A method for measuring the attitude of ocean buoys based on multi-view vision and solar vector.

By combining multi-view vision with solar vectors, multiple imaging modules and a zenith imaging module are used to extract features of the sea-line and solar images. These features are then combined with GPS information for weighted fusion, solving the problem of insufficient attitude measurement accuracy of ocean buoys under complex sea conditions and achieving low-cost, high-precision attitude calculation.

CN123111183APending Publication Date: 2026-11-17NANTONG UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611312772.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-08-27
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

Existing methods for measuring the attitude of marine buoys are not accurate enough in complex sea conditions and rely on high-cost and high-power INS and GNSS systems, making it difficult to achieve stable attitude measurement of long-term unattended platforms.

Method used

By combining multi-view vision with solar vectors, multiple imaging modules and a zenith imaging module are deployed on the buoy to extract features of the sea-line and solar images. A weighted fusion algorithm is used to calculate the attitude angle, and information obtained from GPS is combined to perform autonomous attitude calculation.

Benefits of technology

It improves the buoy's adaptability and attitude measurement accuracy in complex sea conditions, reduces costs, avoids the effects of magnetic fields and inertial navigation drift, and is suitable for long-term unattended platforms.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123111183A_ABST
    Figure CN123111183A_ABST
Patent Text Reader

Abstract

This invention discloses a method for measuring the attitude of an ocean buoy based on multi-view vision and solar vectors. The method includes: extracting the horizon line from a horizon image, obtaining the intercept and slope, and calculating the root mean square error (RMSE) of fitting each horizon line; extracting the center of the solar spot from a solar image and calculating the solar vector in the camera coordinate system; calculating the pitch and roll angles based on the intercept difference and slope of the horizon lines; determining the weights of each estimate based on the RMSE, and performing weighted fusion of the pitch and roll angles to output the final buoy pitch and roll angles; converting the solar vector to a horizontal coordinate system using the buoy pitch and roll angles to obtain the measured solar azimuth angle; and calculating the buoy's yaw angle from the difference between the reference solar azimuth angle and the measured solar azimuth angle. This invention achieves fully autonomous, drift-free visual measurement of buoy attitude, has strong anti-interference capabilities, low cost, and is suitable for long-term ocean monitoring.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of optical measurement technology, specifically to a method for measuring the attitude of marine buoys based on multi-view vision and solar vector. Background Technology

[0002] Attitude measurement of ocean buoys is a crucial foundation for marine environmental monitoring, satellite communication alignment, and platform stability control. Currently, buoy attitude measurement primarily relies on Global Navigation Satellite System (GNSS) attitude modules and high-precision Inertial Navigation Systems (INS). GNSS attitude measurement methods are limited by the finite size of the buoy, resulting in short baselines that lead to insufficient measurement accuracy, and are susceptible to multipath effects. High-precision INS are costly and power-intensive, making them unsuitable for long-term unattended buoy platforms. While low-cost Micro-Electro-Mechanical Systems (MEMS) sensors can be used for attitude estimation, gyroscopes suffer from drift errors, accelerometers are susceptible to non-gravitational acceleration interference in dynamic marine environments, and magnetometers are affected by hard and soft magnetic interference. In dynamic marine environments, attitude calculation errors are prone to divergence.

[0003] In recent years, methods for measuring the attitude of surface platforms using visual information have emerged. These methods typically involve using a monocular camera to capture images of the sea-line in front of the platform and calculating the attitude based on the changes in the sea-line's position in the image. However, monocular methods require extremely high reliability in sea-line extraction and are easily affected by the platform's heave (rise and fall) motion. Furthermore, in severe sea conditions, when the sea-line is jagged or exhibits an irregular curve, the measurement accuracy of a single feature drops sharply. Summary of the Invention

[0004] Purpose of the invention: The purpose of this invention is to solve the problems existing in the prior art and provide a method for measuring the attitude of marine buoys based on multi-view vision and solar vector, thereby improving the system's adaptability and overall measurement accuracy under complex sea conditions.

