Ship parallel wet transport hose curvature radius distribution reconstruction method and system

Abstract

This application discloses a method and system for reconstructing the curvature radius distribution of a wet transport hose used in parallel operations between ships, relating to the fields of deep-sea mining and marine operations. The method uses a binocular camera and a laser linear array to acquire synchronous images and linear laser stripes of the hose crossing area, and collects attitude information from inertial measurement units and motion reference units on the decks of the two ships. Platform motion compensation is then applied to the synchronous images. Stereo matching and point cloud reconstruction are performed on the linear laser stripes and the compensated synchronous images, and discrete point sequences are extracted from the hose centerline. Based on the extracted discrete point sequences, the curvature radius distribution along the hose is calculated using a discrete curvature formula, and the current minimum radius is obtained. A formula for predicting the minimum radius is established based on a trend term coupled with sea state amplification using a time window, predicting the minimum curvature radius of the hose. This application reduces false alarms and missed alarms in hose bending monitoring, can predict the minimum curvature radius of the hose, and supports hose bending monitoring and prediction under high sea states.

Description

Technical Field

[0001] This application relates to the fields of deep-sea mining and marine operations technology, and in particular to a method and system for reconstructing the radius of curvature distribution of a wet transport hose used by a ship. Background Technology

[0002] During wet transport operations alongside ships, the hoses connecting the two vessels are highly susceptible to localized small-radius bends in the crossing area due to the combined effects of waves, ocean currents, and ship motion. This can lead to hose fatigue damage, reduced transport efficiency, and even leaks. Currently, most mainstream solutions for reconstructing the hose curvature radius distribution rely on single-sided camera monitoring or manual inspection, which has the following significant limitations: Monocular vision lacks depth information: A single-sided camera cannot accurately acquire the three-dimensional spatial shape of the hose, and the curvature estimation relies on prior assumptions, resulting in a large error; Manual inspections have poor real-time performance: they cannot be continuously monitored under high sea states, resulting in a high rate of missed reports; Lack of motion compensation: The six-degree-of-freedom motion of the ship leads to image blurring, inconsistent measurement references, and unstable curvature estimation; Lack of predictive capability: Existing solutions can only provide curvature estimates at the current moment and cannot predict the risk of hose bending in the short term.

[0003] Therefore, there is an urgent need for a method and system for reconstructing the curvature radius distribution of wet transport hoses for ships, which can be systematically optimized through digital technology to make hose curvature monitoring and prediction more accurate and adaptable to more complex sea conditions. Summary of the Invention

[0004] The purpose of this application is to provide a method and system for reconstructing the curvature radius distribution of wet transport hoses used in ship berthing, which solves the problems of low accuracy in hose curvature monitoring, lack of motion compensation, and inability to predict the trend of minimum radius changes in the current technology.

[0005] To achieve the above objectives, this application provides a method for reconstructing the radius of curvature distribution of wet transport hoses used in parallel berthing of ships, comprising the following steps: S1. By using binocular cameras and laser arrays arranged at both ends of the hose crossing area, synchronous images and line laser stripes of the hose crossing area are acquired, and attitude information of inertial measurement units (IMU) and motion reference units (MRU) on the decks of the two ships is collected to perform platform motion compensation on the synchronous images. S2. Perform stereo matching and point cloud reconstruction on the line laser stripes and the synchronized image after platform motion compensation, and extract the discrete point array of the hose centerline; S3. Based on the discrete point list of the hose centerline, calculate the radius of curvature distribution along the hose using the discrete curvature formula, and obtain the current minimum radius; S4. Establish a formula for predicting the minimum radius based on the time window-based trend term coupled with sea state amplification, and predict the minimum curvature radius of the hose.

[0006] Preferably, step S1 specifically includes: A binocular camera and a laser linear array are arranged at both ends of the hose crossing area to acquire the initial synchronization image and the initial line laser stripe, and distortion removal and epipolar correction are performed to obtain the synchronization image and line laser stripe. Inertial measurement units and motion reference units were deployed on the decks of both ships to collect attitude data of the two ships. Based on the attitude data of the two ships, platform motion compensation is performed on the synchronized image to obtain the synchronized image after platform motion compensation.

