Underwater salvage positioning monitoring method

Abstract

The application discloses a kind of underwater salvage positioning monitoring methods, belong to underwater positioning monitoring technical field.For solving the problem of ship positioning, grab tracking and underwater imaging data dispersion in existing salvage operation, lack unified monitoring system, the ship positioning information and attitude data are obtained by GNSS receiver and attitude instrument and error correction;Grab position information is obtained by ultra-short baseline communication;Two-dimensional scanning sonar is lowered into water to obtain two-dimensional acoustic image;Three-dimensional image of grab and sunken ship target is obtained by three-dimensional real-time sonar;Sound velocity profile instrument is used to measure sound velocity to correct the error of sonar image;The corrected ship positioning information, grab position information, two-dimensional acoustic image, three-dimensional image and sunken ship background image are fused and displayed in the same coordinate system, and the grab is commanded to grab the sunken ship target according to the fusion information.The application is mainly used for accurate positioning and real-time monitoring in sunken ship salvage operation.

Description

Technical Field

[0001] This invention belongs to the field of underwater positioning and monitoring technology, specifically relating to an underwater salvage positioning and monitoring method. Background Technology

[0002] In underwater salvage operations, precise positioning of the sunken vessel and real-time monitoring of the grab bucket are crucial for ensuring salvage efficiency and operational safety. Traditional salvage operations rely primarily on a combination of coarse positioning by surface vessels and underwater exploration by divers, which suffers from low positioning accuracy, low operational efficiency, and high safety risks. While underwater salvage positioning and monitoring methods are constantly being updated with advancements in sonar and satellite positioning technologies, numerous technical challenges remain in practical applications.

[0003] Salvage operations involve vessels constantly moving on the water's surface due to wind, waves, and currents. The vessel positioning information directly output by the GNSS receiver includes dynamic errors. Because of the vessel's rolling and displacement, there is a deviation between the GNSS receiving antenna position and the vessel's actual operating position, leading to a decrease in positioning accuracy. Traditional methods typically use single-point positioning or differential positioning to obtain the vessel's position, but these fail to effectively eliminate the impact of changes in vessel attitude on positioning accuracy, making it difficult to meet the precise positioning requirements of salvage operations.

[0004] Existing underwater salvage positioning and monitoring technologies have shortcomings in areas such as ship dynamic error correction, precise grab positioning, sound speed error correction, multi-source data fusion, and real-time operation guidance, making it difficult to meet the requirements for positioning accuracy and real-time performance in complex waterway shipwreck salvage operations. Summary of the Invention

[0005] One object of the present invention is to solve at least the above-mentioned problems and / or defects, and to provide at least the advantages described below.

[0006] One objective of this invention is to address the problem in existing underwater salvage positioning and monitoring methods where multiple data sources, such as ship positioning, grab tracking, and underwater imaging, are scattered and independent, lacking a unified monitoring system, and thus failing to achieve accurate positioning and real-time monitoring of the entire salvage operation process.

[0007] One objective of this invention is to address the problem that when a ship rolls and shifts under the influence of wind, waves, and currents, the ship positioning information directly output by the GNSS receiver contains dynamic errors and cannot accurately reflect the ship's actual operating position.

[0008] One objective of this invention is to address the problem that in ultra-short baseline positioning systems, factors such as changes in sound speed, multipath interference, beacon attitude changes, and multi-level coordinate transformations lead to a decrease in the underwater positioning accuracy of the grab bucket, making it difficult to meet the precise positioning requirements of the grab bucket.

[0009] One objective of this invention is to solve the problem that the actual sound velocity in the operating water area differs from the preset nominal sound velocity of the equipment during two-dimensional scanning sonar and three-dimensional real-time sonar imaging, leading to errors in the calculation of sound wave propagation distance and distortion of target position in the image.

[0010] One objective of this invention is to solve the problem that the coordinate systems of multiple data sources, such as ship positioning information, grab bucket position information, two-dimensional acoustic images, three-dimensional images, and shipwreck background images, are not unified and cannot be fused and displayed under the same reference, which affects the accurate determination of the relative position of the grab bucket and the shipwreck target.

[0011] One objective of this invention is to address the problem of how sound velocity profile data can be applied to sonar image error correction.

[0012] One objective of this invention is to solve the problem of how to update the position coordinates of each pixel in a sonar image based on the new distance value and the original angle information after the sound wave propagation distance has been recalculated, thereby achieving image geometric correction.

[0013] One objective of this invention is to solve the problem that the coordinate systems of two-dimensional acoustic images and three-dimensional images are not unified and cannot be directly superimposed and fused.

[0014] One objective of this invention is to address the problem that when the grab bucket approaches the sunken ship target, operators cannot accurately judge the safe distance based solely on visual judgment, which can easily lead to collisions between the grab bucket and non-target structures such as the immersed tube.

[0015] One object of the present invention is to provide an underwater salvage positioning and monitoring method, comprising the following steps: Step S1: Install a GNSS receiver, attitude sensor and inertial navigation system on the main construction vessel, install an ultra-short baseline beacon on the grab bucket, and install an ultra-short baseline transducer on the side of the main construction vessel to complete the deployment of the positioning and monitoring system; Step S2: Obtain the original ship positioning information of the main construction vessel through the GNSS receiver, obtain the real-time attitude data of the main construction vessel through the attitude sensor and inertial navigation system, and perform dynamic error correction on the original ship positioning information based on the real-time attitude data to obtain the corrected ship positioning information. Step S3: Through communication between the ultra-short baseline transducer and the ultra-short baseline beacon, the grab bucket is tracked in real time to obtain the grab bucket's position information; Step S4: Lower the two-dimensional scanning sonar installed at the stern of the main construction vessel into the water to scan the area around the sunken ship and obtain a two-dimensional acoustic image of the area around the sunken ship. Step S5: By transmitting sound waves underwater and receiving echoes through a three-dimensional real-time sonar installed on the side of the main construction vessel, a three-dimensional image of the relative positional relationship between the grab bucket and the sunken ship target is generated. Step S6: Measure the sound velocity in water in real time using a sound velocity profiler, input the measured sound velocity data into the two-dimensional scanning sonar and the three-dimensional real-time sonar, and perform real-time error correction on the two-dimensional acoustic image and the three-dimensional image to obtain the corrected two-dimensional acoustic image and the corrected three-dimensional image. Step S7: Overlay and fuse the corrected ship positioning information, grab bucket position information, corrected two-dimensional acoustic image and corrected three-dimensional image, as well as the pre-imported background image of the sunken ship and immersed tube position, in the same coordinate system. Based on the fused information, determine the relative position of the grab bucket and the sunken ship target in real time, and command the grab bucket to grab the sunken ship target.

[0016] Preferably, in the underwater salvage positioning and monitoring method, the specific implementation process of step S2 includes: Step S2-1: Obtain the original vessel positioning information of the main construction vessel in real time through a GNSS receiver. The original vessel positioning information includes longitude coordinates, latitude coordinates, and geodetic coordinates. Step S2-2: Acquire real-time attitude data of the main construction vessel through the attitude sensor and inertial navigation system. The real-time attitude data includes roll angle, pitch angle and bow angle. Step S2-3: Calculate the offset of the GNSS receiving antenna installation position relative to the ship reference point based on real-time attitude data, and use a coordinate rotation algorithm to convert the antenna position offset caused by the ship attitude change into a displacement in the geodetic coordinate system. Step S2-4: Subtract the displacement calculated in step S2-3 from the original ship positioning information to obtain the corrected ship positioning information.

[0017] Preferably, in the underwater salvage positioning and monitoring method, the specific implementation process of step S3 includes: Step S3-1: Transmit an acoustic interrogation signal to the US. beacon through the US. transducer and receive the response signal transmitted by the beacon. Calculate the slant range between the transducer and the beacon based on the time difference between the transmission of the interrogation signal and the reception of the response signal. Calculate the horizontal and vertical azimuth angles of the beacon relative to the transducer based on the phase difference of the received response signal. Then, determine the three-dimensional coordinates of the grab bucket in the local coordinate system of the transducer through polar coordinate transformation based on the slant range, horizontal azimuth angle, and vertical azimuth angle. Step S3-2: Based on the pre-measured offset of the transducer installation position relative to the ship reference point, convert the three-dimensional coordinates of the grab bucket in the local coordinate system of the transducer into the three-dimensional coordinates in the ship coordinate system. Step S3-3: Combining real-time attitude data, the three-dimensional coordinates of the grab bucket in the ship coordinate system are converted into three-dimensional coordinates in the station center coordinate system through a coordinate rotation algorithm; Step S3-4: Combining the original ship positioning information, the three-dimensional coordinates of the grab bucket in the station center coordinate system are converted into the grab bucket position information in the geodetic coordinate system using a coordinate translation algorithm.

[0018] Preferably, in the underwater salvage positioning and monitoring method, the specific implementation process of real-time error correction of the two-dimensional acoustic image and the three-dimensional image in step S6 to obtain the corrected two-dimensional acoustic image and the corrected three-dimensional image includes: Step S6-1: Measure the sound velocity values ​​at different depths in the working water area in real time using a sound velocity profiler to generate sound velocity profile data; Step S6-2: Based on the sound velocity profile data, calculate the actual sound velocity of the sound wave on the propagation path between the two-dimensional scanning sonar transducer and the physical position corresponding to each pixel of the two-dimensional acoustic image, and the actual sound velocity of the sound wave on the propagation path between the three-dimensional real-time sonar transducer and the physical position corresponding to each pixel of the three-dimensional image, respectively, and replace the preset nominal sound velocities of the two-dimensional scanning sonar and the three-dimensional real-time sonar. Step S6-3: Substitute the actual sound velocity corresponding to the two-dimensional scanning sonar into the sonar ranging formula to recalculate the sound wave propagation distance corresponding to each pixel in the two-dimensional acoustic image, and substitute the actual sound velocity corresponding to the three-dimensional real-time sonar into the sonar ranging formula to recalculate the sound wave propagation distance corresponding to each pixel in the three-dimensional image. Step S6-4: Update the position coordinates of each pixel in the two-dimensional acoustic image according to the sound wave propagation distance corresponding to each pixel in the recalculated two-dimensional acoustic image, and update the position coordinates of each pixel in the three-dimensional image according to the sound wave propagation distance corresponding to each pixel in the recalculated three-dimensional image. Step S6-5: Output the updated 2D acoustic image and the updated 3D image as the corrected 2D acoustic image and the corrected 3D image.

[0019] Preferably, in the underwater salvage positioning and monitoring method, the specific implementation process of step S7 includes: Step S7-1: Establish a unified geodetic coordinate system as the fusion benchmark, and convert the corrected ship positioning information, grab bucket position information, corrected two-dimensional acoustic image, corrected three-dimensional image, and background image of the sunken ship and immersed tube position to the geodetic coordinate system. Step S7-2: Perform image registration on the corrected two-dimensional acoustic image and the corrected three-dimensional image, extract the feature points of the grab bucket and the sunken ship target in the image, and calculate the relative position of the grab bucket and the sunken ship target based on the coordinates of the feature points; Step S7-3: Spatially correlate the corrected ship positioning information with the grab bucket position information to determine the spatial position of the grab bucket relative to the main construction vessel, and combine the corrected ship positioning information to convert the grab bucket position information to a unified geodetic coordinate system. Step S7-4: The corrected two-dimensional acoustic image and the corrected three-dimensional image registered in step S7-2, the grab bucket position information converted in step S7-3, and the pre-imported background image of the sunken ship and immersed tube position are superimposed according to the coordinate correspondence to generate a fused comprehensive positioning image. Step S7-5: In the fused integrated positioning image, mark the relative distance and azimuth angle between the grab bucket and the sunken ship target in real time, and transmit the fused integrated positioning image to the crane control room display terminal in real time, so that the operator can direct the grab bucket to grab the target according to the relative position of the grab bucket and the sunken ship target in the fused integrated positioning image.