[0005] Technical solution: The present invention provides a method for measuring the attitude of an ocean buoy based on multi-view vision and solar vector, comprising:

[0006] S1. At least four imaging modules are deployed at multiple horizontal positions of the buoy, with the optical axis of each imaging module pointing outward in the horizontal direction, for synchronously acquiring images of the sea-sky line; a zenith imaging module is deployed at the top of the buoy, with its optical axis pointing vertically upward, for acquiring images of the sun;

[0007] S2. Extract the sea-line straight line from each of the sea-line images, obtain the intercept and slope of each sea-line in its respective image coordinate system, and calculate the root mean square error of the fitting of each sea-line straight line; at the same time, extract the center of the solar spot from the solar image and calculate the solar vector in the camera coordinate system.

[0008] S3. Calculate the first estimated value of the pitch angle and the second estimated value of the roll angle based on the intercept difference and slope of the sea-line extracted by the first imaging module. Calculate the first estimated value of the roll angle and the second estimated value of the pitch angle based on the intercept difference and slope of the sea-line extracted by the second imaging module. The first imaging module and the second imaging module are perpendicular to each other.

[0009] S4. Determine the weight of each estimate based on the root mean square error, perform weighted fusion of the first estimate and the second estimate, and output the final buoy pitch angle and roll angle.

[0010] S5. Using the pitch angle and roll angle, the solar vector in the camera coordinate system is transformed to the horizontal coordinate system to obtain the measured solar azimuth angle; combined with the latitude, longitude and time provided by the global positioning system receiver, the reference solar azimuth angle is calculated through ephemeris, and the yaw angle of the buoy is calculated from the difference between the reference solar azimuth angle and the measured solar azimuth angle.

[0011] Further, in step S1, the at least four imaging modules are deployed at intervals along the circumference of the buoy, including four imaging modules deployed on the front, rear, left and right sides of the buoy, with the optical axis azimuth angles of adjacent imaging modules differing by 90°; the zenith imaging module is installed at the center of the top of the buoy.

[0012] Furthermore, in step S3, the first imaging module includes two imaging modules deployed on the front and rear sides of the buoy, and the second imaging module includes two imaging modules deployed on the left and right sides of the buoy.

[0013] The first estimated value of the pitch angle is calculated from the difference in intercepts of the sea-line lines of the front and rear imaging modules, and the first estimated value of the roll angle is calculated from the difference in intercepts of the sea-line lines of the left and right imaging modules.

[0014] The second estimated value of the pitch angle is obtained by converting the slope of the sea-line line of the left and right imaging modules, and the second estimated value of the roll angle is obtained by converting the slope of the sea-line line of the front and rear imaging modules.

[0015] Furthermore, the formulas for calculating the first estimated value of the pitch angle and the first estimated value of the roll angle are as follows:

[0016] ;

[0017] in, This is the first estimated value of the pitch angle. This is the first estimated value for the roll angle. , , , The intercepts of the sea-line spectrum extracted by the front, rear, left, and right imaging modules are respectively. The pixel focal length of the imaging module;

[0018] Furthermore, the formulas for calculating the second estimate of the pitch angle and the second estimate of the roll angle are as follows:

[0019] ;

[0020] in, The second estimated value of the pitch angle. Second estimate of roll angle , , , The slope of the sea-line system extracted from the imaging modules on the right, left, front, and rear sides.

[0021] Further, in step S4, determining the weight of each estimated value based on the root mean square error includes:

[0022] Define the first estimate of the pitch angle Dependency quality Second estimate of pitch angle Dependency quality for:

[0023] ;

[0024] First estimate of pitch angle weight Second estimate of pitch angle weight for:

[0025] ;

[0026] in, , , , These represent the root mean square error when the front, rear, left, and right imaging modules extract the sea-line area, respectively.

[0027] Furthermore, in step S4, determining the weights of each estimated value based on the root mean square error also includes:

[0028] First estimate of roll angle Corresponding quality indicators Second estimate of roll angle Corresponding quality indicators ;

[0029] ;

[0030] First estimate of roll angle weight Second estimate of roll angle weight for:

[0031] .

[0032] Further, in step S4, the weighted fusion of the first estimate and the second estimate is performed using the following formula:

[0033] ;

[0034] in, For the merged pitch angle, This is the combined roll angle.