[0007] Preferably, the step of performing platform motion compensation on the synchronization image based on the attitude data of the two ships to obtain the platform motion-compensated synchronization image specifically includes: Based on the exposure time and reference time of the binocular camera, and combined with the attitude rotation matrix in the attitude data, the compensation rotation matrix of the left synchronized image and the compensation rotation matrix of the right synchronized image are calculated. Based on the compensation rotation matrix of the left synchronized image, the compensation rotation matrix of the right synchronized image, and the intrinsic parameter matrix of the binocular camera, construct the pixel domain homography matrix of the left synchronized image and the pixel domain homography matrix of the right synchronized image. The left and right synchronized images are resampled using the pixel-domain homography matrix of the left synchronized image and the pixel-domain homography matrix of the right synchronized image, respectively, to obtain the left and right synchronized images after platform motion compensation, i.e., the synchronized images after platform motion compensation.

[0008] Preferably, the expression for the compensation rotation matrix of the left-synchronized image is: ; ; in, The compensation rotation matrix for the left-synchronized image. For reference time, t L This is the exposure time for the left camera. t R This is the exposure time for the right camera. The deck attitude rotation matrix is ​​the reference time. This is the deck attitude rotation matrix corresponding to the exposure time of the left camera; The expression for the compensation rotation matrix of the right-synchronized image is: ; in, The compensation rotation matrix for the right-synchronized image. This is the deck attitude rotation matrix corresponding to the exposure time of the right camera.

[0009] Preferably, the expression for the pixel-domain homography matrix of the left-synchronized image is: ; in, This is the pixel-domain homography matrix of the left-synchronized image. The intrinsic parameter matrix of the left camera. The compensation rotation matrix for the left-synchronized image. This is the inverse of the intrinsic parameter matrix of the left camera; The expression for the pixel-domain homography matrix of the right-synchronized image is: ; in, This is the pixel-domain homography matrix of the right-synchronized image. The intrinsic parameter matrix of the right camera. The compensation rotation matrix for the right-synchronized image. It is the inverse of the intrinsic parameter matrix of the right camera.

[0010] Preferably, step S2 specifically includes: Stereo matching is performed on the synchronized images after platform motion compensation, and the point cloud depth is calculated based on parallax to obtain the three-dimensional point cloud on the surface of the hose. Based on the 3D point cloud, the center of the hose cross section is located by the center of the line laser bright band and the standard plate in each hose cross section. The discrete point series of the center point is extracted along the longitudinal direction and spline fitting is performed to generate a continuous center line, that is, the discrete point series of the hose center line is obtained.

[0011] Preferably, the formula for calculating the current minimum radius is: ; ; ; ; ; in, U min The current minimum radius, i Index of discrete nodes along the centerline. For the first i The distribution values ​​of the radius of curvature corresponding to each node For discrete curvature estimator, The lower limit threshold of curvature, Forward difference vector, It is the backward difference vector. For the first i +1 discrete nodes along the centerline For the first i Discrete nodes along the centerline For the first i -1 discrete nodes along the centerline.

[0012] Preferably, the formula for predicting the minimum radius is: ; ; in, To predict the minimum radius, t For time, The minimum radius at the current time. The term representing the decreasing trend of the minimum radius. For nonlinear redundancy coefficients, For forward-looking time window, For sea state coupling amplification, As a safety amplification factor for high sea states, For curvature sensitivity, It is the acceleration due to gravity. For the deck coordinate system horizontal x Directional acceleration, The deck coordinate system is perpendicular y Directional acceleration, Depth in the deck coordinate system z The acceleration along the direction line is given by RMS, which is the root mean square.