[0020] Preferably, in the underwater salvage positioning and monitoring method, the specific implementation process of step S6-2 includes: Sound velocity profilers are deployed near two-dimensional scanning sonar and three-dimensional real-time sonar to measure the sound velocity at different depths in the operating water area in real time and generate sound velocity profile data. Based on the sound velocity profile data, the average sound velocity along the propagation path of the sound wave from the transducer of the two-dimensional scanning sonar to the physical location of each pixel in the two-dimensional acoustic image is calculated, which is taken as the actual sound velocity of the two-dimensional scanning sonar. Based on the sound velocity profile data, the average sound velocity along the propagation path of the sound wave from the transducer of the three-dimensional real-time sonar to the physical location corresponding to each pixel in the three-dimensional image is calculated, which is taken as the actual sound velocity of the three-dimensional real-time sonar. The calculated actual sound velocities of the two-dimensional scanning sonar and the three-dimensional real-time sonar are used to replace the preset nominal sound velocities of the two-dimensional scanning sonar and the three-dimensional real-time sonar, respectively.

[0021] Preferably, in the underwater salvage positioning and monitoring method, the specific implementation process of updating the position coordinates of each pixel in the two-dimensional acoustic image and the position coordinates of each pixel in the three-dimensional image in step S6-4 includes: For two-dimensional scanning sonar, the horizontal and vertical azimuth angles corresponding to each pixel in the two-dimensional acoustic image are obtained. The recalculated sound wave propagation distance is used as the radial distance. Taking the transducer position as the origin, the updated position coordinates of each pixel in the local coordinate system of the transducer are determined according to the directions indicated by the horizontal and vertical azimuth angles. The original position coordinates of the pixel are replaced with the updated position coordinates. For 3D real-time sonar, the beam pointing angle and elevation angle corresponding to each pixel in the 3D image are obtained. The recalculated sound wave propagation distance is used as the radial distance. Taking the transducer position as the origin, the updated position coordinates of each pixel in the local coordinate system of the transducer are determined according to the direction indicated by the beam pointing angle and elevation angle. The original position coordinates of the pixel are replaced with the updated position coordinates.

[0022] Preferably, in the underwater salvage positioning and monitoring method, the specific implementation process of image registration of the corrected two-dimensional acoustic image and the corrected three-dimensional image in step S7-2 includes: Extract the edge contours of the shipwreck structure and the contours of the grab bucket from the corrected 2D acoustic image as 2D registration features; Extract the point cloud of the shipwreck structure and the point cloud of the grab bucket from the corrected 3D image as 3D registration features; The two-dimensional registration features are matched with the three-dimensional registration features, and the rotation and translation parameters between the two-dimensional acoustic image and the three-dimensional image are calculated. Based on the rotation and translation parameters, the corrected two-dimensional acoustic image and the corrected three-dimensional image are aligned to a unified geodetic coordinate system.

[0023] Preferably, in the underwater salvage positioning and monitoring method, in step S7-5, an early warning is issued when the relative distance between the grab bucket and the sunken ship target is less than a preset safety threshold.

[0024] The present invention has at least the following beneficial effects: This invention constructs a complete positioning and monitoring system by integrating a GNSS receiver, attitude sensor, inertial navigation system, ultra-short baseline transducer, two-dimensional scanning sonar, three-dimensional real-time sonar, and sound velocity profiler on the main construction vessel. This system achieves integrated ship positioning, grab tracking, underwater imaging, and sound velocity correction. This method integrates multi-source data acquisition into a unified technical framework, avoiding system compatibility issues caused by the fragmentation and independence of various monitoring methods, and providing a system foundation for subsequent data fusion and real-time monitoring.

[0025] This invention acquires real-time ship attitude data and uses a coordinate rotation algorithm to dynamically correct the raw ship positioning information output by the GNSS receiver, eliminating antenna position offsets caused by changes in ship roll, pitch, and heading. The corrected ship positioning information accurately reflects the actual position of the ship's reference point, improving ship positioning accuracy and providing a precise ship position benchmark for subsequent grab position coordinate conversion.

[0026] This invention utilizes acoustic communication between an ultra-short baseline transducer and a beacon to obtain the three-dimensional coordinates of the grab bucket in the transducer's local coordinate system. Combining this with the transducer's installation position offset, real-time ship attitude data, and original ship positioning information, a three-level coordinate transformation is performed: from the transducer's local coordinate system to the ship's coordinate system, from the ship's coordinate system to the station-centered coordinate system, and from the station-centered coordinate system to the geodetic coordinate system. This transformation method comprehensively considers multiple factors, including the transducer's installation position, changes in ship attitude, and the ship's position, achieving accurate calculation of the grab bucket's position.

[0027] This invention uses a sound velocity profiler to measure the sound velocity at different depths in the operating water area in real time, generating sound velocity profile data. Based on this data, the actual sound velocity along the propagation path between the transducer and the corresponding physical location of each pixel is calculated, replacing the preset nominal sound velocity of the sonar equipment. This method eliminates the ranging error caused by the difference between the actual sound velocity in the water area and the nominal sound velocity, and updates the position coordinates of each pixel based on the recalculated sound wave propagation distance, achieving geometric correction of the sonar image and making the target location in the image closer to the actual physical location.

[0028] This invention establishes a unified geodetic coordinate system as the fusion benchmark, transforming the corrected ship positioning information, grab bucket position information, corrected 2D acoustic images, corrected 3D images, and background images of the sunken ship and immersed tube locations into the same coordinate system. After image registration, spatial correlation, and overlay fusion, a comprehensive positioning image is generated. This method solves the problem of inconsistent coordinate systems among multi-source data, enabling operators to intuitively obtain the relative positional relationship between the grab bucket and the sunken ship target within the same image, providing clear and accurate visual guidance for grab bucket operations.

[0029] This invention utilizes sound velocity profilers deployed near two-dimensional scanning sonar and three-dimensional real-time sonar to measure the vertical distribution of sound velocity in the operating water area in real time. Based on the sound velocity profile data, the average sound velocity along the propagation path from the transducer to the corresponding physical location of each pixel is calculated and used as the actual sound velocity to replace the device's preset nominal sound velocity. This method transforms sound velocity profile data into actual sound velocity parameters that can be directly used for sonar ranging correction, providing accurate basic data for subsequent recalculation of sound wave propagation distance.

[0030] This invention obtains the horizontal and vertical azimuth angles corresponding to each pixel in a two-dimensional acoustic image, and the beam pointing and pitch angles corresponding to each pixel in a three-dimensional image. Using the recalculated sound wave propagation distance as the radial distance, and with the transducer position as the origin, the updated position coordinates of each pixel in the transducer's local coordinate system are determined according to the direction indicated by the angle information. This method achieves geometric reconstruction of the pixel positions in the image after sound velocity correction, ensuring that the geometric structure of the sonar image remains consistent with the actual underwater spatial structure.

[0031] This invention extracts the edge contours of the shipwreck structure and the outline of the grab bucket from the corrected 2D acoustic image as 2D registration features, and extracts the point clouds of the shipwreck structure and the grab bucket from the corrected 3D image as 3D registration features. After matching the 2D and 3D features, rotation and translation parameters are calculated, and finally, the 2D acoustic image and the 3D image are aligned to a unified geodetic coordinate system. This method achieves accurate registration of heterogeneous images, providing a spatial alignment basis for multi-source image overlay and fusion.

[0032] This invention presets a safety threshold in the fused integrated positioning image. When the relative distance between the grab bucket and the sunken ship target is less than this threshold, an automatic warning is issued. This method provides operators with a distance warning function, assisting them in adjusting their grabbing actions in a timely manner as the grab bucket approaches the sunken ship target, thus avoiding collisions between the grab bucket and non-target structures such as the immersed tube.

[0033] Other advantages, objectives and features of the present invention will become apparent in part from the following description, and in part from those skilled in the art through study and practice of the invention. Attached Figure Description

[0034] Figure 1 This is a flowchart of the underwater salvage positioning and monitoring method described in this invention. Detailed Implementation

[0035] The present invention will now be described in further detail so that those skilled in the art can implement it based on the description.

[0036] like Figure 1 As shown, the present invention provides an underwater salvage positioning and monitoring method, comprising the following steps: Step S1: Install a GNSS receiver, attitude sensor and inertial navigation system on the main construction vessel, install an ultra-short baseline beacon on the grab bucket, and install an ultra-short baseline transducer on the side of the main construction vessel to complete the deployment of the positioning and monitoring system; Step S2: Obtain the original ship positioning information of the main construction vessel through the GNSS receiver, obtain the real-time attitude data of the main construction vessel through the attitude sensor and inertial navigation system, and perform dynamic error correction on the original ship positioning information based on the real-time attitude data to obtain the corrected ship positioning information. Step S3: Through communication between the ultra-short baseline transducer and the ultra-short baseline beacon, the grab bucket is tracked in real time to obtain the grab bucket's position information; Step S4: Lower the two-dimensional scanning sonar installed at the stern of the main construction vessel into the water to scan the area around the sunken ship and obtain a two-dimensional acoustic image of the area around the sunken ship. Step S5: By transmitting sound waves underwater and receiving echoes through a three-dimensional real-time sonar installed on the side of the main construction vessel, a three-dimensional image of the relative positional relationship between the grab bucket and the sunken ship target is generated. Step S6: Measure the sound velocity in water in real time using a sound velocity profiler, input the measured sound velocity data into the two-dimensional scanning sonar and the three-dimensional real-time sonar, and perform real-time error correction on the two-dimensional acoustic image and the three-dimensional image to obtain the corrected two-dimensional acoustic image and the corrected three-dimensional image. Step S7: Overlay and fuse the corrected ship positioning information, grab bucket position information, corrected two-dimensional acoustic image and corrected three-dimensional image, as well as the pre-imported background image of the sunken ship and immersed tube position, in the same coordinate system. Based on the fused information, determine the relative position of the grab bucket and the sunken ship target in real time, and command the grab bucket to grab the sunken ship target.

[0037] Existing underwater salvage positioning and monitoring methods typically employ decentralized and independent monitoring approaches. Ship positioning relies on GNSS receivers to acquire location information, grab tracking utilizes ultra-short baseline positioning systems to determine underwater position, and underwater environmental perception relies on sonar equipment to acquire image information. During salvage operations, operators need to observe the ship's position, grab's position, and sonar images on different display terminals, relying on experience to comprehensively judge the relative position of the grab and the sunken ship. This decentralized monitoring method has the following shortcomings: each monitoring system operates independently, lacking correlation between data; operators need to monitor multiple display terminals simultaneously, resulting in low information acquisition efficiency; data formats and coordinate systems are inconsistent across different systems, making comprehensive judgment difficult; and a unified technical framework is lacking for integrating and processing multi-source data.

[0038] The underwater salvage positioning and monitoring method of the present invention constructs a unified positioning and monitoring system through the following specific embodiments.

[0039] A positioning and monitoring system is deployed on the main construction vessel. A GNSS receiver is installed in an unobstructed area above the bridge. This receiver receives satellite signals and differential correction signals, enabling the acquisition of the main construction vessel's raw positioning information. An attitude sensor and an inertial navigation system are installed near the vessel's center of gravity. The attitude sensor measures the vessel's roll and pitch angles in real time, while the inertial navigation system measures the vessel's heading angle and linear and angular accelerations in three directions in real time. An ultra-short baseline transducer is installed on the side of the main construction vessel near the crane's operating area. This location is chosen to be far from the vessel's engine and propeller to reduce mechanical noise interference with acoustic communications. A beacon protection cage is welded to the center of the grab bucket, and the ultra-short baseline beacon is installed inside the cage, ensuring that the beacon is not near the wire rope and remains vertically upward. A two-dimensional scanning sonar is installed at the stern of the main construction vessel. The sonar is lowered into the water using a cage suspension method, with counterweights at the bottom of the cage to reduce equipment sway caused by currents and waves. A 3D real-time sonar transducer was installed on the port aft side of the main construction vessel, a location conducive to acquiring 3D images of the grab bucket operation area. Sound velocity profilers were deployed near the 2D scanning sonar and the 3D real-time sonar to measure sound velocity values ​​at different depths in the operating water area in real time. With the installation of these devices, the basic deployment of the positioning and monitoring system was completed.

[0040] After system deployment, equipment debugging and data communication testing were conducted. After powering on each device, the data output was checked for normality. It was confirmed that the GNSS receiver could stably output raw ship positioning information, the attitude sensor and inertial navigation system could continuously output real-time attitude data, the ultra-short baseline transducer and beacon could establish a stable acoustic communication link, the 2D scanning sonar and 3D real-time sonar could acquire underwater images normally, and the sound velocity profiler could output sound velocity profile data in real time. A local area network was established on the main construction vessel, connecting all positioning and monitoring equipment to the network to achieve real-time data transmission and sharing. Unified coordinate system parameters, including the reference ellipsoid, central meridian, and projection method, were set in the navigation and positioning software to prepare for subsequent data fusion processing.

