Water wave surface measurement software and hardware system based on multi-line structured light and monocular camera

Abstract

The invention discloses a water wave surface measurement software and hardware system based on multi-line structured light and a monocular camera, and belongs to the technical field of ship wave surface measurement. The method comprises the following steps: building and calibrating a system, arranging a multi-line laser and a monocular camera to form a hardware system, using a checkerboard marker to fix camera parameters and fitting a laser plane equation; water wave image acquisition: acquiring laser stripe images reflected by water waves; preprocessing the image, and enhancing and denoising the image; stripe center extraction and sorting: extracting stripe center coordinates with sub-pixel precision and carrying out clustering sorting to correspond to a laser plane; and three-dimensional wave surface reconstruction: converting the coordinates into three-dimensional points and carrying out interpolation to obtain wave surface morphology data. According to the method, the multi-line structured light is combined with the monocular vision, so that the interference of the broken wave surface is overcome, the problem of feature matching is avoided, and the high-precision and high-robustness measurement of the high-dynamic water wave transient full-field three-dimensional morphology is realized.

Description

Technical Field

[0001] This invention relates to the field of ship wave surface measurement technology, specifically to a hardware and software system for water wave surface measurement based on multi-line structured light and a monocular camera. Background Technology

[0002] When a ship is sailing at high speed, complex wave-making phenomena are generated around the hull, especially in the near-field region where highly nonlinear and unsteady air-water two-phase flows, such as wave breaking, occur. These breaking waves have a crucial impact on the ship's drag performance and fluid stealth. Therefore, in hull optimization design, it is necessary to accurately measure the breaking waves in the near-field of the ship in a large towing tank.

[0003] Current measurement technologies are mainly divided into single-point and field measurement technologies. Single-point measurement technologies, such as servo wave height meters and ultrasonic wave height meters, can only acquire wave height data that changes over time at a single point, and cannot capture the three-dimensional wave surface morphology of the entire transient field. In addition, in the highly dynamic environment of wave breaking generating splash droplets and cavitation, these single-point instruments are susceptible to interference, resulting in severe distortion of measurement data.

[0004] Stereo vision-based field measurement techniques are an effective means of acquiring instantaneous 3D wavefront morphology. Existing solutions include: directly using the reflected light spots on the water wave surface as feature points for binocular matching, but this method is suitable for large-scale, small-view-angle scenes, and feature matching is difficult in the narrow view of a towed pool; releasing foam or particles into the water to create artificial textures, but this causes water pollution; projecting lattice structured light onto the wavefront, however, the droplets and cavitation generated by the broken wavefront can seriously interfere with the recognition of discrete light points. Another approach is to use a binocular camera with a single-line laser to reconstruct the 3D morphology by matching the laser stripes in two images, but this method suffers from a high feature matching error rate due to differences in the viewing angles and reflections of the two cameras.

[0005] In summary, existing technologies generally face the challenge of simultaneously achieving full-field, transient, high-precision, and anti-interference measurements when dealing with highly dynamic and strongly interfered broken wave measurements. Summary of the Invention

[0006] The purpose of this invention is to provide a hardware and software system for measuring water wave surface based on multi-line structured light and a monocular camera, so as to overcome the shortcomings of the prior art and achieve high-precision, anti-interference measurement of the transient three-dimensional shape of the breaking waves of high-speed ships.

[0007] To achieve the above objectives, the technical solution adopted by the present invention is as follows: A method for measuring the wavefront of water waves based on multi-line structured light and a monocular camera includes the following steps: S1. System setup and calibration: Deploy a multi-line laser that projects multiple parallel laser lines and a high-speed monocular camera to form a measurement hardware system. Use a checkerboard target to take pictures at different spatial positions in the measured area to calibrate the internal and external parameters of the camera and fit the laser plane equation corresponding to each laser line. S2. Water wave image acquisition: Remove the target, project the laser line emitted by the multi-line laser onto the surface of the water wave to be measured, and acquire the laser stripe image reflected from the water wave surface through the monocular camera; S3. Image preprocessing: The acquired original image of laser stripes is preprocessed to enhance stripe features and suppress interference. The preprocessing includes contrast enhancement, background suppression, response normalization, and disconnection of lines. S4. Stripe Center Extraction and Sorting: Extract the pixel coordinates of the center line of each laser stripe from the preprocessed image with sub-pixel precision, and cluster and sort the pixel coordinates belonging to the same laser line to establish a correspondence with the laser plane equation. S5. Three-dimensional wavefront reconstruction: Based on the internal and external parameters of the camera and the laser plane equation, the pixel coordinates of the center line of each laser stripe are converted into three-dimensional spatial coordinates, and two-dimensional interpolation is performed to obtain three-dimensional wavefront topography data.