[0013] This application provides a system for reconstructing the curvature radius distribution of wet transport hoses used in parallel berthing of ships, for implementing the aforementioned method for reconstructing the curvature radius distribution of wet transport hoses used in parallel berthing of ships, including: The acquisition and motion compensation module is used to acquire synchronous images and line laser stripes of the hose crossing area through binocular cameras and laser arrays arranged at both ends of the hose crossing area, and to acquire attitude information of inertial measurement units and motion reference units on the decks of the two ships, and to perform platform motion compensation on the synchronous images. The point cloud reconstruction and centerline extraction module is used to perform stereo matching and point cloud reconstruction of the line laser stripes and the synchronized image after platform motion compensation, and to extract the discrete point array of the hose centerline. The curvature calculation module is used to calculate the curvature radius distribution along the hose based on a discrete point list of the hose centerline using a discrete curvature formula, and to obtain the current minimum radius. The prediction module is used to establish a formula for predicting the minimum radius based on the time window-based trend term coupled with sea state amplification, and to predict the minimum curvature radius of the hose.

[0014] This application also provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor, when calling the computer program in the memory, implements the content of the above-described method for reconstructing the curvature radius distribution of a ship's wet transport hose.

[0015] In summary, the method and system for reconstructing the curvature radius distribution of a wet transport hose for ships provided in this application have the following advantages compared to traditional technologies: By using a binocular camera and a line laser array, discrete points along the hose's centerline are extracted to obtain the overall three-dimensional curvature radius distribution of the hose, replacing single-point or local curvature estimation in existing solutions, significantly reducing false alarms and missed alarms in hose bend monitoring; by using attitude information collected by an inertial measurement unit and a motion reference unit to perform motion compensation on synchronized images, the interference of ship swaying and heave on visual measurements is effectively eliminated, ensuring that the curvature reconstruction results remain stable and reliable even under high sea states; by establishing a prediction minimum radius formula based on a trend term coupled with sea state amplification using a time window, the minimum curvature radius of the hose can be predicted, supporting hose bend monitoring and prediction under high sea states.

[0016] The technical solution of this application will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating a method for reconstructing the radius of curvature of a wet transport hose used for berthing of a ship, as described in an embodiment of this application. Figure 2 This is a schematic diagram illustrating the specific operation of acquiring synchronized images, line laser stripes, and attitude information to perform platform motion compensation in this embodiment of the application. Figure 3 This is a schematic diagram illustrating the specific operations of point cloud reconstruction and centerline extraction in the embodiments of this application; Figure 4 This is a schematic diagram of the instrument arrangement for the ship-to-ship wet transport hose curvature radius distribution reconstruction system in this embodiment of the application; wherein, 1, binocular camera; 2, laser linear array; 3, hose; 4, auxiliary camera; 5, standard plate; 6, IMU; 7, MRU; Figure 5 This is a schematic diagram illustrating the specific operation of curvature calculation and minimum radius prediction in the embodiments of this application; Figure 6 This is a block diagram of a ship-to-ship wet transport hose curvature radius distribution reconstruction system according to an embodiment of this application. Detailed Implementation

[0018] The technical methods of this application will be further described below with reference to the accompanying drawings and embodiments. It should be noted that, unless otherwise specifically stated, the relative arrangement, numerical expressions, and values ​​of the components and steps described in these embodiments do not limit the scope of this application.

[0019] The following description of at least one exemplary embodiment is merely illustrative and is not intended to limit the scope of this application or its application or use.

[0020] Techniques, systems, and equipment known to those skilled in the art may not be discussed in detail, but where appropriate, they should be considered part of the instruction manual.

[0021] In all the examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not as limitations. Therefore, other examples of exemplary embodiments may have different values.

[0022] Unless otherwise defined, the technical or scientific terms used in this application shall have the ordinary meaning as understood by a person of ordinary skill in the art to which this application pertains.

[0023] This application provides a method for reconstructing the radius of curvature distribution of wet transport hoses used in parallel operations on ships, such as... Figure 1 As shown, it includes the following steps: S1. By using binocular cameras and laser arrays arranged at both ends of the hose crossing area, synchronous images and line laser stripes of the hose crossing area are acquired, and attitude information of inertial measurement units (IMUs) and motion reference units (MRUs) on the decks of the two ships is collected to perform platform motion compensation on the synchronous images.