[0041] After the salvage operation begins, the system enters real-time monitoring mode. The GNSS receiver continuously outputs the original vessel positioning information of the main construction vessel at a fixed frequency, including the vessel's longitude, latitude, and geodetic coordinates. The attitude sensor and inertial navigation system continuously output real-time attitude data of the main construction vessel at a higher frequency, including the vessel's roll, pitch, and heading angles. The ultra-short baseline transducer transmits interrogation signals to the beacon on the underwater grab bucket at a set period and receives the response signals returned by the beacon. The slant distance between the transducer and the beacon is calculated using the round-trip time of the signals, and the horizontal and vertical azimuth angles of the beacon relative to the transducer are calculated using the phase difference of the received signals, thereby determining the three-dimensional coordinates of the grab bucket in the transducer's local coordinate system. After the two-dimensional scanning sonar is lowered to the predetermined depth, it performs a fan-shaped scan of the area around the sunken ship to acquire a two-dimensional acoustic image of the area around the sunken ship. The three-dimensional real-time sonar transmits sound waves underwater and receives the echoes to generate a three-dimensional image of the relative positional relationship between the grab bucket and the sunken ship target. The sound velocity profiler measures the sound velocity at different depths in the operating water area in real time, generates sound velocity profile data, and inputs it into the two-dimensional scanning sonar and the three-dimensional real-time sonar.

[0042] In the data processing stage, the system comprehensively processes the raw ship positioning information and real-time attitude data. Based on the real-time attitude data, a correction model is established to assess the impact of ship attitude changes on the GNSS receiving antenna position. The antenna position offset caused by changes in ship roll, pitch, and heading is calculated, and this offset is removed from the raw ship positioning information to obtain the corrected ship positioning information. The corrected ship positioning information more accurately reflects the actual position of the ship's reference point.

[0043] The grab position information acquired by the ultra-short baseline transducer undergoes multi-stage coordinate transformation. First, based on the offset of the transducer's installation position relative to the ship's reference point, the grab's three-dimensional coordinates in the transducer's local coordinate system are converted to three-dimensional coordinates in the ship's coordinate system. Then, combined with real-time attitude data, a coordinate rotation algorithm is used to convert the grab's three-dimensional coordinates in the ship's coordinate system to three-dimensional coordinates in the station-centered coordinate system. Finally, combined with the original ship positioning information, a coordinate translation algorithm is used to convert the grab's three-dimensional coordinates in the station-centered coordinate system to grab position information in the geodetic coordinate system. After these transformations, the grab's position information and the ship's positioning information are in the same coordinate system.

[0044] The sound velocity data output by the sound velocity profiler is used to correct sonar images. The measured sound velocity data is input into the 2D scanning sonar and the 3D real-time sonar, replacing the preset nominal sound velocity, to perform real-time error correction on the 2D acoustic images and 3D images. This eliminates geometric distortion caused by changes in sound velocity, resulting in corrected 2D acoustic images and corrected 3D images. In the corrected sonar images, the shape and position of the shipwreck structure and grab bucket are closer to their actual physical state.

[0045] Finally, multi-source data fusion and display are performed. The corrected ship positioning information, grab bucket position information, corrected 2D acoustic images, corrected 3D images, and pre-imported background images of the sunken ship and immersed tunnel locations are converted to the same coordinate system and overlaid for fusion display. The fused comprehensive positioning image simultaneously displays the ship's position, grab bucket position, sunken ship structure, immersed tunnel location, and the relative positional relationship between the grab bucket and the sunken target. Operators use the fused information to determine the relative position of the grab bucket and the sunken target in real time and direct the grab bucket to grasp the sunken target.

[0046] This invention integrates multiple positioning and monitoring devices on the main construction vessel, consolidating vessel positioning, grab tracking, underwater imaging, and sound velocity correction into a unified technical framework, thus constructing a complete salvage positioning and monitoring system. Compared with existing decentralized monitoring methods, this invention enables operators to simultaneously acquire vessel position, grab position, and underwater image information on a single display terminal, avoiding visual fatigue and information omissions caused by observing multiple display terminals separately, and improving the monitoring efficiency and accuracy of salvage operations.

[0047] In a preferred embodiment, the underwater salvage positioning and monitoring method includes the following steps in step S2: Step S2-1: Real-time acquisition of the original ship positioning information of the main construction vessel using a GNSS receiver, including longitude coordinates, latitude coordinates, and geodetic coordinates; Step S2-2: Real-time acquisition of the real-time attitude data of the main construction vessel using an attitude sensor and inertial navigation system, including roll angle, pitch angle, and heading angle; Step S2-3: Calculation of the offset of the GNSS receiving antenna installation position relative to the ship reference point based on the real-time attitude data, and conversion of the antenna position offset caused by the ship attitude change into a displacement in the geodetic coordinate system using a coordinate rotation algorithm; Step S2-4: Subtraction of the displacement calculated in step S2-3 from the original ship positioning information to obtain the corrected ship positioning information.

[0048] In current underwater salvage operations, the positioning of construction vessels typically relies on the direct output of position information from a GNSS receiver. The GNSS receiver, mounted above the ship's bridge, receives satellite signals and differential correction signals, outputting the ship's longitude, latitude, and geodetic coordinates in real time. Operators use these coordinates to determine the ship's position in the work area and direct it to the designated work point. While this positioning method meets general accuracy requirements under static or low-dynamic conditions, it suffers from the following shortcomings in actual salvage operations: The ship continuously experiences roll, pitch, and heading changes due to wind, waves, and currents; the GNSS antenna shifts spatially with the ship's movement, resulting in an inconsistency between the antenna position and the ship's reference point; the raw positioning information output by the GNSS receiver is the coordinates of the antenna position, not the coordinates of the ship's reference point, making it impossible to accurately reflect the ship's actual working position; and positioning errors caused by changes in ship attitude are not effectively corrected, affecting the positioning accuracy of subsequent grab position calculations and salvage operations.

[0049] The present invention performs dynamic error correction on ship positioning information through the following specific embodiments.

[0050] A fixed ship reference point is established on the main construction vessel. This reference point is typically chosen from the ship's center of gravity or a benchmark point marked on the ship's design drawings, serving as a unified reference for ship positioning. An attitude sensor and an inertial navigation system (INS) are installed near the reference point. The attitude sensor measures the ship's roll and pitch angles in real time, while the INS measures the ship's heading angle. A GNSS receiver is installed in an unobstructed area above the ship's bridge to receive satellite signals and differential correction signals, acquiring the main construction vessel's raw ship positioning information in real time. This raw ship positioning information includes longitude, latitude, and geodetic coordinates, reflecting the spatial coordinates of the GNSS receiving antenna's location.

[0051] After the equipment is installed, initial parameter measurements are performed. The three-dimensional offset of the GNSS receiving antenna's installation position relative to the ship's reference point is measured. This offset includes the longitudinal offset along the bow direction, the lateral offset along the starboard side direction, and the vertical offset perpendicular to the deck direction. The measured offsets are recorded and saved as fixed parameters for subsequent calculations.

[0052] Once the salvage operation commenced, the system entered real-time data acquisition and processing mode. The GNSS receiver continuously outputs the raw vessel positioning information of the main construction vessel at a fixed frequency, including the geodetic coordinates of the antenna position at the current moment. The attitude sensor and inertial navigation system continuously output real-time attitude data of the main construction vessel at a higher frequency, including the vessel's roll, pitch, and bow angles at the current moment. This attitude data reflects the vessel's degree of tilt and heading relative to the horizontal plane.

[0053] In the data processing stage, the antenna position offset caused by changes in ship attitude is calculated based on real-time attitude data. When the ship rolls, the antenna undergoes lateral displacement around the ship's longitudinal axis; when the ship pitches, the antenna undergoes longitudinal displacement around the ship's transverse axis; when the ship's heading changes, the antenna undergoes a horizontal position change around the ship's vertical axis. These displacement components caused by attitude changes are comprehensively calculated to obtain the total offset of the antenna position relative to the ship's reference point. This offset is then converted from an angular quantity in the geodetic coordinate system to a length quantity, yielding the antenna position offset in the geodetic coordinate system.

[0054] Finally, the calculated displacement is subtracted from the original ship positioning information output by the GNSS receiver. Specifically, the longitude component of the displacement is subtracted from the longitude coordinates, the latitude component is subtracted from the latitude coordinates, and the elevation component is subtracted from the geodetic coordinates. The new coordinates obtained after these subtractions are the corrected ship positioning information. This corrected ship positioning information reflects the actual position of the ship's reference point in the geodetic coordinate system, not the position of the GNSS receiving antenna.

[0055] Through the above-described embodiments, this invention eliminates the impact of ship attitude changes on GNSS positioning accuracy. Compared with existing technologies, this invention no longer directly uses the raw positioning information output by the GNSS receiver as the ship's position. Instead, it fully considers the antenna position offset caused by changes in ship roll, pitch, and heading, and dynamically corrects the raw positioning information using real-time attitude data. The corrected ship positioning information can more accurately reflect the actual position of the ship's reference point, providing a more reliable ship positioning benchmark for salvage operations. This method requires no additional hardware equipment; it only corrects existing GNSS output through attitude data, and has the advantages of being simple to implement and having significant effects.

[0056] In a preferred embodiment, the underwater salvage positioning and monitoring method includes the following specific implementation process for step S3: Step S3-1: Transmitting an acoustic interrogation signal to an ultra-short baseline beacon via an ultra-short baseline transducer and receiving a response signal transmitted by the beacon; calculating the slant range between the transducer and the beacon based on the time difference between the transmission of the interrogation signal and the reception of the response signal; calculating the horizontal and vertical azimuth angles of the beacon relative to the transducer based on the phase difference of the received response signal; and then determining the position of the grab bucket relative to the transducer using polar coordinate transformation based on the slant range, horizontal azimuth angle, and vertical azimuth angle. Step S3-2: Based on the pre-measured offset of the transducer installation position relative to the ship reference point, convert the three-dimensional coordinates of the grab bucket in the local coordinate system of the transducer to the three-dimensional coordinates in the ship coordinate system; Step S3-3: Combine real-time attitude data, and use a coordinate rotation algorithm to convert the three-dimensional coordinates of the grab bucket in the ship coordinate system to the three-dimensional coordinates in the station-centered coordinate system; Step S3-4: Combine the original ship positioning information, and use a coordinate translation algorithm to convert the three-dimensional coordinates of the grab bucket in the station-centered coordinate system to the grab bucket position information in the geodetic coordinate system.

[0057] Underwater positioning of grab buckets typically employs an ultra-short baseline (USBR) positioning system. This system consists of a transducer mounted on the construction vessel and a beacon mounted on the grab bucket. The relative position of the grab bucket is determined by measuring the slant distance and azimuth angle between the transducer and the beacon. Operators use the grab bucket coordinates output by the transducer to determine its underwater position and direct crane operators to adjust the grab bucket's position. This positioning method has the following shortcomings in practical applications: the grab bucket coordinates output by the transducer are relative to the transducer's local coordinate system, not the absolute position in the geodetic coordinate system; the vessel's continuous movement under wind and waves causes the transducer to sway and shift with the vessel, making it impossible to directly convert the grab bucket's relative coordinates to absolute coordinates; the grab bucket's position information is not correlated with the vessel's attitude and position, failing to accurately reflect the grab bucket's true position in the geodetic coordinate system; and the lack of a unified transformation framework during multi-level coordinate transformations affects the accuracy and reliability of grab bucket positioning.

[0058] The present invention achieves accurate acquisition of grab bucket position information and coordinate transformation through the following specific embodiments.

[0059] A reference point is established on the main construction vessel, serving as a unified benchmark for vessel positioning and coordinate transformation. An attitude sensor and inertial navigation system are installed near the reference point to measure the vessel's roll, pitch, and heading angles in real time. A GNSS receiver is installed above the bridge to acquire the main construction vessel's raw positioning information in real time. An ultra-short baseline transducer is installed on the side of the main construction vessel near the crane's operating area, in a location chosen away from the vessel's engine and propeller to minimize mechanical noise interference with acoustic communications. An ultra-short baseline beacon is installed above the grab bucket, ensuring it remains vertically upward and is not obstructed by the grab bucket's wire rope.