[0008] Furthermore, the calibration of the camera's internal and external parameters described in S1 specifically includes: A checkerboard target was placed in the test area, and target images were acquired at multiple different positions and orientations, with the laser off and the laser on respectively. The checkerboard corner points are identified from the target image with the laser off. The camera's internal parameter matrix and distortion coefficients are calculated based on Zhang Zhengyou's calibration method, as well as the target's external parameter matrix at different positions. Extract the center subpixel coordinates of each laser line projected onto the target plane from the target image with the laser turned on; Based on the external parameter matrix, the center sub-pixel coordinates of each laser line are transformed to the world coordinate system to obtain a three-dimensional spatial point set; By performing plane fitting on the three-dimensional spatial point set generated by the same laser line at different target positions, the laser plane equation corresponding to each laser line is obtained.

[0009] Furthermore, the extraction process for the center sub-pixel coordinates of each laser line is as follows: using the Steger algorithm based on the Hessian matrix, the normal direction of the stripes is determined by calculating the second derivative of the image, and sub-pixel positioning is performed in this direction to obtain the accurate center point coordinates.

[0010] Furthermore, the background suppression described in S3 uses morphological opening operations to extract the non-uniform background image and subtracts the original image from the background image to eliminate the influence of uneven illumination.

[0011] Furthermore, the response normalization described in S3 specifically includes: Calculate the Hessian response matrix of the image; Gaussian filtering is applied to the Hessian response matrix to obtain its smoothed result; The Hessian response matrix and its smoothing result are divided at the pixel level to enhance the stripe feature.

[0012] Furthermore, the pixel coordinates of the center line of each laser stripe mentioned in S4 are extracted as follows: using the Steger algorithm based on the Hessian matrix, the normal direction of the stripe is determined by calculating the second derivative of the image, and sub-pixel positioning is performed in this direction to obtain the accurate center point coordinates.

[0013] Furthermore, the clustering described in S4 employs a neighborhood search algorithm, which searches the eight neighborhoods of each central pixel to merge connected points into the same laser stripe.

[0014] Furthermore, the sorting described in S4 involves fitting the pixel coordinates of each laser stripe with a quadratic curve and sorting them according to the relative position of the fitted curve and the image edge to match the corresponding laser plane equation.

[0015] Another object of the present invention is to provide a water wave surface measurement system based on multi-line structured light and a monocular camera, wherein the system, when executed, implements the aforementioned water wave surface measurement method based on multi-line structured light and a monocular camera, including: The hardware subsystem includes a multi-line laser for projecting multiple parallel laser lines onto the water wave surface, a high-speed monocular camera for acquiring laser stripe images, and a workstation for controlling the camera and storing and processing data. The software subsystem, integrated into the workstation, is used to perform camera and laser plane calibration, image preprocessing, stripe center extraction and sorting, and three-dimensional wavefront reconstruction algorithms.

[0016] Another object of the present invention is to provide an electronic device, including a memory and a processor, wherein the memory is used to store a computer program, and the processor runs the computer program to enable the electronic device to perform the aforementioned method for measuring water wave surface based on multi-line structured light and a monocular camera.

[0017] The present invention provides a hardware and software system for water wave surface measurement based on multi-line structured light and a monocular camera, which has the following significant advantages compared with the prior art: 1. Strong anti-interference ability and applicable to complex broken wave surfaces: This invention uses line structured light as a feature source. Compared with dot matrix light or natural light spots, continuous laser stripes are more robust in high dynamic broken wave surfaces with droplet and cavitation interference. Features are not easily lost, solving the technical problem that traditional stereo vision feature points are easily interfered with.