[0024] Furthermore, such as Figure 2 As shown, step S1 specifically includes the following steps: S11. A binocular camera and a laser linear array are positioned at both ends of the flexible tube crossing area to acquire initial synchronization images and initial line laser stripes. Distortion correction and epipolar correction are then performed to obtain the synchronization images and line laser stripes. Specifically, the initial synchronization images and initial line laser stripes are acquired, distortion correction and epipolar correction are performed, and it is determined whether the effect of the initial synchronization images and initial line laser stripes is acceptable. If acceptable, subsequent platform motion compensation is performed; otherwise, the standard plate position is adjusted, and acquisition is repeated until an acceptable initial synchronization image and initial line laser stripes are obtained.

[0025] S12. Install IMUs and MRUs on the decks of both ships to collect attitude data. Specifically, after collecting the attitude data, determine whether the motion of the two ships is stable. If stable, perform subsequent platform motion compensation; otherwise, stop the machine immediately.

[0026] S13. Based on the attitude data of the two ships, perform platform motion compensation on the synchronized image to obtain the image after platform motion compensation. Specific content includes: Based on the exposure time and reference time of the stereo camera, and combined with the pose rotation matrix in the pose data, the compensation rotation matrix for the left-synchronized image and the right-synchronized image are calculated. The expression for the compensation rotation matrix of the left-synchronized image is as follows: ; ; in, The compensation rotation matrix for the left-synchronized image. For reference time, t L This is the exposure time for the left camera. t R This is the exposure time for the right camera. The deck attitude rotation matrix is ​​the reference time. This is the deck attitude rotation matrix corresponding to the exposure time of the left camera.

[0027] The expression for the compensation rotation matrix of the right-synchronized image is: ; in, The compensation rotation matrix for the right-synchronized image. This is the deck attitude rotation matrix corresponding to the exposure time of the right camera.

[0028] Based on the compensated rotation matrix of the left synchronized image, the compensated rotation matrix of the right synchronized image, and the intrinsic parameter matrix of the stereo camera, the pixel-domain homography matrix of the left synchronized image and the pixel-domain homography matrix of the right synchronized image are constructed. The expression for the pixel-domain homography matrix of the left synchronized image is as follows: ; in, This is the pixel-domain homography matrix of the left-synchronized image. The intrinsic parameter matrix of the left camera. The compensation rotation matrix for the left-synchronized image. It is the inverse of the intrinsic parameter matrix of the left camera.

[0029] The expression for the pixel-domain homography matrix of a right-synchronized image is: ; in, This is the pixel-domain homography matrix of the right-synchronized image. The intrinsic parameter matrix of the right camera. The compensation rotation matrix for the right-synchronized image. It is the inverse of the intrinsic parameter matrix of the right camera.

[0030] The left and right synchronized images are resampled using the pixel-domain homography matrix of the left synchronized image and the pixel-domain homography matrix of the right synchronized image, respectively, to obtain the left and right synchronized images after platform motion compensation, i.e., the synchronized images after platform motion compensation.

[0031] S2. Perform stereo matching and point cloud reconstruction on the line laser stripes and the synchronous image after platform motion compensation, and extract the discrete point sequence of the hose centerline.

[0032] Furthermore, such as Figure 3 As shown, step S2 specifically involves: S21. Perform stereo matching on the synchronized image after platform motion compensation, calculate the point cloud depth based on parallax, and obtain the three-dimensional point cloud on the surface of the hose.

[0033] S22. Based on the 3D point cloud, the center of the hose cross-section is located using the center of the line laser bright band and the standard plate. A discrete point sequence of center points is extracted longitudinally and spline fitted to generate a continuous centerline, thus obtaining the discrete point sequence of the hose centerline. Specifically, the spline fitting process requires determining whether the fitting effect is acceptable. If acceptable, the discrete point sequence of the hose centerline is output; otherwise, the spline fitting conditions are adjusted, the discrete point sequence of center points is re-established, and spline fitting is performed again until an acceptable fitting effect is obtained.

