processing module is used to apply a reverse heading offset to at least one of the first survey line and the second survey line and re-geocode it.

[0156] The second generation module is used to verify the alignment result through several common feature points. When the remaining lateral displacement difference of several common feature points is less than a preset threshold, or when the root mean square error of the total offset vector of several common feature points reaches the minimum, the final heading offset is locked, and the corrected geocoded mosaic map and underwater target positioning coordinates are generated.

[0157] It should be noted that although several units or sub-units of the device have been mentioned in the detailed description above, this division is merely exemplary and not mandatory. In fact, according to embodiments of this application, the features and functions of two or more units described above can be embodied in one unit. Conversely, the features and functions of one unit described above can be further divided and embodied by multiple units.

[0158] Based on the same inventive concept as the above method embodiments, this application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it enables the electronic device to implement the control method described in the above embodiments.

[0159] In one embodiment, the electronic device may be a server, and in this embodiment, the structure of the electronic device may be as follows: Figure 4 As shown, it includes a memory, a communication module, and one or more processors.

[0160] Memory is used to store computer programs executed by the processor. Memory can be mainly divided into a program storage area and a data storage area. The program storage area can store the operating system and programs required to run instant messaging functions, etc.; the data storage area can store various instant messaging information and operation instruction sets, etc.

[0161] Memory can be volatile memory, such as random access memory (RAM); memory can also be non-volatile memory, such as read-only memory, flash memory, hard disk drive (HDD), or solid-state drive (SSD); or memory can be any other medium capable of carrying or storing a desired computer program having the form of instructions or data structures and accessible by a computer, but is not limited thereto. Memory can be a combination of the above-mentioned types of memory.

[0162] A processor may include one or more central processing units (CPUs) or digital processing units, etc. A processor is used to implement the aforementioned data processing methods when a computer program stored in memory is invoked.

[0163] The communication module is used to communicate with terminal devices and other servers.

[0164] This application embodiment does not limit the specific connection medium between the above-described memory, communication module, and processor. This application embodiment... Figure 4 The memory and processor are connected via a bus, and the bus is in... Figure 4 The connections between other components are illustrated with arrows and are for illustrative purposes only, not as limiting information. Buses can be categorized as address buses, data buses, control buses, etc. For ease of description, Figure 4 The text uses only one arrow to describe it, but does not indicate that there is only one bus or one type of bus.

[0165] Based on the same inventive concept as the above-described method embodiments, embodiments of the present invention also provide a computer-readable storage medium for storing a computer program. When the computer program is run on a computer, it enables an electronic device to implement the control methods described in the above embodiments. The computer-readable storage medium can be a readable signal medium or a readable storage medium. A readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of readable storage media (a non-exhaustive list) include: an electrical connection having one or more wires, a portable disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof.

[0166] Based on the same inventive concept as the above-described method embodiments, embodiments of the present invention also provide a computer program product. The computer program product includes a computer program that, when run on an electronic device, causes the electronic device to perform the steps of the control methods described above according to various exemplary embodiments of this application. The program product may take the form of any combination of one or more readable media. These computer program commands can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing device to produce a machine, such that the commands executed by the processor of the computer or other programmable data processing device generate a process for implementing... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0167] While the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the invention. Those skilled in the art can make various modifications and refinements without departing from the spirit and scope of the invention. Therefore, the scope of protection of the present invention shall be determined by the claims.

Claims

1. A method for position correction and localization of side-scan sonar images, characterized in that, include: Acquire raw side-scan sonar data for the first and second survey lines, which are two survey lines formed by side-scan sonar collecting data on the same underwater target along different tracks. The side-scan sonar data of the first and second survey lines are subjected to seabed line tracking, radiation distortion correction and slant range correction. Preliminary geocoding was performed using the original navigation data and attitude data, and then the data was overlaid to generate a preliminary mosaic map. In the preliminary mosaic map, common feature points of the same underwater target are identified, and the lateral displacement difference of the common feature points in the two survey line images is measured and the average slope distance is estimated. The lateral displacement difference is the displacement difference measured along the lateral direction of the survey line in the geocoded mosaic map coordinate system. The average slope is the average slope distance obtained by averaging the slope distances of the common feature points corresponding to the first survey line and the second survey line, respectively. The initial heading offset is obtained based on the lateral displacement difference and the average slant distance. Apply a reverse heading offset to at least one of the first survey line and the second survey line and regeocode it; The alignment results are verified by several common feature points. When the remaining lateral displacement difference of several common feature points is less than the preset threshold, or when the root mean square error of the total offset vector of several common feature points reaches the minimum, the final heading offset is locked, and the corrected geocoded mosaic map and underwater target positioning coordinates are generated.

2. The method for side-scan sonar image position correction and localization according to claim 1, characterized in that, The common feature points are derived from the port side sonar image data of the first survey line and the starboard side sonar image data of the second survey line when corresponding to the same underwater target, or from the starboard side sonar image data of the first survey line and the port side sonar image data of the second survey line when corresponding to the same underwater target. The seabed line tracking is used to determine the seabed echo boundary, the radiation distortion correction is used to correct the grayscale difference of the sonar image, and the slant range correction is used to convert the sonar slant range observation into a planar distance observation. The obtained side-scan sonar image data, corrected for seabed line, grayscale, and slant range, is used to identify common feature points of the same underwater target.