[0018] 2. Avoids the problem of binocular matching and achieves high reconstruction accuracy: This invention adopts a monocular vision scheme, which uses the pre-calibrated laser plane equation as a spatial constraint to directly convert the two-dimensional image coordinates into three-dimensional world coordinates. This completely avoids the feature matching error problem caused by differences in viewing angle and reflection in binocular vision, and significantly improves the accuracy and reliability of three-dimensional reconstruction.

[0019] 3. Achieves transient full-field measurement and is environmentally friendly: This invention can acquire three-dimensional wavefront data of the entire laser coverage area in a single measurement, achieving transient full-field measurement and overcoming the limitations of single-point wave height meters. Furthermore, compared to methods that release pollutants, this invention does not alter the aquatic environment, making it more environmentally friendly.

[0020] 4. Simple system structure and low cost: Using a monocular camera instead of a binocular camera reduces the complexity and cost of the hardware system, while simplifying the software algorithm process. Attached Figure Description

[0021] Figure 1 This is a flowchart of the method of the present invention; Figure 2 This is a schematic diagram of the system composition of the present invention; Figure 3 This is a schematic diagram of the original stripe image obtained in an embodiment of the present invention; Figure 4 This is a schematic diagram of the striped image after preprocessing according to an embodiment of the present invention; Figure 5 This is a schematic diagram of the line feature pixel coordinates extracted in an embodiment of the present invention; Figure 6 This is a three-dimensional coordinate schematic diagram of the line feature in an embodiment of the present invention; Figure 7 This is a schematic diagram of the water wave surface morphology after two-dimensional interpolation according to an embodiment of the present invention; In the picture: 1-Laser, 2-DOE lens, 3-Camera, 4-Workstation. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0023] This embodiment provides a method for measuring the wavefront of water waves based on multi-line structured light and a monocular camera, such as... Figure 1 As shown, it includes the following steps: System setup and calibration: A multi-line laser that projects multiple parallel laser lines and a high-speed monocular camera are set up to form a measurement hardware system. A checkerboard target is used to take pictures at different spatial positions in the area to be measured in order to calibrate the internal and external parameters of the camera and fit the laser plane equation corresponding to each laser line. Water wave image acquisition: Remove the target, project the laser line emitted by the multi-line laser onto the surface of the water wave to be measured, and acquire the laser stripe image reflected from the water wave surface through the monocular camera; Image preprocessing: The acquired original image of the laser stripes is preprocessed to enhance stripe features and suppress interference. The preprocessing includes contrast enhancement, background suppression, response normalization, and disconnection of lines. Stripe center extraction and sorting: Extract the pixel coordinates of the center line of each laser stripe from the preprocessed image with subpixel precision, and cluster and sort the pixel coordinates belonging to the same laser line to establish a correspondence with the laser plane equation; Three-dimensional wavefront reconstruction: Based on the internal and external parameters of the camera and the laser plane equation, the pixel coordinates of the center line of each laser stripe are converted into three-dimensional spatial coordinates, and two-dimensional interpolation is performed to obtain three-dimensional wavefront topography data.

[0024] The following is a detailed description with reference to specific embodiments.

[0025] S1. System Setup and Calibration: A multi-line laser that projects multiple parallel laser lines and a high-speed monocular camera are set up to form a measurement hardware system. A checkerboard target is used to take pictures at different spatial positions in the measured area to calibrate the internal and external parameters of the camera and fit the laser plane equation corresponding to each laser line.

[0026] like Figure 2As shown, the hardware system of this invention mainly includes: a multi-line laser 1, a high-speed monocular camera 3, and a data processing workstation 4. Additionally, a checkerboard target is required for calibration. The multi-line laser 1 has a DOE lens 2 installed at its front end, which converts a single laser beam into multiple parallel laser lines, projecting them onto the water wave area to be measured. The camera 3 is controlled by the workstation 4 and is responsible for image acquisition. First, the relative position and angle of the laser 1 and camera 3 need to be adjusted to ensure that the camera's field of view clearly covers all laser stripes projected onto the water wave area, and to minimize overexposure caused by water wave reflections.