[0034] S3. Based on the discrete point list of the hose centerline, calculate the radius of curvature distribution along the hose using the discrete curvature formula, and obtain the current minimum radius. The formula for calculating the current minimum radius is: ; ; ; ; ; in, U min The current minimum radius, i Index of discrete nodes along the centerline. Let be the radius of curvature distribution value corresponding to the i-th node. For discrete curvature estimator, This is the lower limit threshold for curvature (used to suppress abnormally large radii or numerical instability caused by noise). Forward difference vector, It is the backward difference vector. For the first i +1 discrete nodes along the centerline For the first i Discrete nodes along the centerline For the first i -1 discrete nodes along the centerline.

[0035] S4. Establish a formula for predicting the minimum radius based on the time window-based trend term coupled with sea state amplification, and predict the minimum curvature radius of the hose. The formula for calculating the minimum radius is as follows: ; ; in, To predict the minimum radius, t For time, The minimum radius at the current time. The term representing the decreasing trend of the minimum radius. For nonlinear redundancy coefficients, For forward-looking time window, For sea state coupling amplification, As a safety amplification factor for high sea states, For curvature sensitivity, It is the acceleration due to gravity. For the deck coordinate system horizontal x Directional acceleration, The deck coordinate system is perpendicular y Directional acceleration, Depth in the deck coordinate system z The acceleration along the direction line is given by RMS, which is the root mean square.

[0036] In an exemplary embodiment of this application, as follows Figure 4 As shown, a set of binocular cameras 1 and a laser array 2 are arranged on each of the two ship decks, looking down at the area crossed by the flexible tube. The pitch angle range is 15°-30°. The area that the left and right cameras of binocular camera 1 can simultaneously see occupies no less than 80% of the area crossed by the flexible tube. At the middle position of the area crossed by the flexible tube, the common field of view of binocular camera 1 leaves a redundancy of no less than 1.5 times the diameter of flexible tube 3 on both sides. The baseline of binocular camera 1 is 15-35cm; the working distance is 6-20m; the parallax is no less than 40 pixels (center section), avoiding hanging points and railings. If necessary, 1-2 auxiliary viewing angles, i.e., auxiliary cameras 4, are added. The intrinsic parameter matrix K of the left and right cameras (preferably including focal length) is completed using a 9×6 checkerboard grid (30-40mm squares). f x and f y Main point c x and c yThe binocular extrinsic parameters (relative pose between the left and right cameras) are calibrated, and the reprojection error is checked using epipolar correction.

[0037] Constructing a deck coordinate system :level x The direction is towards the bow, perpendicular. y Direction to starboard, depth z With the direction vertically upward, use 3-4 deck targets for extrinsic parameter fitting, and transform the binocular camera coordinates to the deck coordinate system using rigid body transformation. Linear laser scanning is used to scan the checkerboard pattern, and the laser plane equation is fitted. ,in In the deck coordinate system Three-dimensional points in It is the unit normal vector ( ), This is the signed distance from the plane to the origin, and the line width and visibility at the working distance are verified. Five labels are placed every 0.5-1.0m along the hose, and the label spacing is recorded. With hose outer diameter Three points were verified using a soft ruler, and the mean and standard deviation were taken as the geometric priors for the actual on-site scale and robust filtering.