[0060] Initial parameter measurements were performed after equipment installation. The three-dimensional offset of the ultra-short baseline transducer's installation position relative to the ship's reference point was measured. This offset included the longitudinal offset along the bow direction, the lateral offset along the starboard direction, and the vertical offset perpendicular to the deck direction. The measured offsets were recorded and saved as fixed parameters for subsequent coordinate transformations.

[0061] After the salvage operation begins, the system enters real-time data acquisition and coordinate transformation mode. The ultra-short baseline transducer transmits acoustic interrogation signals to the beacon on the underwater grab bucket at a set interval. Upon receiving the interrogation signal, the beacon immediately transmits a response signal back to the transducer. The transducer records the transmission time of the interrogation signal and the reception time of the response signal, calculating the time difference between the two. Based on the speed of sound propagation in water and the time difference, the slant distance between the transducer and the beacon is calculated. Simultaneously, the transducer's receiving array detects the phase distribution of the response signal and calculates the horizontal and vertical azimuth angles of the beacon relative to the transducer based on the phase difference of the signals received by each array element. The horizontal azimuth angle represents the projection direction of the beacon onto the transducer's horizontal plane, and the vertical azimuth angle represents the pitch angle of the beacon relative to the transducer's horizontal plane.

[0062] Based on the calculated slant range, horizontal azimuth, and vertical azimuth, the three-dimensional coordinates of the grab bucket in the transducer's local coordinate system are determined using a polar coordinate transformation method. In the transducer's local coordinate system, with the transducer center as the origin and the transducer's own installation direction as the coordinate axis direction, the three-dimensional coordinates of the grab bucket are determined by the horizontal and vertical projections of the slant range.

[0063] After obtaining the three-dimensional coordinates of the grab bucket in the transducer's local coordinate system, the first coordinate transformation is performed. Based on the pre-measured offset of the transducer's installation position relative to the ship's reference point, the grab bucket's coordinates are transformed from the transducer's local coordinate system to the ship's coordinate system. The ship's coordinate system has the ship's reference point as its origin, with the X-axis pointing towards the bow, the Y-axis pointing towards the starboard side, and the Z-axis perpendicular to the deck and upwards. This coordinate system moves synchronously with the ship's motion and serves as an intermediate bridge connecting the transducer's local coordinate system and the geodetic coordinate system.

[0064] After the first coordinate transformation, a second coordinate transformation is performed. Real-time ship attitude data from the attitude sensor and inertial navigation system is acquired, including the current roll, pitch, and heading angles. A coordinate rotation algorithm is used to convert the grab's three-dimensional coordinates in the ship's hull coordinate system to a stationary coordinate system based on the ship's attitude data. The stationary coordinate system uses north as the X-axis, east as the Y-axis, and vertical as the Z-axis, with the origin located at the ship's reference point. This transformation process eliminates the influence of ship roll, pitch, and heading changes on the grab's coordinates, decoupling the grab's coordinates from the ship's attitude.

[0065] Specifically, step S3-3: Combining real-time attitude data, the three-dimensional coordinates of the grab bucket in the ship's coordinate system are converted to three-dimensional coordinates in the stationary coordinate system using a coordinate rotation algorithm. This coordinate system has the ship's reference point as its origin, with the X-axis pointing due east, the Y-axis pointing due north, and the Z-axis perpendicular to the reference ellipsoid pointing towards the zenith. The coordinate rotation algorithm comprehensively considers the ship's roll, pitch, and heading angles to construct a rotation matrix R from the ship's coordinate system to the stationary coordinate system. The above coordinate transformation formula can be implemented using the Euler angle rotation matrix method, which is well-known in the field.

[0066] After the second coordinate transformation, a third coordinate transformation is performed. The raw ship positioning information, including the geodetic coordinates of a point near the ship's reference point, is acquired from the real-time output of the GNSS receiver. A coordinate translation algorithm is used to convert the grab's three-dimensional coordinates in the station-centered coordinate system to its position in the geodetic coordinate system, based on the ship's position. The geodetic coordinate system uses national or internationally recognized coordinate systems such as CGCS2000 or WGS84, expressing the absolute position of a point using latitude, longitude, and elevation. This transformation process translates the grab's coordinates from the ship's reference point to the origin of the geodetic coordinate system, obtaining the grab's absolute position coordinates in the geodetic coordinate system.

[0067] Specifically, steps S3-4: Combining the original ship positioning information, the three-dimensional coordinates of the grab bucket in the station-centered coordinate system are converted to the grab bucket position information in the geodetic coordinate system (CGCS2000 or WGS84 ellipsoidal coordinate system) using a coordinate transformation algorithm. This conversion process consists of two steps: Step 1: Conversion from station-centered coordinates to spatial rectangular coordinates. Using the geodetic coordinates of the ship's reference point as a benchmark, the station-centered coordinates of the grab bucket are converted to spatial rectangular coordinates according to the conversion relationship between the station-centered coordinate system and the spatial rectangular coordinate system. Step 2: Conversion from spatial rectangular coordinates to geodetic coordinates. Using an iterative method or a direct solution method, the spatial rectangular coordinates are converted to geodetic coordinates, obtaining the longitude B, latitude L, and elevation H of the grab bucket in the geodetic coordinate system. Through the above complete coordinate transformation process, the grab bucket position information is accurately converted from the transducer local coordinate system through the ship coordinate system and the station-centered coordinate system to the geodetic coordinate system, providing a unified spatial reference for subsequent multi-source data fusion. This conversion process is a well-known coordinate transformation method in this field. Through the three coordinate transformations described above, the grab's position information is gradually transformed from the transducer's local coordinate system to the geodetic coordinate system. The parameters for each transformation are derived from pre-measured or real-time collected data, including transducer installation position offset, real-time ship attitude data, and real-time ship positioning information. The transformation process fully considers multiple factors such as transducer installation position, ship attitude changes, and ship position to ensure the accuracy and reliability of the grab's position information.

[0068] Compared to existing technologies, this invention no longer uses the relative coordinates of the grab bucket directly output by the ultra-short baseline transducer as the final positioning result. Instead, it establishes a three-level transformation framework from the transducer's local coordinate system to the geodetic coordinate system. The transformation from the transducer's local coordinate system to the ship's coordinate system eliminates the influence of the transducer's installation position; the transformation from the ship's coordinate system to the station-center coordinate system eliminates the influence of changes in the ship's attitude; and the transformation from the station-center coordinate system to the geodetic coordinate system introduces the ship's absolute position information. This multi-level coordinate transformation method places the grab bucket's position information and the ship's positioning information in the same coordinate system, providing a unified spatial reference for subsequent multi-source data fusion and display. Operators can directly obtain the grab bucket's absolute position in the geodetic coordinate system without manual calculation based on the transducer's relative coordinates and the ship's position, thus improving the accuracy of grab bucket positioning and operational efficiency.

[0069] In a preferred embodiment, the underwater salvage positioning and monitoring method, specifically implementing real-time error correction of the two-dimensional acoustic image and the three-dimensional image in step S6 to obtain the corrected two-dimensional acoustic image and the corrected three-dimensional image, includes the following steps: Step S6-1: Real-time measurement of sound velocity values ​​at different depths in the working water area using a sound velocity profiler to generate sound velocity profile data; Step S6-2: Based on the sound velocity profile data, calculating the actual sound velocity of the sound wave on the propagation path between the two-dimensional scanning sonar transducer and the physical positions corresponding to each pixel point of the two-dimensional acoustic image, and the actual sound velocity of the sound wave on the propagation path between the three-dimensional real-time sonar transducer and the physical positions corresponding to each pixel point of the three-dimensional image, respectively replacing the preset nominal sound velocities of the two-dimensional scanning sonar and the three-dimensional real-time sonar. Step S6-3: Substitute the actual sound velocity corresponding to the two-dimensional scanning sonar into the sonar ranging formula to recalculate the sound wave propagation distance corresponding to each pixel in the two-dimensional acoustic image, and substitute the actual sound velocity corresponding to the three-dimensional real-time sonar into the sonar ranging formula to recalculate the sound wave propagation distance corresponding to each pixel in the three-dimensional image; Step S6-4: Update the position coordinates of each pixel in the two-dimensional acoustic image according to the recalculated sound wave propagation distance corresponding to each pixel in the two-dimensional acoustic image, and update the position coordinates of each pixel in the three-dimensional image according to the recalculated sound wave propagation distance corresponding to each pixel in the three-dimensional image; Step S6-5: Output the updated two-dimensional acoustic image and the updated three-dimensional image as the corrected two-dimensional acoustic image and the corrected three-dimensional image.

[0070] Sonar equipment is the primary tool for acquiring underwater environmental information. Two-dimensional scanning sonar acquires two-dimensional acoustic images of the area surrounding a shipwreck through fan-shaped scanning, while three-dimensional real-time sonar forms a three-dimensional image of the grab bucket and the shipwreck target through dense beams. These sonar devices are preset with fixed nominal sound speeds at the factory, typically using typical sound speed values ​​in freshwater or seawater, such as 1500 m / s. The equipment calculates the target distance based on the time difference between the transmitted sound wave and the received echo, and then generates a sonar image. This method of preset sound speed has the following shortcomings in practical applications: the sound speed in seawater varies due to temperature, salinity, and hydrostatic pressure, and there are differences in sound speeds in different waters and at different depths; the sound speed in the actual operating water area is inconsistent with the preset nominal sound speed of the equipment, leading to errors in the calculation of sound wave propagation distance; distance calculation errors cause geometric distortion of the shape and position of the target in the sonar image, resulting in a deviation between the actual position of the target and the position displayed in the image; this deviation is exacerbated with increasing water depth and scanning angle, affecting the accurate judgment of the relative position of the grab bucket and the shipwreck target.

[0071] The present invention performs real-time error correction on sonar images through the following specific embodiments.

[0072] Sound velocity profilers were deployed near the two-dimensional scanning sonar transducer and the three-dimensional real-time sonar transducer. A sound velocity profiler is a specialized device capable of measuring sound velocity at different depths. It uses the ringing method to directly measure the propagation time of sound waves over a fixed distance, and then calculates the sound velocity at that depth. The sound velocity profiler was lowered into the water, placing it at a similar depth to the sonar transducer, ensuring that the measured sound velocity data accurately reflects the acoustic environment of the sonar operating area.

[0073] Once the salvage operation began, the sound velocity profiler entered continuous measurement mode. The equipment measured the sound velocity at different depths in the operating area at a fixed frequency in real time, starting near the surface and proceeding downwards until the maximum depth detectable by sonar was reached. The corresponding sound velocity value was recorded at each measurement depth, forming a set of depth-sound velocity correspondences, i.e., sound velocity profile data. This data reflects the true distribution pattern of sound velocity with depth in the operating area.

[0074] After obtaining the sound velocity profile data, the actual sound velocity is calculated. For two-dimensional scanning sonar, considering the transducer installation depth and the water depth range covered by the scan, the actual sound velocity along the propagation path from the transducer to the corresponding physical location of each pixel in the two-dimensional acoustic image is determined based on the sound velocity profile data. Since the sound wave propagation path may cross water layers of different depths, the sound velocity along the path is not a constant value and needs to be calculated comprehensively based on the sound velocity of each water layer along the path. For three-dimensional real-time sonar, similarly, based on the transducer installation depth and the spatial range covered by the beam, the actual sound velocity along the propagation path from the transducer to the corresponding physical location of each pixel in the three-dimensional image is calculated. The calculated actual sound velocities are then used to replace the preset nominal sound velocities in the two-dimensional scanning sonar and three-dimensional real-time sonar devices, respectively.

[0075] After replacing the actual sound speed with the actual sound speed, the sound wave propagation distance is recalculated. For two-dimensional scanning sonar, the calculated actual sound speed is substituted into the sonar ranging formula. The basic principle of sonar ranging is that the product of the round-trip time of the sound wave from the transducer to the target and back to the transducer and the sound speed is equal to twice the target distance. Based on the round-trip time and actual sound speed recorded in the original echo signal, the sound wave propagation distance corresponding to each pixel in the two-dimensional acoustic image is recalculated. For three-dimensional real-time sonar, the actual sound speed is also substituted into the ranging formula to recalculate the sound wave propagation distance corresponding to each pixel in the three-dimensional image.