[0035] Furthermore, the calculation process for the yaw angle is as follows:

[0036] Obtain the measured solar vector in the camera coordinate system Using pitch angle and roll angle Construct rotation matrix The horizontal solar vector is obtained as follows:

[0037] ;

[0038] Measuring the solar azimuth angle:

[0039] ;

[0040] in, , This refers to the horizontal component of the horizontal solar vector.

[0041] The latitude, longitude, and time are obtained through the Global Positioning System, and the reference solar azimuth angle is calculated using a solar ephemeris position algorithm. ;

[0042] Calculate the yaw angle:

[0043] .

[0044] Further, in step S2, the root mean square error is calculated as follows: for the N edge points participating in the sea-line straight line fitting, the vertical residual of each point relative to the fitted straight line is calculated, and then the root mean square value of the vertical residual is calculated.

[0045] Beneficial effects: Compared with the prior art, the present invention has the following significant advantages:

[0046] This invention utilizes at least four cameras and simultaneously leverages both the height and tilt characteristics of the sea surface. Based on the real-time RMSE extracted from the sea surface of each camera, the fusion weights are dynamically adjusted. This automatically suppresses interference in a particular observation direction, improving the system's adaptability and overall measurement accuracy under complex sea conditions. The height difference method eliminates common-mode interference from heave motion through differential calculation, while the tilt method is free from heave interference. The weighted fusion of these two methods further enhances the stability and reliability of the output attitude. A zenith camera and GPS ephemeris reference enable autonomous, drift-free measurement of the buoy's yaw angle, forming a complete three-dimensional attitude vision measurement system. The heading measurement is independent of the magnetic compass, unaffected by magnetic field interference and inertial navigation drift, offering low cost and high accuracy, making it particularly suitable for long-term unmanned platforms such as ocean buoys. Attached Figure Description

[0047] Figure 1 This is a flowchart of the marine buoy attitude testing method of the present invention.

[0048] Figure 2 This is a top view of the external structure of the buoy in an embodiment of the present invention.

[0049] Figure 3 This is a front view of the external structure of the buoy in an embodiment of the present invention.

[0050] Figure 4 This is a cross-sectional view of the internal structure of the buoy in an embodiment of the present invention.

[0051] Figure 5 The image shows the intercept and slope of the fitted line of the imaging module in this embodiment of the invention.

[0052] Figure 6 This is a solar image taken by the zenith imaging module covered by the Baader membrane according to an embodiment of the present invention.

[0053] Figure 7 The diagram shows the simulation results of attitude angle estimation for roll angle, pitch angle, and yaw angle in an embodiment of the present invention.

[0054] Figure 8 This is a diagram showing the estimation error of the simulated attitude angles (roll, pitch, and yaw) in an embodiment of the present invention. Detailed Implementation

[0055] The technical solution of the present invention will be further described below with reference to the accompanying drawings.

[0056] This embodiment uses a certain type of cylindrical marine meteorological buoy as an application platform to provide a marine buoy attitude measurement system for implementing the method of the present invention. Taking a horizontally deployed four imaging modules as an example, it includes:

[0057] Five imaging modules are used to acquire images of the sea-line and the sun;

[0058] The GPS module is used to provide latitude, longitude, and time, and to calculate the reference solar vector;

[0059] Baader film is used to prevent overexposure in solar imaging;

[0060] The main control processor is used for data processing and to execute the method of this invention.

[0061] The power module supplies power to the above components.

[0062] like Figure 2 , Figure 3 and Figure 4 As shown, in the ocean buoy attitude measurement system, the five imaging modules are: the first imaging module 3 installed at the front of the buoy, the second imaging module 4 installed at the left side of the buoy, the third imaging module 1 installed at the rear of the buoy, the fourth imaging module 2 installed at the right side of the buoy, and the fifth imaging module 5 installed above the buoy. The optical axes of the four imaging modules installed horizontally on the buoy point outwards along the horizontal direction, used to simultaneously acquire four sea-sky images; the fifth imaging module is a zenith imaging module, with its optical axis pointing vertically upwards, used to acquire solar images. Because sunlight is prone to overexposure, the imaging module needs to be covered with a Baader film. The five imaging modules are synchronously triggered via a hardware synchronization cable, and the image data is transmitted to the main control processor 6 (STM32H7 series) for processing.