3. The method for side-scan sonar image position correction and localization according to claim 1, characterized in that, The preliminary geocoding includes determining the reference location of the side-scan sonar observation data based on the tow point coordinates. The tow point coordinates are used as the reference point for subsequent tow body coordinate calculations or geocoding of the sonar observation data. The tow point coordinates are obtained according to the following formula: ； In the formula, For geographic coordinate system, For the ship's hull, a fixed coordinate system is established. For drag point object identification, For shipborne satellite positioning antennas, These are the planar coordinates of the drag point in the geographic coordinate system. The coordinates of the shipborne satellite positioning antenna in the geographic coordinate system are given. Let be the planar coordinates of the shipborne satellite positioning antenna in the ship's fixed coordinate system. Let be the planar coordinates of the towing point in the fixed coordinate system of the ship's hull. This is the rotation matrix corresponding to the heading angle. This is the rotation matrix corresponding to the pitch angle. This is the rotation matrix corresponding to the roll angle. For heading angle, For pitch angle, The roll angle, and These are the two orthogonal coordinate components of the planar coordinate system.

4. The method for side-scan sonar image position correction and localization according to claim 3, characterized in that, In the initial geocoding, under the condition that the tow body's orientation is approximated by the survey vessel's heading, where this condition corresponds to the tow cable's horizontal projection direction being approximated by the survey vessel's heading angle, the tow body coordinates are obtained according to the following formula: ； In the formula, For towing the sonar tow body. The coordinates of the towed sonar body in the geographic coordinate system are given. These are the planar coordinates of the drag point in the geographic coordinate system. This refers to the length of the tow cable. The angle of inclination of the tow cable relative to the vertical direction. Relative to geographic coordinate system The heading angle of the surveying vessel measured along the axial direction. and These are the two orthogonal coordinate components of the planar coordinate system.

5. The method for side-scan sonar image position correction and localization according to claim 1, characterized in that, The lateral position error caused by the heading angle error is characterized by the following formula: In the formula, The positional error is in the direction perpendicular to the survey line's heading. This refers to the slant range of a single survey line in a side-scan sonar. This refers to the heading angle error; When the system-recorded heading deviates from the actual heading, the heading compensation amount applied in post-processing is determined according to the following formula: In the formula, This is the heading compensation amount applied during post-processing. The negative sign indicates that the direction of the heading compensation is opposite to the direction of the deviation, so that the system records the deviation of the heading relative to the actual heading.

6. The method for side-scan sonar image position correction and localization according to claim 1, characterized in that, The common feature points are the corresponding image features of the same underwater target that can be identified in both the first and second survey line images; The common feature points include feature points located at the far end of the underwater target, where the far end is the region closer to the edge of the side-scan sonar range relative to the near end region of the corresponding survey line; the initial heading offset is estimated according to the following formula: ； In the formula, This is the initial heading offset. The difference in lateral displacement of the same common feature point in the first and second survey line images. The average slant distance of this common feature point relative to the position of the surveying vessel. It is an arcsine function, a constant. This represents the bilateral misalignment contribution term formed by the first and second survey lines for the same common feature point.

7. The method for side-scan sonar image position correction and localization according to claim 6, characterized in that, Applying a reverse heading offset to at least one of the first survey line and the second survey line includes: The initial heading offset is used as the first heading correction parameter; When the misalignment of one of the first or second survey lines is greater than that of the other survey line, a reverse heading offset is applied to the survey line with the greater misalignment; wherein, the degree of misalignment is determined based on the remaining lateral displacement difference of common feature points, the root mean square error of the total offset vector, or the deviation from the coordinates of the check point; When both the first survey line and the second survey line have systematic offsets related to the corresponding navigation attitude data, a heading offset amount opposite to the direction of their respective heading errors is applied to the first survey line and the second survey line respectively.

8. The method for side-scan sonar image position correction and localization according to claim 7, characterized in that, Re-geocoding includes: After each heading offset is applied, the geocoding results of the first and second survey lines are regenerated based on the original navigation data, attitude data, and the applied heading offset, and the mosaic map is updated; wherein, the input parameters for each heading offset include the original navigation data, attitude data, the heading offset of the first survey line, and the heading offset of the second survey line. The remaining lateral displacement difference of common feature points is remeasured in the updated mosaic. The heading offset is adjusted based on the remaining lateral displacement difference and geocoding is repeated. The heading offset corresponding to the smallest remaining lateral displacement difference is determined as the candidate final heading offset. When there are several common feature points, the heading offset corresponding to the smallest sum of squares of the remaining lateral displacement differences of the several common feature points or the smallest root mean square error of the total offset vector is determined as the candidate final heading offset.

9. A method for side-scan sonar image position correction and localization according to claim 8, characterized in that, The alignment results were verified using multiple common feature points, including: Under the influence of the candidate final heading offset, check the alignment of multiple common feature points at different positions in the updated mosaic map; When several common feature points distributed in different parts of an underwater target or different regions of a mosaic form corresponding alignment results with the candidate final heading offset, the candidate final heading offset is determined to have global validity and is locked as the final heading offset; where alignment means that the remaining lateral displacement difference is less than a preset threshold, or the root mean square error of the total offset vector is less than a preset threshold. When at least one common feature point does not form a corresponding alignment result, continue to adjust the heading offset to avoid using the local alignment result as the final result.

10. A method for side-scan sonar image position correction and localization according to claim 1, characterized in that, The generated corrected geocoded mosaic map and underwater target location coordinates include: The first and second survey lines are re-geocoded using the final heading offset, and the final heading offset, the corrected geocoded mosaic map, the underwater target plane position coordinates, and the accuracy assessment results are output. Among them, the planar position coordinates of the underwater target are the geographic coordinates corresponding to the geometric center of the underwater target, the preset positioning point, or the confirmed common feature points; The accuracy assessment results include the deviation before and after the checkpoint correction, the eastward offset of the feature point, the northward offset of the feature point, and the root mean square error of the total offset vector. The underwater target is any one of the following: a large cylindrical barrel, a pile foundation, a pipeline, a shipwreck, or an artificial reef, and has an identifiable geometric shape.

Images