[0076] After obtaining the recalculated acoustic wave propagation distance, the pixel position coordinates are updated. For 2D scanning sonar, the scanning angle corresponding to each pixel during the original imaging is recorded, including the horizontal azimuth and vertical elevation angles. Based on the recalculated acoustic wave propagation distance as the radial distance, with the transducer position as the origin, and following the direction indicated by the originally recorded scanning angles, the updated position coordinates of each pixel in the transducer's local coordinate system are determined. For 3D real-time sonar, the beam pointing angle and elevation angle corresponding to each pixel during the original imaging are recorded. Similarly, using the recalculated acoustic wave propagation distance as the radial distance, the updated position coordinates of each pixel are determined. The updated position coordinates replace the original position coordinates of the pixel, completing the image geometric correction.

[0077] Finally, the processed 2D and 3D acoustic images are output as corrected 2D acoustic images and corrected 3D images. In the corrected sonar images, the position coordinates of each pixel have been corrected according to the actual sound speed, and the shape and position of the shipwreck structure and grab bucket in the image are closer to the actual physical state.

[0078] Compared to existing technologies, this invention no longer relies on a fixed nominal sound velocity preset by the sonar equipment for imaging. Instead, it introduces a sound velocity profiler to measure the actual sound velocity distribution in the operating water area in real time, and corrects the sonar image pixel by pixel based on the measured sound velocity data. This method fully considers the objective law of sound velocity variation with depth, eliminating ranging errors caused by sound velocity differences. The geometric distortion of the corrected sonar image is effectively suppressed, and the shape and position of the target in the image more accurately reflect the actual underwater situation. Operators can determine the relative position of the grab bucket and the sunken ship target based on the corrected image, obtaining more reliable spatial position information and providing more accurate visual guidance for the grab bucket operation.

[0079] In a preferred embodiment, the underwater salvage positioning and monitoring method includes the following steps in step S7: Step S7-1: Establishing a unified geodetic coordinate system as the fusion reference, converting the corrected ship positioning information, grab bucket position information, corrected two-dimensional acoustic image, corrected three-dimensional image, and background image of the sunken ship and immersed tube position to the geodetic coordinate system; Step S7-2: Performing image registration on the corrected two-dimensional acoustic image and the corrected three-dimensional image, extracting feature points of the grab bucket and sunken ship targets in the image, and calculating the relative position of the grab bucket and sunken ship targets based on the coordinates of the feature points; Step S7-3: Spatially associating the corrected ship positioning information with the grab bucket position information to determine the grab bucket relative to the main construction vessel. The spatial location of the grab bucket is determined, and combined with the corrected ship positioning information, the grab bucket position information is converted to a unified geodetic coordinate system; Step S7-4: The corrected two-dimensional acoustic image and the corrected three-dimensional image registered in step S7-2, the grab bucket position information converted in step S7-3, and the pre-imported background map of the sunken ship and immersed tube position are superimposed according to the coordinate correspondence to generate a fused comprehensive positioning image; Step S7-5: In the fused comprehensive positioning image, the relative distance and azimuth angle between the grab bucket and the sunken ship target are marked in real time, and the fused comprehensive positioning image is transmitted to the crane control room display terminal in real time, so that the operator can direct the grab bucket grabbing operation according to the relative position of the grab bucket and the sunken ship target in the fused comprehensive positioning image.

[0080] The present invention achieves the fusion display of multi-source data through the following specific embodiments.

[0081] First, establish a unified geodetic coordinate system as the fusion benchmark. Choose either the CGCS2000 or WGS84 coordinate system as the unified geodetic coordinate system, specifying its reference ellipsoid parameters, central meridian longitude, projection method, and coordinate origin. Convert the corrected ship positioning information to the geodetic coordinate system. Since the corrected ship positioning information already exists in the geodetic coordinate system, only its coordinate system needs to be confirmed to be consistent with the unified coordinate system. Convert the grab bucket position information to the geodetic coordinate system. The grab bucket position information already exists in the geodetic coordinate system after multiple coordinate transformations; similarly, only coordinate system consistency needs to be confirmed. Convert the corrected 2D acoustic images and corrected 3D images to the geodetic coordinate system. The sonar images originally exist in their respective transducer local coordinate systems; coordinate transformation is required based on transducer installation location, ship attitude, and ship position to ensure that the geodetic coordinates of each pixel in the image can be calculated. Convert the pre-imported background images of the sunken ship and immersed tunnel locations to the geodetic coordinate system. These background images are usually drawn based on historical measurement data; their coordinate system needs to be confirmed to be consistent with the unified coordinate system.

[0082] After coordinate unification, image registration is performed. Feature information is extracted from the corrected 2D acoustic image, including the edge contours of the shipwreck structure, the geometric features of the bow and stern, the turning points of the hull, and the outline shape of the grab bucket. Feature information is extracted from the corrected 3D image, including the point cloud distribution of the shipwreck structure, point cloud clusters at the bow and stern, dense point cloud regions at the hull turning points, and the point cloud centroid and tooth tip clusters of the grab bucket. The features extracted from the 2D image are matched with those extracted from the 3D image to find corresponding feature point pairs. Based on the matched feature point pairs, rotation and translation parameters between the 2D and 3D images are calculated. Based on the calculated rotation and translation parameters, the corrected 2D acoustic image and the corrected 3D image are aligned to a unified geodetic coordinate system, ensuring that the same physical location in both images corresponds to the same coordinate value.

[0083] During image registration, spatial correlation is performed between the grab bucket's position information and the vessel's positioning information. The corrected grab bucket position information is compared with the corrected vessel positioning information to determine the spatial positional relationship of the grab bucket relative to the main construction vessel. This relative positional relationship includes the horizontal distance, vertical depth, and bearing of the grab bucket and the vessel. Combined with the corrected vessel positioning information, the grab bucket's position information is finally confirmed in a unified geodetic coordinate system, ensuring that the grab bucket's position is consistent with the grab bucket's position displayed in the sonar image.

[0084] After the above processing, multi-source data overlay and fusion are performed. The registered and corrected 2D acoustic image, the registered and corrected 3D image, the converted grab bucket position information, and the pre-imported background map of the sunken ship and immersed tunnel locations are overlaid according to their coordinate correspondence. During the overlay process, a unified geodetic coordinate system is used as a reference to combine data from different sources within the same spatial area. The 2D acoustic image provides environmental information about the area surrounding the sunken ship, the 3D image provides the three-dimensional morphology information of the grab bucket and the sunken ship target, the grab bucket position information provides the precise coordinates of the grab bucket, and the background map provides the location of fixed structures such as the immersed tunnel. Through overlay and fusion, a comprehensive positioning image containing multiple types of information is generated.

[0085] In actual operation, the USBL grab positioning information (high-precision point coordinates) obtained in step S3 may deviate from the position of the grab point cloud in the 3D sonar image generated in step S5. When this deviation occurs, the following fusion strategy is adopted: (1) USBL positioning information is used as the reference for the actual position of the grab bucket. After multi-level coordinate transformation, the accuracy of USBL positioning is not affected by factors such as water turbidity and sonar beam angle, and its reliability is higher.

[0086] (2) During the overlay and fusion process, the USBL positioning markers (such as crosshairs) are first displayed in the fused image according to their geodetic coordinates. Then, the grab point cloud in the 3D sonar image is translated as a whole so that its geometric center coincides with the USBL positioning markers. This translation amount is combined with the translation parameters calculated during the 3D image registration process.

[0087] (3) If the deviation exceeds the preset threshold (e.g., 0.5 meters), the system will issue a prompt. The operator can check whether the USBL beacon is blocked, whether the sonar beam is abnormal, etc., and choose whether to use the USBL positioning as the standard for forced correction.

[0088] The above strategy ensures the accuracy and consistency of the grab position information in the fused image.

[0089] After generating the integrated positioning image, it is annotated and transmitted in real time. The relative distance and azimuth angle between the grab bucket and the sunken ship target are calculated and annotated in real time within the integrated positioning image. The relative distance refers to the straight-line distance between the geometric center of the grab bucket and the feature point of the sunken ship target, and the azimuth angle refers to the horizontal angle of the grab bucket relative to the sunken ship target. The annotated integrated positioning image is transmitted in real time to the display terminal in the crane control room. Crane operators can observe the integrated positioning image through the display terminal, simultaneously seeing the ship's position, the grab bucket's position, the sunken ship structure, the immersed tube position, and the relative distance and azimuth angle between the grab bucket and the sunken ship target. Based on the information in the image, operators can directly determine the direction and distance the grab bucket should move and direct the grab bucket to grab the sunken ship target.

[0090] It should be noted that in the fused image, the grab position information obtained in step S3 is marked with an independent positioning marker (such as a cursor, crosshair, or highlight), which is displayed together with the existing grab point cloud outline in the 3D image. This positioning marker serves as a precise position reference for the grab, assisting the operator in determining the actual position of the grab and corroborating it with the grab outline in the sonar image.

[0091] In step S7-1, the specific implementation method for converting the corrected two-dimensional acoustic image and the corrected three-dimensional image to the geodetic coordinate system is as follows: For the corrected two-dimensional acoustic image, the position coordinates of each pixel have been updated to the local coordinate system of the two-dimensional scanning sonar transducer in step S6-4. To convert the coordinates in this local coordinate system to a unified geodetic coordinate system, the following coordinate transformations are required: (1) Based on the pre-measured three-dimensional offset Δ of the two-dimensional scanning sonar transducer installation position relative to the ship reference point x 2D Δ y 2D Δ z 2D The coordinates of the pixel in the local coordinate system of the transducer ( x T , y T , z T Convert to coordinates in the ship's hull coordinate system. x S , y S , z S The transformation relationship is a translation transformation: (2) Combine the real-time attitude data (roll angle) of the main construction vessel ϕ Pitch angle i Bow angle ψ Using the same coordinate rotation algorithm as in step S3-3, the coordinates of the pixels in the ship's coordinate system are converted to those in the station-centered coordinate system. x E , y N , z U The station-centered coordinate system has the ship's reference point as its origin, with the X-axis pointing due east, the Y-axis pointing due north, and the Z-axis pointing to the zenith. The rotation formula is: Where the rotation matrix R( ϕ , i , ψ It is obtained by multiplying the three basic rotation matrices in sequence: The basic rotation matrices are: The specific form of the aforementioned rotation matrix is ​​a technique known in the art.

[0092] (3) Combine the corrected ship positioning information (i.e., the longitude of the ship reference point in the geodetic coordinate system) B 0. Latitude L 0. Elevation H 0), convert the station center coordinates to spatial rectangular coordinates ( X , Y , Z ), then convert to geodetic coordinates ( B , L , H The conversion process is exactly the same as steps S3-4, and will not be repeated here. Finally, the coordinates of each pixel in the two-dimensional acoustic image in a unified geodetic coordinate system are obtained.

[0093] For the corrected 3D image, the coordinate transformation of each pixel is performed using the same method, requiring only the installation offset (Δ) of the 3D real-time sonar transducer. x 3D ,Δ y 3D ,Δ z 3D This replaces the offset of the two-dimensional sonar. Through the above transformation, all pixels in the two-dimensional acoustic image and the three-dimensional image are assigned coordinates in a unified geodetic coordinate system, thereby achieving superposition with ship positioning information and grab position information under the same reference.

[0094] This invention unifies all data into a single coordinate system for fusion display. By establishing a unified coordinate system, image registration, spatial correlation, and overlay fusion, it integrates ship positioning information, grab bucket position information, 2D acoustic images, 3D images, and the sunken ship background image into a comprehensive positioning image. Operators only need to observe one display terminal to obtain all necessary information, eliminating the need to switch between multiple terminals and perform manual calculations. The comprehensive positioning image intuitively displays the relative positional relationship between the grab bucket and the sunken ship target, and marks the distance and azimuth angle in real time, providing clear and accurate visual guidance for grab bucket operations. This method improves information acquisition efficiency, reduces the judgment burden on operators, and enhances the controllability and accuracy of salvage operations.