[0063] like Figure 1 As shown, the following is a detailed description of a method for measuring the attitude of an ocean buoy based on multi-view vision and solar vectors, including the following steps:

[0064] (1) Extract the straight line of the sea line from each sea line image, obtain the intercept and slope of each sea line in its respective image coordinate system, and calculate the root mean square error of the fitting of each sea line; at the same time, extract the center of the solar spot from the solar image and calculate the solar vector in the camera coordinate system.

[0065] like Figure 5 As shown, the origin of the image coordinate system is at the top left corner, and the horizontal axis... To the right, vertical axis Downward. For the four imaging modules in the front, rear, left, and right directions, edge detection is used to obtain a series of points located on the edge of the sea-line. , ,… ,... .in These are the column coordinates (horizontal direction) of the image. These are row coordinates (vertical direction). Using the least squares method to fit a straight line to these points, we obtain the equation of the line:

[0066] ;

[0067] in It's the slope. This refers to the intercept. In addition to outputting the intercept and tilt angle of the sea-line, the image processing unit simultaneously calculates the root mean square error (RMSE) of the straight-line fitting for each sea-line: for the N candidate points of the sea-line participating in the fitting, it calculates the vertical residual of each point relative to the fitted straight line.

[0068] ;

[0069] Root mean square error of straight-line fitting of sea surface:

[0070] ;

[0071] This value directly reflects the clarity and straightness of the horizon line; the smaller the RMSE, the higher the extraction quality.

[0072] For the zenith imaging module, the coordinates of the solar center are calculated using the gray-scale centroid method. :

[0073] ;

[0074] in, For pixels grayscale value, Its pixel coordinates, This represents the total number of pixels within the light spot area.

[0075] Based on the camera intrinsic parameter matrix :

[0076] ;

[0077] in, The x-coordinate of the principal point The y-coordinate of the principal point The focal length is in the x-direction. The focal length is in the y-direction;

[0078] Convert pixel coordinates to coordinates in the normalized camera coordinate system. :

[0079] ;

[0080] Normalizing the obtained vector yields the measured solar vector in the camera coordinate system:

[0081] .

[0082] (2) Calculate the first estimated value of pitch angle and the second estimated value of roll angle based on the intercept difference and slope of the sea-line line extracted by the first imaging module, and calculate the first estimated value of roll angle and the second estimated value of pitch angle based on the intercept difference and slope of the sea-line line extracted by the second imaging module; the first imaging module and the second imaging module are perpendicular to each other.

[0083] The first imaging module includes two imaging modules deployed on the front and rear sides of the buoy, and the second imaging module includes two imaging modules deployed on the left and right sides of the buoy;

[0084] The first estimated value of the pitch angle is calculated from the difference in intercepts of the sea-line lines of the front and rear imaging modules, and the first estimated value of the roll angle is calculated from the difference in intercepts of the sea-line lines of the left and right imaging modules.

[0085] The second estimated value of the pitch angle is obtained by converting the slope of the sea-line line of the left and right imaging modules, and the second estimated value of the roll angle is obtained by converting the slope of the sea-line line of the front and rear imaging modules.

[0086] Define the buoy's coordinate system: the X-axis points towards the bow, the Y-axis points towards starboard, and the Z-axis points downwards. Pitch angle. For rotation about the Y-axis, roll angle The system rotates around the X-axis. The optical axis of the front camera is approximately along the X-axis, and that of the right camera is along the Y-axis. The sea-line coordinates of the front and rear cameras shift vertically, and the change in intercept reflects the pitch angle. Meanwhile, the horizon lines of the left and right cameras rotate, and the slope reflects the pitch angle. The vertical shift of the horizon line of the left and right cameras reflects the change in roll angle. Meanwhile, the horizon lines of the front and rear cameras rotate, and the slope reflects the roll angle. .

[0087] First estimate of pitch angle First estimate of roll angle The calculation formula is as follows:

[0088] ;

[0089] in, , , , The intercepts of the sea-line spectrum extracted from the front, rear, left, and right imaging modules are respectively. This refers to the pixel focal length of the imaging module.