[0038] Simultaneously acquire two synchronized images from the left and right cameras at the same time. Simultaneously, six-DOF attitude data from IMU6 and MRU7 on both ship decks were acquired. Synchronized images were then analyzed. First, distortion correction and epipolar correction are performed. High-brightness stripes formed by the laser beam on the surface of the flexible tube 3 and the corresponding target points are detected. Cross-image matching is then performed to find the corresponding points in the right synchronous image for the stripe points and target points in the left synchronous image. The acceptableness of the synchronous image and laser stripe effect is determined based on whether the time stamp difference between the left and right synchronous images is less than one frame period, whether the high-brightness stripes are continuous, whether the high-brightness stripes are visible, and whether the consistency between the left and right synchronous images is not less than 80%. Otherwise, the target spacing is adjusted. Therefore, platform motion compensation is performed based on the attitude outputs of the corresponding deck IMU 6 and MRU 7. The preferred method is to unify the left and right synchronized images to the same reference attitude as follows: Let the exposure times of the left and right cameras be respectively... t L and t R Take reference time The deck attitude rotation matrix is ​​obtained from IMU 6 and MRU 7. Calculate the compensated rotation of the left and right synchronized images: ; ; And from the intrinsic parameter matrices of the left and right cameras and Construct the pixel-domain homography matrix: ; ; The left and right synchronized images were resampled separately to obtain synchronized images after platform motion compensation. .

[0039] like Figure 5 As shown, the synchronized image after platform motion compensation The horizontal pixel difference, i.e., parallax Further calculate the point cloud depth ,in Z For point cloud depth, f For the focal length of a binocular camera, B The baseline is used for both eyes and then substituted back into the deck coordinate system. , expressed in the unified coordinate system of the deck.

[0040] Use the center of the bright band and the marker to locate the center of the hose profile at each hose cross-section, and mark the center point longitudinally. The spline is smoothed and fitted using B-spline or Cardinal spline curves to generate a continuous centerline. The spline fitting effect is judged to be acceptable based on the error between the spline and discrete points not exceeding 20-50 mm and the absence of obvious peaks or outliers in the curvature. Otherwise, the spline smoothing parameters or smoothing method are adjusted. If acceptable, for a given set of three points... ,definition: ; ; Using the defined discrete curvature estimate With radius distribution : ; ; For the radius of curvature sequence Noise reduction is achieved using a two-stage filtering process: median filtering and Savitzky-Golay filtering. The optimal filtering window contains 7-11 points to obtain a more stable radius of curvature sequence and the current minimum radius of curvature. ; Obtain the radius of curvature sequence With the current minimum radius Using the attitude and displacement information output by the deck IMU and MRU, the point cloud and centerline are transformed from the camera coordinate system to the deck coordinate system. A time-domain low-pass filter is then applied, with a cutoff frequency preferably between 1.5 and 2.0 Hz. Zero drift and temperature drift corrections are performed using a fixed target on the deck, with a correction cycle preferably every 10-15 minutes, or a reset correction is performed after an alarm is triggered, to ensure the long-term stability of coordinate transformation and centerline estimation.

[0041] Defining the Future Minimum radius for prediction in seconds: ; in, Estimated by the 2-3s sliding window slope.

[0042] Sea state coupled amplification The resultant acceleration is calculated from the combined acceleration measured by the deck IMU and MRU and converted to curvature sensitivity. The combined acceleration ratio is defined using the roll, pitch, and heave measurements from the deck IMU and MRU, representing the ratio of the current combined acceleration to the standard gravitational acceleration: ; And define its root mean square within a 10-20s sliding window: .

[0043] Define sea state coupled amplification : ; in, Defined as the conservative reduction in the minimum radius of curvature of the hose caused by the root mean square of the unit resultant acceleration, in meters (m), it is obtained by calibration using the root mean square of the resultant acceleration and the change in the minimum radius of curvature of the hose, which are synchronously collected during the commissioning period. The safety amplification factor for high sea states is determined based on field experience / experiments.

[0044] To determine whether the predicted minimum radius is acceptable, the ultimate failure radius is set based on hose bending failure tests or ultimate bending data. Determine if the predicted minimum radius is greater than the ultimate failure radius. If it is acceptable, output the predicted minimum radius; otherwise, adjust the curvature sensitivity. Safety amplification factor under high sea states and forward-looking time window .

[0045] This application provides a system for reconstructing the curvature radius distribution of wet transport hoses used in parallel berthing of ships, for implementing the aforementioned method for reconstructing the curvature radius distribution of wet transport hoses used in parallel berthing of ships. Figure 6 As shown, it includes: The acquisition and motion compensation module is used to acquire synchronized images and line laser stripes of the hose crossing area through binocular cameras and laser arrays arranged at both ends of the hose crossing area, and to acquire attitude information of inertial measurement units and motion reference units on the decks of the two ships, and to perform platform motion compensation on the synchronized images.