[0095] In a preferred embodiment, the underwater salvage positioning and monitoring method includes the following specific implementation process for step S6-2: deploying sound velocity profilers near the two-dimensional scanning sonar and the three-dimensional real-time sonar to measure the sound velocity values ​​at different depths in the operating water area in real time, generating sound velocity profile data; calculating the average sound velocity along the propagation path of the sound wave from the transducer of the two-dimensional scanning sonar to the physical location corresponding to each pixel point of the two-dimensional acoustic image based on the sound velocity profile data, and using this as the actual sound velocity of the two-dimensional scanning sonar; calculating the average sound velocity along the propagation path of the sound wave from the transducer of the three-dimensional real-time sonar to the physical location corresponding to each pixel point of the three-dimensional image based on the sound velocity profile data, and using this as the actual sound velocity of the three-dimensional real-time sonar; and replacing the preset nominal sound velocities of the two-dimensional scanning sonar and the three-dimensional real-time sonar with the calculated actual sound velocities of the two-dimensional scanning sonar and the three-dimensional real-time sonar, respectively.

[0096] Sonar equipment typically uses a fixed nominal sound velocity value when calculating target distance. Two-dimensional scanning sonar and three-dimensional real-time sonar have a preset nominal sound velocity at the factory, generally 1500 m / s as the standard value. The equipment calculates the target distance based on the time difference between the transmitted sound wave and the received echo, and then generates a sonar image. This fixed sound velocity ranging method has the following shortcomings in practical applications: the sound velocity in seawater is not a constant value, but varies with temperature, salinity, and hydrostatic pressure; there are significant differences in sound velocity in different waters, seasons, and depths; when the nominal sound velocity does not match the actual sound velocity, the distance calculation results will have systematic errors; the shape and position of the target in the sonar image will be geometrically distorted, affecting the accurate identification of underwater structures; operators find it difficult to judge the degree of image distortion and cannot adjust the sound velocity setting according to environmental changes.

[0097] The present invention realizes the conversion of sound velocity profile data into actual sound velocity parameters through the following specific embodiments.

[0098] Sound velocity profilers are deployed near the two-dimensional scanning sonar transducer and the three-dimensional real-time sonar transducer. A sound velocity profiler is a specialized instrument capable of measuring the velocity of sound at different depths. It uses the ringing method to directly measure the propagation time of sound waves over a fixed distance, and then calculates the velocity of sound at that depth. The sound velocity profiler is lowered into the water so that its measurement depth range covers the sonar detection area. The placement of the sound velocity profiler should be as close as possible to the sonar transducer to ensure that the measured velocity data represents the acoustic environment along the actual propagation path of the sound waves.

[0099] Once the salvage operation begins, the sound velocity profiler enters continuous measurement mode. The equipment starts measuring point by point downwards from near the water surface, recording a sound velocity value at regular intervals until the maximum sonar detection depth is reached. During the measurement process, the sound velocity profiler simultaneously records the depth information and corresponding sound velocity value at each measurement point, forming a set of depth-sound velocity correlation data, i.e., sound velocity profile data. This data reflects the true distribution pattern of sound velocity with depth in the operating water area in the form of tables or curves. The update frequency of the sound velocity profile data can be set according to operational needs; the measurement frequency can be increased in areas with rapidly changing hydrological conditions, and appropriately decreased in areas with relatively stable hydrological conditions.

[0100] After obtaining the sound velocity profile data, the actual sound velocity is calculated. For a two-dimensional scanning sonar, the installation depth of its transducer and the water depth range corresponding to the physical location of each pixel in the two-dimensional acoustic image are first determined. Each pixel of the two-dimensional scanning sonar corresponds to an echo signal in a certain direction underwater, and this echo signal comes from physical locations at different distances and depths. Based on the transducer installation depth and the target depth corresponding to the pixel, the range of water layers traversed by the propagation path of the sound wave from the transducer to that physical location is determined. The sound velocity values ​​of each water layer along this propagation path are extracted from the sound velocity profile data, and these sound velocity values ​​are combined to obtain the average sound velocity along the entire propagation path. This average sound velocity reflects the overall speed level of the sound wave propagating along this path and is used as the actual sound velocity of the two-dimensional scanning sonar.

[0101] For 3D real-time sonar, a similar method is used to calculate the actual sound velocity. Each pixel in a 3D real-time sonar corresponds to an echo signal in a specific beam direction, originating from a physical location at a specific direction and distance. Based on the installation depth of the 3D real-time sonar transducer and the target depth corresponding to the pixel, the range of water layers traversed by the sound wave propagation path is determined. The sound velocity values ​​of each water layer along this path are extracted from the sound velocity profile data, and the average sound velocity along the entire path is calculated as the actual sound velocity of the 3D real-time sonar. Since the beam coverage of a 3D real-time sonar is wide, the propagation paths corresponding to pixels in different directions and at different distances may differ. Therefore, it is necessary to calculate the actual sound velocity for each pixel or each beam direction separately, forming sound velocity correction parameters that match the actual sound velocity distribution.

[0102] After calculating the actual sound velocity, the sound velocity parameters are replaced. The calculated actual sound velocity from the 2D scanning sonar is input into the 2D scanning sonar device, replacing the original nominal sound velocity in the device. Similarly, the calculated actual sound velocity from the 3D real-time sonar is input into the 3D real-time sonar device, replacing the original nominal sound velocity in the device. The replacement of the sound velocity parameters can be done manually in the device settings interface or automatically through the data interface. The replaced actual sound velocity will be used for subsequent sound wave propagation distance calculations and image generation.

[0103] Compared to existing technologies, this invention no longer relies on a fixed nominal sound velocity preset at the factory for sonar equipment. Instead, it uses a sound velocity profiler to measure the sound velocity distribution data of the operating water area in real time and calculates the average sound velocity along the sound wave propagation path as the actual sound velocity. This method converts sound velocity profile data into actual sound velocity parameters that can be directly used for sonar ranging correction, providing accurate basic data for subsequent recalculation of sound wave propagation distance. Through real-time measurement and calculation, the actual sound velocity can dynamically track changes in the sound velocity of the operating water area, ensuring that the sonar equipment always operates under sound velocity conditions that match the actual environment. Operators do not need to understand complex sound velocity variation patterns or manually adjust equipment parameters; the system automatically completes sound velocity measurement, calculation, and parameter substitution, improving the automation and accuracy of sonar image correction.

[0104] In a preferred embodiment, the underwater salvage positioning and monitoring method, specifically implementing step S6-4 of updating the position coordinates of each pixel in the two-dimensional acoustic image and the position coordinates of each pixel in the three-dimensional image, includes: for two-dimensional scanning sonar, acquiring the horizontal and vertical azimuth angles corresponding to each pixel in the two-dimensional acoustic image, using the recalculated sound wave propagation distance as the radial distance, taking the transducer position as the origin, and determining the updated position coordinates of each pixel in the local coordinate system of the transducer according to the directions indicated by the horizontal and vertical azimuth angles, and replacing the original position coordinates of the pixel with the updated position coordinates; for three-dimensional real-time sonar, acquiring the beam pointing angle and pitch angle corresponding to each pixel in the three-dimensional image, using the recalculated sound wave propagation distance as the radial distance, taking the transducer position as the origin, and determining the updated position coordinates of each pixel in the local coordinate system of the transducer according to the directions indicated by the beam pointing angle and pitch angle, and replacing the original position coordinates of the pixel with the updated position coordinates.

[0105] Sonar images are typically used directly for target identification and location determination after generation. Two-dimensional scanning sonar calculates the target distance based on the time difference between the emitted and received sound waves and a preset nominal sound velocity, combining this with the scanning angle to generate a two-dimensional acoustic image. Three-dimensional real-time sonar calculates the target distance in each beam direction based on the beam emission direction and time difference, generating a three-dimensional image. After obtaining the actual sound velocity using a sound velocity profiler and recalculating the sound wave propagation distance, the new distance value needs to be applied to image reconstruction. In existing technologies, the recalculated distance often directly replaces the distance value in the image, without considering how to reconstruct the pixel position coordinates based on the new distance value and the original angle information, resulting in incomplete updates to the image's geometric structure.

[0106] The present invention achieves the updating of pixel position coordinates in sonar images through the following specific embodiments.

[0107] After recalculating the sound wave propagation distance, the pixel position coordinates of the two-dimensional acoustic image are updated. For two-dimensional scanning sonar, the angle information corresponding to each pixel in the original two-dimensional acoustic image is first acquired. Two-dimensional scanning sonar acquires underwater images using a fan-shaped scanning method, and each pixel records its corresponding horizontal and vertical azimuth angles during imaging. The horizontal azimuth angle represents the horizontal deflection angle of the pixel relative to the transducer's horizontal direction, and the vertical azimuth angle represents the pitch angle of the pixel relative to the transducer's horizontal plane. This angle information is precisely controlled by the device's scanning mechanism during sonar scanning, exhibiting high stability and reliability. The horizontal and vertical azimuth angles of each pixel are extracted from the original image data as the direction reference for position updating.

[0108] After acquiring the angle information of each pixel, the position coordinates are updated by combining it with the recalculated sound wave propagation distance. The recalculated sound wave propagation distance is used as the radial distance, with the transducer position as the origin. The radial distance represents the straight-line distance from the transducer to the physical location corresponding to the pixel. Extending the radial distance from the origin along the directions indicated by the horizontal and vertical azimuth angles corresponding to the pixel, the resulting spatial point is the updated position of the pixel's physical location in the transducer's local coordinate system. This process is equivalent to re-determining the spatial position of the pixel based on the new distance value and a fixed angle value. This position update operation is performed point-by-point for all pixels in the two-dimensional scanning sonar image.

[0109] After the position calculation is completed, the original position coordinates of the pixel are replaced with the updated position coordinates. Each pixel in the original 2D acoustic image stores its original position coordinates, which are calculated based on the nominal sound speed. The newly calculated updated position coordinates are then written into the image data, replacing the original position coordinates. After replacing the positions of all pixels, the overall geometry of the 2D acoustic image is updated, and the new position of each point in the image reflects the true distance under actual sound speed conditions.

[0110] For 3D real-time sonar, a similar method is used to update the pixel position coordinates. 3D real-time sonar forms a 3D image through dense beamforming, with each pixel corresponding to an echo signal along a specific beam direction. The beam pointing angle and elevation angle corresponding to each pixel in the 3D image are obtained. The beam pointing angle represents the projection direction of the beam onto the horizontal plane, and the elevation angle represents the tilt angle of the beam relative to the horizontal plane. This angular information is precisely controlled by the beamforming system of the 3D real-time sonar and is the fundamental parameter for determining the beam direction.

[0111] After acquiring the angle information of each pixel, the recalculated sound wave propagation distance is used as the radial distance, with the transducer position as the origin. Following the direction indicated by the beam pointing angle and elevation angle corresponding to that pixel, a radial distance is extended from the origin along this direction. The resulting spatial point is the updated position of that pixel's physical location in the transducer's local coordinate system. This position update operation is performed point-by-point for all pixels in the 3D real-time sonar image.

[0112] After the position calculation is completed, the original position coordinates of the pixel are replaced with the updated position coordinates. Each pixel in the original 3D real-time sonar image stores its original position coordinates calculated based on the nominal sound speed. The newly calculated updated position coordinates are written into the image data, replacing the original position coordinates. After the positions of all pixels are replaced, the geometry of the 3D real-time sonar image is updated, and the new position of each point in the image reflects the true distance under actual sound speed conditions.

[0113] It should be noted that the pixel position coordinates updated in step S6-4 are coordinates located in the local coordinate system of their respective sonar transducers and have not yet been converted to the geodetic coordinate system.

[0114] Through the above implementation methods, this invention achieves a complete update of the pixel position coordinates in a sonar image after sound speed correction. Compared with existing technologies, this invention not only recalculates the sound wave propagation distance but also fully utilizes the original angle information recorded during sonar imaging. It combines the new distance value with the fixed angle value and determines the updated position coordinates of each pixel through spatial geometric reconstruction. This method preserves the scanning geometry of the sonar image, ensuring that the relative shape and positional relationship of targets in the image remain correct after correction. In the updated sonar image, the position coordinates of each pixel correspond to the true distance under actual sound speed conditions, effectively correcting the geometric distortion of the image and providing an accurate data foundation for subsequent multi-source data fusion and target location determination.