[0090] Second estimate of pitch angle Second estimate of roll angle The calculation formula is as follows:

[0091] ;

[0092] in, , , , The slope of the sea-line system extracted from the right, left, front, and rear cameras.

[0093] Define the first estimate of the pitch angle Dependency quality Second estimate of pitch angle Dependency quality for:

[0094] ;

[0095] in, , , , These represent the root mean square error when the front, rear, left, and right imaging modules extract the sea-line area, respectively.

[0096] First estimate of pitch angle weight Second estimate of pitch angle weight for:

[0097] ;

[0098] First estimate of roll angle Corresponding quality indicators Second estimate of roll angle Corresponding quality indicators The calculation formula is:

[0099] ;

[0100] First estimate of roll angle weight Second estimate of roll angle weight for:

[0101] ;

[0102] Among them, the weighting coefficient , The sum is 1. , The sum is 1. Each weight coefficient is determined in real time based on the sea-line extraction quality index upon which the corresponding estimate depends; the higher the quality, the greater the weight. When the extraction quality of a camera in a certain direction drops sharply due to reasons such as wave obstruction or strong reflection, the weight of its corresponding estimate is automatically reduced to avoid contaminating the fusion result.

[0103] (3) Determine the weight of each estimate based on the root mean square error, perform weighted fusion of the first estimate and the second estimate, and output the final buoy pitch angle and roll angle.

[0104] Pitch angle after fusion Roll angle The calculation formula is:

[0105] .

[0106] (4) Use the pitch angle and roll angle to convert the solar vector in the camera coordinate system to the horizontal coordinate system to obtain the measured solar azimuth angle; combine the latitude and longitude and time provided by the global positioning system receiver, calculate the reference solar azimuth angle through ephemeris, and calculate the yaw angle of the buoy from the difference between the reference solar azimuth angle and the measured solar azimuth angle.

[0107] Utilizing the merged pitch angle and roll angle Construct rotation matrix The horizontal solar vector is obtained as follows:

[0108] ;

[0109] Measuring the solar azimuth angle:

[0110] ;

[0111] The latitude, longitude, and time are obtained using GPS, and the reference solar azimuth angle is calculated using a solar ephemeris position algorithm. ;

[0112] Calculate the yaw angle:

[0113] .

[0114] This embodiment verifies its superior performance through MATLAB simulation. As shown in Figure 8, the calculated roll angle error under steady-state conditions is ±0.63°, pitch angle error is ±0.63°, and yaw angle error is ±0.82°. Figure 7 As shown, the attitude angle error calculated using this method is small, and the estimated value tends to be consistent with the true value.

Claims

1. A method for measuring the attitude of an ocean buoy based on multi-view vision and solar vector, characterized in that, include: S1. At least four imaging modules are deployed at multiple horizontal positions of the buoy, with the optical axis of each imaging module pointing outward in the horizontal direction, for synchronously acquiring images of the sea-sky horizon; a zenith imaging module is deployed at the top of the buoy, with its optical axis pointing vertically upward, for acquiring images of the sun. S2. Extract the sea-line straight line from each of the sea-line images, obtain the intercept and slope of each sea-line in its respective image coordinate system, and calculate the root mean square error of the fitting of each sea-line straight line; at the same time, extract the center of the solar spot from the solar image and calculate the solar vector in the camera coordinate system. S3. Calculate the first estimated value of the pitch angle and the second estimated value of the roll angle based on the intercept difference and slope of the sea-line line extracted by the first imaging module. Calculate the first estimated value of the roll angle and the second estimated value of the pitch angle based on the intercept difference and slope of the sea-line line extracted by the second imaging module. The first imaging module and the second imaging module are perpendicular to each other. S4. Determine the weight of each estimate based on the root mean square error, perform weighted fusion of the first estimate and the second estimate, and output the final buoy pitch angle and roll angle. S5. Using the pitch angle and roll angle, the solar vector in the camera coordinate system is transformed to the horizontal coordinate system to obtain the measured solar azimuth angle; combined with the latitude, longitude and time provided by the global positioning system receiver, the reference solar azimuth angle is calculated through ephemeris, and the yaw angle of the buoy is calculated from the difference between the reference solar azimuth angle and the measured solar azimuth angle.