[0046] The point cloud reconstruction and centerline extraction module is used to perform stereo matching and point cloud reconstruction of the line laser stripes and the synchronous image after platform motion compensation, and to extract the discrete point series of the hose centerline.

[0047] The curvature calculation module is used to calculate the curvature radius distribution along the hose line based on a discrete point list of the hose centerline using the discrete curvature formula, and to obtain the current minimum radius.

[0048] The prediction module is used to establish a formula for predicting the minimum radius based on the time window-based trend term coupled with sea state amplification, and to predict the minimum curvature radius of the hose.

[0049] This application provides an electronic device, including a memory and a processor. The memory stores a computer program, and when the processor calls the computer program in the memory, it implements the content of the above-mentioned method for reconstructing the curvature radius distribution of ship-to-wet transport hoses.

[0050] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application and not to limit them. Although this application has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of this application, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of this application.

Claims

1. A method for reconstructing the radius of curvature distribution of a wet transport hose used for alongside ships, characterized in that, Includes the following steps: S1. By using binocular cameras and laser arrays arranged at both ends of the hose crossing area, synchronous images and line laser stripes of the hose crossing area are obtained, and attitude information of inertial measurement units and motion reference units on the decks of the two ships is collected to perform platform motion compensation on the synchronous images. S2. Perform stereo matching and point cloud reconstruction on the line laser stripes and the synchronized image after platform motion compensation, and extract the discrete point array of the hose centerline; S3. Based on the discrete point list of the hose centerline, calculate the radius of curvature distribution along the hose using the discrete curvature formula, and obtain the current minimum radius; S4. Establish a formula for predicting the minimum radius based on the time window-based trend term coupled with sea state amplification, and predict the minimum curvature radius of the hose. The formula for predicting the minimum radius is: ； ； in, To predict the minimum radius, t For time, The minimum radius at the current time. The term representing the decreasing trend of the minimum radius. For nonlinear redundancy coefficients, For forward-looking time window, For sea state coupling amplification, As a safety amplification factor for high sea states, For curvature sensitivity, It is the acceleration due to gravity. For the deck coordinate system horizontal x Directional acceleration, The deck coordinate system is perpendicular y Directional acceleration, Depth in the deck coordinate system z The acceleration along the direction line is given by RMS, which is the root mean square.

2. The method for reconstructing the radius of curvature distribution of a wet transport hose used for ship berthing according to claim 1, characterized in that, Step S1 specifically includes: A binocular camera and a laser linear array are arranged at both ends of the hose crossing area to acquire the initial synchronization image and the initial line laser stripe, and distortion removal and epipolar correction are performed to obtain the synchronization image and line laser stripe. Inertial measurement units and motion reference units were deployed on the decks of both ships to collect attitude data of the two ships. Based on the attitude data of the two ships, platform motion compensation is performed on the synchronized image to obtain the synchronized image after platform motion compensation.

3. The method for reconstructing the radius of curvature distribution of a wet transport hose used for ship berthing according to claim 2, characterized in that, The step of performing platform motion compensation on the synchronization image based on the attitude data of the two ships to obtain the platform motion compensated synchronization image specifically includes: Based on the exposure time and reference time of the binocular camera, and combined with the attitude rotation matrix in the attitude data, the compensation rotation matrix of the left synchronized image and the compensation rotation matrix of the right synchronized image are calculated. Based on the compensation rotation matrix of the left synchronized image, the compensation rotation matrix of the right synchronized image, and the intrinsic parameter matrix of the binocular camera, construct the pixel domain homography matrix of the left synchronized image and the pixel domain homography matrix of the right synchronized image. The left and right synchronized images are resampled using the pixel-domain homography matrix of the left synchronized image and the pixel-domain homography matrix of the right synchronized image, respectively, to obtain the left and right synchronized images after platform motion compensation, i.e., the synchronized images after platform motion compensation.