[0115] In a preferred embodiment, the underwater salvage positioning and monitoring method, specifically implementing the image registration process of the corrected two-dimensional acoustic image and the corrected three-dimensional image in step S7-2, includes: extracting the edge contour of the sunken ship structure and the contour of the grab bucket from the corrected two-dimensional acoustic image as two-dimensional registration features; extracting the point cloud of the sunken ship structure and the point cloud of the grab bucket from the corrected three-dimensional image as three-dimensional registration features; matching the two-dimensional registration features with the three-dimensional registration features, and calculating the rotation parameters and translation parameters between the two-dimensional acoustic image and the three-dimensional image; and aligning the corrected two-dimensional acoustic image and the corrected three-dimensional image to a unified geodetic coordinate system based on the rotation parameters and translation parameters.

[0116] In current underwater salvage operations, two-dimensional acoustic images and three-dimensional images are typically presented on separate display terminals. Images of the area surrounding the sunken ship acquired by two-dimensional scanning sonar are displayed in planar form, reflecting the outline and distribution of underwater structures; images of the grab and the sunken ship target acquired by three-dimensional real-time sonar are displayed in stereo form, reflecting the target's spatial shape and relative position. Operators need to observe these two types of images separately, relying on experience to correlate the contour information in the two-dimensional image with the spatial information in the three-dimensional image to determine the actual positional relationship between the grab and the sunken ship target. This separate display method has the following shortcomings: two-dimensional and three-dimensional images use different coordinate systems and reference bases, resulting in a lack of spatial correspondence between the images; the contour features in the two-dimensional image are difficult to directly correspond to the point cloud features in the three-dimensional image, leading to subjectivity and uncertainty in manual matching; the two types of images cannot be superimposed under the same spatial reference, making it difficult for operators to gain a comprehensive spatial understanding; and the image registration process relies on manual experience, resulting in low efficiency and difficulty in guaranteeing accuracy.

[0117] The present invention achieves automatic registration of the corrected two-dimensional acoustic image and the corrected three-dimensional image through the following specific embodiments.

[0118] Before image registration, it is necessary to establish the spatial transformation relationship between the 2D scanning sonar and the 3D real-time sonar, including the relative positional relationship and the relative orientation relationship, to ensure that the two images can be registered within the same spatial reference frame. Specific details are as follows: (a) Measure the installation parameters of the two sonar transducers in advance.

[0119] For a two-dimensional scanning sonar transducer, the following parameters are measured: (1) The three-dimensional position offset of the origin of the transducer's local coordinate system (i.e., the geometric center of the transducer) relative to the ship's reference point is denoted as (Δ). x 2D ,Δ y 2D ,Δ z 2D This offset is defined along three axes of the ship's coordinate system: Δ x 2D Δ is the offset along the bow direction. y 2D For the offset along the starboard direction, Δ z 2D This is the offset perpendicular to the deck upwards.

[0120] (2) Three attitude angles of the transducer local coordinate system relative to the ship's coordinate system: roll installation angle α 2D Twist installation angle β 2D Bow mounting angle c2D When the transducer is installed so that the three axes of its local coordinate system are parallel to the three axes of the ship's coordinate system, all three attitude angles are zero. If there are deviations due to installation limitations, actual measurements are required.

[0121] For 3D real-time sonar transducers, the same measurements were taken: (1) Three-dimensional position offset relative to the ship reference point (Δ x 3D ,Δ y 3D ,Δ z 3D (2) Attitude angles relative to the ship's coordinate system α 3D , β 3D , c 3D .

[0122] (ii) Calculate the relative position and attitude relationship between the two transducers.

[0123] Relative positional relationship: Taking the two-dimensional scanning sonar transducer as a reference, the positional offset of the three-dimensional real-time sonar transducer relative to the two-dimensional scanning sonar transducer is: Relative attitude relationship: The rotation matrix of the local coordinate system of the 3D real-time sonar transducer relative to the local coordinate system of the 2D scanning sonar transducer is: Where R( α , β , c () is determined by the roll angle α Pitch angle β Bow angle c The constructed rotation matrix has the same form as in step S3-3: (iii) Convert the two sonar images to the same spatial reference frame.

[0124] Using the aforementioned measurement parameters, each pixel in the two-dimensional acoustic image and the three-dimensional image is transformed from its respective transducer local coordinate system to a unified geodetic coordinate system. The specific transformation path is: transducer local coordinate system → hull coordinate system (rotation based on attitude angle, translation based on position offset) → stationary coordinate system (rotation based on real-time ship attitude) → geodetic coordinate system (translation based on geodetic coordinates of the ship reference point and ellipsoid transformation). The transformation formula from the transducer local coordinate system to the hull coordinate system is: In the formula ( x T , y T , z T ) represents the coordinates of the pixel in the local coordinate system of the transducer. x S , y S , z S (Δ) represents the transformed ship coordinates. x ,Δ y ,Δ z R( represents the position offset of the transducer) α , β , c ) is the rotation matrix constructed from the installation attitude angle of the transducer.

[0125] After the above transformation is completed, all pixels in the two-dimensional acoustic image and the three-dimensional image are in a unified geodetic coordinate system and have the same spatial reference datum, thus providing the necessary prerequisites for image registration in step S7-2.

[0126] In a preferred embodiment of the present invention, both sonar transducers are installed in a standard orientation, meaning that the three axes of their local coordinate system are parallel to the three axes of the ship's coordinate system. α 2D = β 2D = c 2D =0, α 3D = β 3D = c 3D =0, relative rotation matrix R relThe coordinates degenerate into an identity matrix. In this case, the relative attitude relationship between the local coordinate systems of the two transducers simplifies to a pure translation relationship, greatly simplifying coordinate transformation and eliminating the need for attitude rotation. This invention does not exclude the possibility of non-standard attitude installations, but standard attitude installations are preferred to reduce system complexity and computational load.

[0127] The complete description of the relative position and attitude relationship between the two sonar transducers ensures that the two-dimensional acoustic image and the three-dimensional image can be converted to the same spatial reference frame, so that the image registration in step S7-2 has a clear spatial reference.

[0128] After obtaining the corrected 2D and 3D acoustic images, feature extraction is performed first. For the corrected 2D acoustic image, image processing algorithms are used to extract the edge contours of the shipwreck structure and the grab bucket contour. The edge contours of the shipwreck structure include feature lines reflecting the geometry of the shipwreck, such as the hull outline, deck edge lines, and hull bend lines. The grab bucket contour includes feature lines reflecting the shape of the grab bucket, such as the overall shape of the grab bucket, the protruding parts of the grab bucket teeth, and the extension direction of the grab bucket arm. These contour features exist in the form of lines or edges and can characterize the projected shape and position of the target on the 2D plane.

[0129] For the corrected 3D image, a point cloud processing algorithm is used to extract the point cloud of the shipwreck structure and the grab bucket. The shipwreck structure point cloud includes a set of feature points reflecting the 3D shape of the shipwreck, such as dense point groups forming the hull surface, prominent point groups at the bow and stern, and point groups at the bends of the hull. The grab bucket point cloud includes a set of feature points reflecting the 3D shape of the grab bucket, such as point groups forming the grab bucket surface, point groups at the tips of the grab bucket teeth, and point groups in the grab bucket's center of gravity region. These point cloud features exist in the form of discrete points and can characterize the actual position and shape of the target in 3D space.

[0130] After feature extraction, feature matching is performed. The edge contours of the shipwreck structure extracted from the 2D acoustic image are matched with the point cloud of the shipwreck structure extracted from the 3D image. During the matching process, the correspondence between the 2D contours and the 3D point cloud is sought; for example, the bow contour line in the 2D image corresponds to the bow point group in the 3D point cloud, and the hull edge line in the 2D image corresponds to the hull point group in the 3D point cloud. Simultaneously, the grab bucket contour extracted from the 2D acoustic image is matched with the grab bucket point cloud extracted from the 3D image, seeking the correspondence between the 2D contours and the 3D point cloud; for example, the grab bucket tooth tip contour in the 2D image corresponds to the grab bucket tooth tip point group in the 3D point cloud.

[0131] Based on the matching correspondence, rotation and translation parameters between the 2D acoustic image and the 3D image are calculated. The rotation parameters describe the angle required to rotate the 2D image to align with the 3D image, including the rotation around the three coordinate axes. The translation parameters describe the displacement required to translate the 2D image to coincide with the 3D image, including the movement along the three coordinate directions. These parameters are calculated by analyzing the spatial differences between the matching features, reflecting the spatial transformation relationship between the two images.

[0132] After obtaining the rotation and translation parameters, image alignment is performed. The corrected 2D acoustic image is rotated according to the calculated rotation parameters to align its direction with the 3D image. Then, it is translated according to the calculated translation parameters to make its position coincide with the 3D image. After rotation and translation processing, the corrected 2D acoustic image and the corrected 3D image are aligned to a unified geodetic coordinate system. In the aligned image, each contour point in the 2D image has the same spatial coordinates as the corresponding point cloud point in the 3D image, achieving positional correspondence between the two images under the same spatial reference.

[0133] Compared to existing technologies, this invention eliminates the reliance on subjective correspondence between 2D and 3D images based on human experience. Instead, it achieves precise image registration through automatic feature extraction and matching. This method fully utilizes contour information from 2D acoustic images and point cloud information from 3D images to correspond and match features from different dimensions, calculating spatial transformation parameters between the images. The registered 2D and 3D images are aligned in the same coordinate system, providing a spatial basis for subsequent overlay and fusion. Operators no longer need to manually compare and calculate between the two images; they can directly observe the registered composite image to obtain more intuitive and accurate spatial location information. This method improves the automation level and registration accuracy of image processing, and enhances the reliability of multi-source data fusion.

[0134] In a preferred embodiment, in the underwater salvage positioning and monitoring method, in step S7-5, an early warning is issued when the relative distance between the grab bucket and the sunken ship target is less than a preset safety threshold.

[0135] In current underwater salvage operations, the process of grabbing a sunken ship relies primarily on the operator's visual judgment and experience. The crane operator observes images on a display terminal, visually estimating the distance between the grab and the sunken target, and judges based on personal experience when to slow down, stop, and initiate the grabbing operation. This manual judgment method has the following shortcomings in practical application: underwater images are affected by water turbidity and lighting conditions, leading to significant errors in visual distance estimation; prolonged screen viewing can cause visual fatigue for operators, reducing the accuracy of distance judgment over time; when the grab approaches the sunken target, operators need to simultaneously monitor the grab's position, the sunken ship's position, and the relative distance, resulting in a heavy information processing load; there is a lack of objective distance warning mechanisms, making it difficult to detect and avoid situations where the grab gets too close to non-target structures such as submerged tunnel sections; and differences in the experience levels of different operators lead to inconsistencies in the stability and safety of the grabbing operation.

[0136] The present invention achieves real-time monitoring and early warning of the relative distance between the grab bucket and the sunken ship target through the following specific embodiments.

[0137] After the fused integrated positioning image is generated, the system continuously calculates the relative distance between the grab bucket and the sunken ship target. The grab bucket's position information comes from the corrected grab bucket position information, which has been accurately determined in the geodetic coordinate system after multi-level coordinate transformation. The sunken ship target's position information comes from the sunken ship feature points extracted from the corrected 3D image, or from a pre-imported background image of the sunken ship's location. The system calculates the straight-line distance between the geometric center of the grab bucket and the feature points of the sunken ship target in real time, obtaining the real-time relative distance between the grab bucket and the sunken ship target. This distance is expressed in length units, objectively reflecting the proximity between the grab bucket and the target.

[0138] During the system initialization phase, a preset safety threshold is established. This threshold is determined based on a comprehensive consideration of factors such as the structural characteristics of the sunken ship, the size and specifications of the grab bucket, and the complexity of the underwater environment. For vulnerable structures such as submerged pipes within the sunken ship, a smaller safety threshold is set, requiring the grab bucket to maintain a greater safe distance. For robust structures within the sunken ship, a relatively larger safety threshold can be set, allowing the grab bucket to approach more closely. The safety threshold can be set and adjusted before operations to meet the safety requirements of different salvage tasks.

[0139] During the salvage operation, the system monitors the relative distance between the grab and the sunken ship in real time and compares this distance with a preset safety threshold. When the calculated relative distance is greater than the safety threshold, the system determines that the grab is within a safe range and continues normal operation. When the relative distance gradually decreases and approaches the safety threshold, the system enters an early warning preparation state. When the relative distance is less than the preset safety threshold, the system immediately triggers the early warning mechanism.