2. The method for measuring the attitude of an ocean buoy based on multi-view vision and solar vectors according to claim 1, characterized in that, In step S1, the at least four imaging modules are deployed at intervals along the circumference of the buoy, including four imaging modules deployed on the front, rear, left and right sides of the buoy, with the optical axis azimuth angles of adjacent imaging modules differing by 90°; the zenith imaging module is installed at the center of the top of the buoy.

3. The method for measuring the attitude of an ocean buoy based on multi-view vision and solar vectors according to claim 1, characterized in that, In step S3, the first imaging module includes two imaging modules deployed on the front and rear sides of the buoy, and the second imaging module includes two imaging modules deployed on the left and right sides of the buoy. The first estimated value of the pitch angle is calculated from the difference in intercepts of the sea-line lines of the front and rear imaging modules, and the first estimated value of the roll angle is calculated from the difference in intercepts of the sea-line lines of the left and right imaging modules. The second estimated value of the pitch angle is obtained by converting the slope of the sea-line line of the left and right imaging modules, and the second estimated value of the roll angle is obtained by converting the slope of the sea-line line of the front and rear imaging modules.

4. The method for measuring the attitude of an ocean buoy based on multi-view vision and solar vectors according to claim 3, characterized in that, The formulas for calculating the first estimated values ​​of the pitch angle and the first estimated value of the roll angle are as follows: ; in, This is the first estimated value of the pitch angle. This is the first estimated value for the roll angle. , , , The intercepts of the sea-line spectrum extracted from the front, rear, left, and right imaging modules are respectively. This refers to the pixel focal length of the imaging module.

5. The method for measuring the attitude of an ocean buoy based on multi-view vision and solar vectors according to claim 3, characterized in that, The formulas for calculating the second estimated values ​​of the pitch angle and the second estimated value of the roll angle are as follows: ; in, The second estimated value of the pitch angle. Second estimate of roll angle , , , The slope of the sea-line is extracted from the imaging modules on the right, left, front, and rear sides.

6. The method for measuring the attitude of an ocean buoy based on multi-view vision and solar vectors according to claim 1, characterized in that, In step S4, the weights of each estimated value are determined based on the root mean square error, including: Define the first estimate of the pitch angle Dependency quality Second estimate of pitch angle Dependency quality for: ; First estimate of pitch angle weight Second estimate of pitch angle weight for: ; in, , , , These represent the root mean square error when the front, rear, left, and right imaging modules extract the sea-line area, respectively.

7. The method for measuring the attitude of an ocean buoy based on multi-view vision and solar vectors according to claim 1, characterized in that, In step S4, determining the weights of each estimated value based on the root mean square error further includes: First estimate of roll angle Corresponding quality indicators Second estimate of roll angle Corresponding quality indicators ; ; First estimate of roll angle weight Second estimate of roll angle weight for: 。 8. The method for measuring the attitude of an ocean buoy based on multi-view vision and solar vectors according to claim 1, characterized in that, In step S4, the first estimate and the second estimate are weighted and fused using the following formula: ; in, For the merged pitch angle, This is the combined roll angle.

9. The method for measuring the attitude of an ocean buoy based on multi-view vision and solar vectors according to claim 1, characterized in that, The calculation process for the yaw angle is as follows: Obtain the measured solar vector in the camera coordinate system Using pitch angle and roll angle Construct rotation matrix The horizontal solar vector is obtained as follows: ; Measuring the solar azimuth angle: ; in, , This refers to the horizontal component of the horizontal solar vector. The latitude, longitude, and time are obtained through the Global Positioning System, and the reference solar azimuth angle is calculated using a solar ephemeris position algorithm. ; Calculate the yaw angle: 。 10. The method for measuring the attitude of an ocean buoy based on multi-view vision and solar vectors according to claim 1, characterized in that, In step S2, the root mean square error is calculated as follows: for the N edge points participating in the sea-line line fitting, the vertical residual of each point relative to the fitted line is calculated, and then the root mean square value of the vertical residual is calculated.