4. The method for reconstructing the radius of curvature distribution of a wet transport hose used for ship berthing according to claim 3, characterized in that, The expression for the compensation rotation matrix of the left-synchronized image is: ； ； in, The compensation rotation matrix for the left-synchronized image. For reference time, t L This is the exposure time for the left camera. t R This is the exposure time for the right camera. The deck attitude rotation matrix is ​​the reference time. This is the deck attitude rotation matrix corresponding to the exposure time of the left camera; The expression for the compensation rotation matrix of the right-synchronized image is: ； in, The compensation rotation matrix for the right-synchronized image. This is the deck attitude rotation matrix corresponding to the exposure time of the right camera.

5. The method for reconstructing the radius of curvature distribution of a wet transport hose used for ship berthing according to claim 3, characterized in that, The expression for the pixel-domain homography matrix of the left-synchronized image is: ； in, This is the pixel-domain homography matrix of the left-synchronized image. The intrinsic parameter matrix of the left camera. The compensation rotation matrix for the left-synchronized image. This is the inverse of the intrinsic parameter matrix of the left camera; The expression for the pixel-domain homography matrix of the right-synchronized image is: ； in, This is the pixel-domain homography matrix of the right-synchronized image. The intrinsic parameter matrix of the right camera. The compensation rotation matrix for the right-synchronized image. It is the inverse of the intrinsic parameter matrix of the right camera.

6. The method for reconstructing the radius of curvature distribution of wet transport hoses for ships alongside each other according to claim 1, characterized in that, Step S2 specifically includes: Stereo matching is performed on the synchronized images after platform motion compensation, and the point cloud depth is calculated based on parallax to obtain the three-dimensional point cloud on the surface of the hose. Based on the 3D point cloud, the center of the hose cross section is located by the center of the line laser bright band and the standard plate in each hose cross section. The discrete point series of the center point is extracted along the longitudinal direction and spline fitting is performed to generate a continuous center line, that is, the discrete point series of the hose center line is obtained.

7. The method for reconstructing the radius of curvature distribution of a wet transport hose used for ship berthing according to claim 1, characterized in that, The formula for calculating the current minimum radius is: ； ； ； ； ； in, U min The current minimum radius, i Index of discrete nodes along the centerline. For the first i The distribution values ​​of the radius of curvature corresponding to each node For discrete curvature estimator, The lower limit threshold of curvature, Forward difference vector, It is the backward difference vector. For the first i +1 discrete nodes along the centerline For the first i Discrete nodes along the centerline For the first i -1 discrete nodes along the centerline.

8. A system for reconstructing the radius of curvature distribution of a wet transport hose used in parallel operations of a ship, characterized in that, A method for reconstructing the radius of curvature distribution of a wet transport hose used in conjunction with a ship, as described in any one of claims 1-7, includes: The acquisition and motion compensation module is used to acquire synchronous images and line laser stripes of the hose crossing area through binocular cameras and laser arrays arranged at both ends of the hose crossing area, and to acquire attitude information of inertial measurement units and motion reference units on the decks of the two ships, and to perform platform motion compensation on the synchronous images. The point cloud reconstruction and centerline extraction module is used to perform stereo matching and point cloud reconstruction of the line laser stripes and the synchronized image after platform motion compensation, and to extract the discrete point array of the hose centerline. The curvature calculation module is used to calculate the curvature radius distribution along the hose based on a discrete point list of the hose centerline using a discrete curvature formula, and to obtain the current minimum radius. The prediction module is used to establish a formula for predicting the minimum radius based on the time window-based trend term coupled with sea state amplification, and to predict the minimum curvature radius of the hose.

9. An electronic device, characterized in that, It includes a memory and a processor, wherein the memory stores a computer program, and the processor, when calling the computer program in the memory, implements the content of the method for reconstructing the curvature radius distribution of a ship's wet transport hose as described in any one of claims 1-7.

Images