[0140] Warnings are issued in multiple forms. On the display terminal, the line connecting the grab bucket and the sunken ship target turns a striking red, or a flashing warning box appears around the grab bucket icon, or red warning text appears at the edge of the screen, alerting operators that the grab bucket is approaching the safety boundary. Simultaneously, the system can emit an alarm sound via audio equipment to attract the operator's attention. For operating equipment equipped with vibration functionality, warning information can also be transmitted via vibration. Multiple warning methods are used in parallel to ensure that operators can promptly perceive when the grab bucket approaches the safety threshold.

[0141] After receiving the warning, the operator adjusts the grab's direction and speed based on the relative position of the grab and the sunken ship target displayed in the integrated positioning image. If the grab gets too close to non-target structures such as the immersed tube, the operator can control the grab to pause its forward movement or move in the opposite direction to avoid a collision. If the grab approaches the intended target location, the operator can precisely control the grab's final landing point with the assistance of the warning prompt to complete the grab operation.

[0142] Compared to existing technologies, this invention no longer relies solely on the operator's visual judgment. Instead, the system automatically calculates the relative distance between the grab and the sunken ship target and compares it with a preset safety threshold. When the distance is less than the threshold, an automatic warning is issued. This method provides operators with an objective distance reference and timely warning information, reducing the impact of visual fatigue and experience differences on judgment accuracy. The warning prompts enable operators to take timely countermeasures when the grab approaches the safety boundary, avoiding collisions between the grab and non-target structures such as immersed tubes. This method enhances the safety of salvage operations and improves the accuracy and reliability of grab operation.

[0143] Although embodiments of the present invention have been disclosed above, they are not limited to the applications listed in the specification and embodiments. It can be applied to various fields suitable for the present invention. Other modifications can be readily implemented by those skilled in the art. Therefore, the present invention is not limited to the specific details without departing from the general concept defined by the claims and their equivalents.

Claims

1. A method for underwater salvage positioning and monitoring, characterized in that, Includes the following steps: Step S1: Install a GNSS receiver, attitude sensor and inertial navigation system on the main construction vessel, install an ultra-short baseline beacon on the grab bucket, and install an ultra-short baseline transducer on the side of the main construction vessel to complete the deployment of the positioning and monitoring system; Step S2: Obtain the original ship positioning information of the main construction vessel through the GNSS receiver, obtain the real-time attitude data of the main construction vessel through the attitude sensor and inertial navigation system, and perform dynamic error correction on the original ship positioning information based on the real-time attitude data to obtain the corrected ship positioning information. Step S3: Through communication between the ultra-short baseline transducer and the ultra-short baseline beacon, the grab bucket is tracked in real time to obtain the grab bucket's position information; Step S4: Lower the two-dimensional scanning sonar installed at the stern of the main construction vessel into the water to scan the area around the sunken ship and obtain a two-dimensional acoustic image of the area around the sunken ship. Step S5: By transmitting sound waves underwater and receiving echoes through a three-dimensional real-time sonar installed on the side of the main construction vessel, a three-dimensional image of the relative positional relationship between the grab bucket and the sunken ship target is generated. Step S6: Measure the sound velocity in water in real time using a sound velocity profiler, input the measured sound velocity data into the two-dimensional scanning sonar and the three-dimensional real-time sonar, and perform real-time error correction on the two-dimensional acoustic image and the three-dimensional image to obtain the corrected two-dimensional acoustic image and the corrected three-dimensional image. Step S7: Overlay and fuse the corrected ship positioning information, grab bucket position information, corrected two-dimensional acoustic image and corrected three-dimensional image, as well as the pre-imported background image of the sunken ship and immersed tube position, in the same coordinate system. Based on the fused information, determine the relative position of the grab bucket and the sunken ship target in real time, and command the grab bucket to grab the sunken ship target.

2. The underwater salvage positioning and monitoring method according to claim 1, characterized in that, The specific implementation process of step S2 includes: Step S2-1: Obtain the original vessel positioning information of the main construction vessel in real time through a GNSS receiver. The original vessel positioning information includes longitude coordinates, latitude coordinates, and geodetic coordinates. Step S2-2: Acquire real-time attitude data of the main construction vessel through the attitude sensor and inertial navigation system. The real-time attitude data includes roll angle, pitch angle and bow angle. Step S2-3: Calculate the offset of the GNSS receiving antenna installation position relative to the ship reference point based on real-time attitude data, and use a coordinate rotation algorithm to convert the antenna position offset caused by the ship attitude change into a displacement in the geodetic coordinate system. Step S2-4: Subtract the displacement calculated in step S2-3 from the original ship positioning information to obtain the corrected ship positioning information.

3. The underwater salvage positioning and monitoring method according to claim 1, characterized in that, The specific implementation process of step S3 includes: Step S3-1: Transmit an acoustic interrogation signal to the US. beacon through the US. transducer and receive the response signal transmitted by the beacon. Calculate the slant range between the transducer and the beacon based on the time difference between the transmission of the interrogation signal and the reception of the response signal. Calculate the horizontal and vertical azimuth angles of the beacon relative to the transducer based on the phase difference of the received response signal. Then, determine the three-dimensional coordinates of the grab bucket in the local coordinate system of the transducer through polar coordinate transformation based on the slant range, horizontal azimuth angle, and vertical azimuth angle. Step S3-2: Based on the pre-measured offset of the transducer installation position relative to the ship reference point, convert the three-dimensional coordinates of the grab bucket in the local coordinate system of the transducer into the three-dimensional coordinates in the ship coordinate system. Step S3-3: Combining real-time attitude data, the three-dimensional coordinates of the grab bucket in the ship coordinate system are converted into three-dimensional coordinates in the station center coordinate system through a coordinate rotation algorithm; Step S3-4: Combining the original ship positioning information, the three-dimensional coordinates of the grab bucket in the station center coordinate system are converted into the grab bucket position information in the geodetic coordinate system using a coordinate translation algorithm.

4. The underwater salvage positioning and monitoring method according to claim 1, characterized in that, The specific implementation process of performing real-time error correction on the two-dimensional acoustic image and the three-dimensional image in step S6 to obtain the corrected two-dimensional acoustic image and the corrected three-dimensional image includes: Step S6-1: Measure the sound velocity values ​​at different depths in the working water area in real time using a sound velocity profiler to generate sound velocity profile data; Step S6-2: Based on the sound velocity profile data, calculate the actual sound velocity of the sound wave on the propagation path between the two-dimensional scanning sonar transducer and the physical position corresponding to each pixel of the two-dimensional acoustic image, and the actual sound velocity of the sound wave on the propagation path between the three-dimensional real-time sonar transducer and the physical position corresponding to each pixel of the three-dimensional image, respectively, and replace the preset nominal sound velocities of the two-dimensional scanning sonar and the three-dimensional real-time sonar. Step S6-3: Substitute the actual sound velocity corresponding to the two-dimensional scanning sonar into the sonar ranging formula to recalculate the sound wave propagation distance corresponding to each pixel in the two-dimensional acoustic image, and substitute the actual sound velocity corresponding to the three-dimensional real-time sonar into the sonar ranging formula to recalculate the sound wave propagation distance corresponding to each pixel in the three-dimensional image. Step S6-4: Update the position coordinates of each pixel in the two-dimensional acoustic image according to the sound wave propagation distance corresponding to each pixel in the recalculated two-dimensional acoustic image, and update the position coordinates of each pixel in the three-dimensional image according to the sound wave propagation distance corresponding to each pixel in the recalculated three-dimensional image. Step S6-5: Output the updated 2D acoustic image and the updated 3D image as the corrected 2D acoustic image and the corrected 3D image.

5. The underwater salvage positioning and monitoring method according to claim 1, characterized in that, The specific implementation process of step S7 includes: Step S7-1: Establish a unified geodetic coordinate system as the fusion benchmark, and convert the corrected ship positioning information, grab bucket position information, corrected two-dimensional acoustic image, corrected three-dimensional image, and background image of the sunken ship and immersed tube position to the geodetic coordinate system. Step S7-2: Perform image registration on the corrected two-dimensional acoustic image and the corrected three-dimensional image, extract the feature points of the grab bucket and the sunken ship target in the image, and calculate the relative position of the grab bucket and the sunken ship target based on the coordinates of the feature points; Step S7-3: Spatially correlate the corrected ship positioning information with the grab bucket position information to determine the spatial position of the grab bucket relative to the main construction vessel, and combine the corrected ship positioning information to convert the grab bucket position information to a unified geodetic coordinate system. Step S7-4: The corrected two-dimensional acoustic image and the corrected three-dimensional image registered in step S7-2, the grab bucket position information converted in step S7-3, and the pre-imported background image of the sunken ship and immersed tube position are superimposed according to the coordinate correspondence to generate a fused comprehensive positioning image. Step S7-5: In the fused integrated positioning image, mark the relative distance and azimuth angle between the grab bucket and the sunken ship target in real time, and transmit the fused integrated positioning image to the crane control room display terminal in real time, so that the operator can direct the grab bucket to grab the target according to the relative position of the grab bucket and the sunken ship target in the fused integrated positioning image.

6. The underwater salvage positioning and monitoring method according to claim 4, characterized in that, The specific implementation process of step S6-2 includes: Sound velocity profilers are deployed near two-dimensional scanning sonar and three-dimensional real-time sonar to measure the sound velocity at different depths in the operating water area in real time and generate sound velocity profile data. Based on the sound velocity profile data, the average sound velocity along the propagation path of the sound wave from the transducer of the two-dimensional scanning sonar to the physical location of each pixel in the two-dimensional acoustic image is calculated, which is taken as the actual sound velocity of the two-dimensional scanning sonar. Based on the sound velocity profile data, the average sound velocity along the propagation path of the sound wave from the transducer of the three-dimensional real-time sonar to the physical location corresponding to each pixel in the three-dimensional image is calculated, which is taken as the actual sound velocity of the three-dimensional real-time sonar. The calculated actual sound velocities of the two-dimensional scanning sonar and the three-dimensional real-time sonar are used to replace the preset nominal sound velocities of the two-dimensional scanning sonar and the three-dimensional real-time sonar, respectively.

7. The underwater salvage positioning and monitoring method according to claim 6, characterized in that, The specific implementation process of updating the position coordinates of each pixel in the two-dimensional acoustic image and the position coordinates of each pixel in the three-dimensional image in step S6-4 includes: For two-dimensional scanning sonar, the horizontal and vertical azimuth angles corresponding to each pixel in the two-dimensional acoustic image are obtained. The recalculated sound wave propagation distance is used as the radial distance. Taking the transducer position as the origin, the updated position coordinates of each pixel in the local coordinate system of the transducer are determined according to the directions indicated by the horizontal and vertical azimuth angles. The original position coordinates of the pixel are replaced with the updated position coordinates. For 3D real-time sonar, the beam pointing angle and elevation angle corresponding to each pixel in the 3D image are obtained. The recalculated sound wave propagation distance is used as the radial distance. Taking the transducer position as the origin, the updated position coordinates of each pixel in the local coordinate system of the transducer are determined according to the direction indicated by the beam pointing angle and elevation angle. The original position coordinates of the pixel are replaced with the updated position coordinates.

8. The underwater salvage positioning and monitoring method according to claim 5, characterized in that, The specific implementation process of image registration between the corrected two-dimensional acoustic image and the corrected three-dimensional image in step S7-2 includes: Extract the edge contours of the shipwreck structure and the contours of the grab bucket from the corrected 2D acoustic image as 2D registration features; Extract the point cloud of the shipwreck structure and the point cloud of the grab bucket from the corrected 3D image as 3D registration features; The two-dimensional registration features are matched with the three-dimensional registration features, and the rotation and translation parameters between the two-dimensional acoustic image and the three-dimensional image are calculated. Based on the rotation and translation parameters, the corrected two-dimensional acoustic image and the corrected three-dimensional image are aligned to a unified geodetic coordinate system.

9. The underwater salvage positioning and monitoring method according to claim 5, characterized in that, In step S7-5, an early warning is issued when the relative distance between the grab bucket and the sunken ship target is less than a preset safety threshold.

Images