[0027] Calibration is crucial for ensuring measurement accuracy. A checkerboard target is placed in the measurement area, and within the desired water wave space, the target is moved to different positions (at least ten) and orientations. At each position, an image containing only the target is first taken with the laser off, and then an image containing laser stripes on the target is taken with the laser on. Using the Zhang Zhengyou calibration method, the camera's internal parameters (focal length, principal point, distortion coefficient, etc.) and external parameters (the camera's relative position to the world coordinate system at each position) are accurately calibrated using the corner point information from the target images taken at all positions. Next, the center pixel coordinates of each laser line are accurately extracted from the target image with laser stripes using the Hessian-Steger algorithm. Using the calibrated camera external parameters, these two-dimensional pixel coordinates are back-projected into three-dimensional space. Since the target plane is known, the three-dimensional point set on which the laser line falls on the target plane can be obtained. Finally, plane fitting is performed on the three-dimensional point sets generated by the same laser line at different target positions to obtain the accurate spatial plane equation of the laser line. This process is repeated to fit individual plane equations for all laser lines and number them sequentially. This process, through precise offline calibration, provides reliable geometric constraints for subsequent monocular 3D reconstruction, fundamentally avoiding the unreliability of online matching. The calibration process is as follows: 1. Place a checkerboard target within the test area, and acquire target images at multiple different positions and orientations, with the laser off and on respectively.

[0028] (1) Fix the checkerboard target at a certain position in the area to be measured to ensure that the camera can capture all corners of the checkerboard. (2) Turn off the laser, fix the target in a certain spatial position, and use the camera to take and save the target image without laser lines; (3) Turn on the laser and project the laser line onto the target. At the same time, the camera takes a picture of the target with the laser line. (4) Move the target position and repeat (2) and (3) above to take pictures of the target patterns with the laser lines turned on and off at least ten different spatial positions; 2. Identify checkerboard corner points from the target image with the laser off. Calculate the camera's intrinsic parameter matrix and distortion coefficients, as well as the target's extrinsic parameter matrix at different positions, based on Zhang Zhengyou's calibration method. First, perform corner point identification on the target pattern under the laser line being turned off. The corner point identification algorithm is as follows: A window of a specific size is moved across the image in various directions, and the autocorrelation function of the gray-level changes within the window during the movement is calculated, as shown in equation (1): in, Let (u,v) be the camera pixel coordinates, (u,v) be the window size, w be the window weight set to 1, and I be the image pixel grayscale value. After Taylor expansion, the autocorrelation function E can be written as: The formula for calculating matrix M is as follows: in and These are the gradients of the pixel grayscale value in the x and y directions, respectively.

[0029] Define a corner response function R to determine corners: Where traceM is the trace of matrix M, and detM is the rank of matrix M. and Let be the eigenvalues ​​of matrix M, and k be an empirical constant, typically taken as 0.04 to 0.06.

[0030] The eigenvalues ​​λ1 and λ2 of matrix M reflect the gradient intensity of the pixel in different directions, and the corner response function R can be defined accordingly. When R is detected to be greater than 0, the corner position can be located and the corner coordinates can be extracted.

[0031] Then, the internal and external parameters of the camera are calibrated using corner coordinates. The calibration method adopts Zhang Zhengyou's calibration algorithm, and the specific process is as follows: The mapping relationship between camera coordinates (x, y) and physical coordinates (X, Y, Z) is represented by the pinhole model, that is: Where K is the intrinsic parameter matrix, and its form is as follows: Where α and β are the fusion of focal length and aspect ratio; γ is the radial distortion coefficient; and R represents the position of the camera principal point in the x and y directions in the image coordinate system, which is the coordinate of the center point of the image; R and t are external parameter matrices, which are the rotation matrix and translation vector, respectively.

[0032] To solve for the internal and external parameter matrices, multiple known pixel coordinates and physical space points need to be substituted into equation (7). This is achieved by substituting the identified checkerboard pixel coordinates and known physical space locations (with the checkerboard's Z-plane set to 0 and a corner point designated as the origin for calculation) into equation (7). Finally, the accuracy of the internal and external parameters is optimized using maximum likelihood estimation.

[0033] 3. Extract the center sub-pixel coordinates of each laser line projected onto the target plane from the target image with the laser activated. The extraction process for the center sub-pixel coordinates of each laser line is as follows: Using the Steger algorithm based on the Hessian matrix, the normal direction of the stripes is determined by calculating the second derivative of the image, and sub-pixel positioning is performed in this direction to obtain accurate center point coordinates. Select one laser line from the target image projecting the laser line for line segment extraction. Use the Hessian-Steger algorithm to extract the center sub-pixel coordinates of the laser line. The Hessian-Steger algorithm is a centerline extraction method based on image processing. The core idea of ​​this algorithm is to determine the normal direction of the light stripe using the Hessian matrix, and then apply Taylor polynomial expansion in this direction to find the sub-pixel position of the light stripe's center. The Hessian matrix is ​​a square matrix composed of the second-order partial derivatives of a multivariate function, which can be used to determine the extrema of a multivariate function. Its form is as follows: in, , , , These are the second-order gradient information of the pixel in the x and y directions, respectively.

[0034] In practice, a Gaussian kernel is used to convolve the image to obtain the second-order partial derivatives and the Hessian matrix. The eigenvalues ​​and eigenvectors are then calculated to obtain the normal direction for each pixel. A Taylor expansion is performed on the image's grayscale distribution along the normal direction, ignoring higher-order terms, to obtain expressions for the first and second-order terms. By setting the first derivative of the Taylor expansion to zero, the sub-pixel position of the ray center is obtained.

[0035] 4. Based on the external parameter matrix, transform the center pixel coordinates of each laser line to the world coordinate system to obtain a three-dimensional spatial point set. After extracting the light stripe images left by the same laser line on different target planes according to the above steps, the center pixel coordinates of the laser line (… , Substituting the aforementioned camera internal and external parameters into equation (7) yields the physical space coordinates at different heights. , , .

[0036] 5. Perform plane fitting on the three-dimensional spatial point set generated by the same laser line at different target positions to obtain the laser plane equation corresponding to each laser line.

[0037] The laser plane equation is fitted based on these physical space coordinates, and its general form is: in , , , The parameters to be fitted for the equation of the i-th laser surface.

[0038] The above extraction and fitting process is performed on all laser lines to obtain the fitting parameters for the plane equations of all laser lines, and these laser lines are numbered in order of their distance from the image edge.

[0039] S2. Water Wave Image Acquisition: Remove the target and project the laser line emitted by the multi-line laser onto the surface of the water wave to be measured. Acquire the laser stripe image reflected from the water wave surface using the monocular camera. After calibration, remove the target. Generate ship-induced waves in the measured area and project the laser emitted by the multi-line laser onto the area of ​​wave breaking. Activate the high-speed monocular camera controlled by the workstation to capture and save the laser stripe image reflected from the water wave surface at a high frame rate. During this process, the laser stripes will be deformed and discontinuous due to the undulation, breaking, and splashing of the wave surface. Figure 3 This is the original image of the laser stripes acquired in this embodiment.

[0040] S3. Image preprocessing: The acquired original image of the laser stripes is preprocessed to enhance stripe features and suppress interference. The preprocessing includes contrast enhancement, background suppression, response normalization, and disconnection of lines.

[0041] Due to the dynamic characteristics of the broken wavefront and optical interference, the original image ( Figure 3The stripes in the image may suffer from uneven brightness, high background noise, and broken stripes. Preprocessing aims to enhance stripe features. First, adaptive histogram equalization is used to enhance contrast, avoiding local overexposure caused by global processing. Second, morphological opening operations are used to extract the background image, and then the enhanced image is subtracted from the background image to effectively suppress non-uniform backgrounds. Then, the Hessian response matrix of the image is calculated and normalized to further enhance the feature response of the stripe region. Finally, for possible broken stripes, morphological closing operations of specific length and direction are used to connect them, ultimately obtaining an image with clear and continuous stripe features, such as... Figure 4 As shown, this series of preprocessing steps can significantly improve the robustness and accuracy of the subsequent stripe center extraction algorithm.

[0042] 1. The contrast of the grayscale image of laser stripes is adjusted by means of an adaptive histogram equalization method. This algorithm calculates the local histogram in blocks and equalizes it, avoiding the local overexposure / underexposure problem caused by global histogram equalization.

[0043] 2. Background suppression is performed on the image to eliminate non-uniform background interference and improve the grayscale distinction between the target area and the background. Background suppression employs morphological opening operations to extract the non-uniform background image, and then subtracts the original image from the background image to eliminate the influence of uneven illumination. First, a morphological opening operation is performed on the contrast-enhanced image obtained in step 1 to extract the background contour, where the structuring element is a disk wider than the stripe width. This effectively preserves the non-uniform distribution characteristics of the background while avoiding over-extraction of the target area. Then, pixel-level subtraction is performed between the contrast-enhanced image and the background image obtained through the morphological operation to obtain the image after eliminating the non-uniform background.

[0044] 3. Calculate the normalized response image and normalize the response of the striped target through Hessian analysis to obtain clearer stripe features. The response normalization specifically includes: calculating the Hessian response matrix of the image; performing Gaussian filtering on the Hessian response matrix to obtain its smoothed result; and performing pixel-level division operations on the Hessian response matrix and its smoothed result to enhance the stripe features.

[0045] First, calculate the Hessian matrix, which is composed of the second-order partial derivatives of the image grayscale values ​​in the x and y directions, as shown in Equation 9; then calculate the Hessian response matrix. The formula is: This formula quantifies the texture feature intensity of the striped target at a pixel by weighted fusion of the partial derivative components.

[0046] After that Gaussian filtering is performed to obtain the smoothed response matrix. The normalized response matrix is ​​obtained by performing pixel-level division on the two. The formula is For the normalized response matrix Dividing by the maximum value and setting values ​​less than 1 to 0, the normalized response matrix enhances the striped areas and displays clearer stripe features.

[0047] 4. Perform a disconnection operation on the obtained normalized response matrix, that is, apply a morphological closing operation to the normalized response matrix. The structuring element is a line of a certain length, which is less than the spacing between the stripes and the angle is basically consistent with the direction of the stripes.

[0048] S4. Stripe Center Extraction and Sorting: Pixel coordinates of the center lines of each laser stripe are extracted from the preprocessed image with sub-pixel precision. Pixel coordinates belonging to the same laser line are clustered and sorted to establish a correspondence with the laser plane equation. The clustering uses a neighborhood search algorithm, searching the eight neighbors of each center pixel to group connected points into the same laser stripe. The sorting uses a quadratic curve to fit the pixel coordinates of each laser stripe and sorts them according to the relative position of the fitted curve to the image edge to match the corresponding laser plane equation.

[0049] For the preprocessed image, the Hessian-Steger algorithm is applied again to extract the pixel coordinates of the stripe center lines with sub-pixel precision. The process is as follows: using the Steger algorithm based on the Hessian matrix, the normal direction of the stripes is determined by calculating the second derivative of the image, and sub-pixel localization is performed in this direction to obtain the precise center point coordinates. The result is as follows. Figure 5 As shown. Then, a neighborhood search algorithm is used to classify the spatially continuous center points, forming independent laser stripes. To correctly correspond with the laser plane during calibration, a quadratic curve is used to fit the pixel coordinates of each stripe, and the stripes are sorted according to their position in the image (e.g., distance from the image edge), thus establishing a one-to-one correspondence with the laser plane equations numbered in step S1. The innovation of this step lies in efficiently and accurately solving the problem of identifying and matching multiple stripes in a single image through curve fitting and sorting. Specifically: The same Hessian-Steger algorithm was used to extract laser stripes from the full-field image of the normalized response matrix to obtain the center coordinates of all laser stripes. The extracted stripe center coordinates are clustered. A neighborhood search algorithm is used to connect the center coordinates of laser stripes along the same laser line and distinguish between different laser lines. Starting from any stripe center pixel coordinate, the algorithm searches for another stripe center coordinate in its eight directions (up, down, left, right, and four diagonal directions). If another exists, it is included in the same laser stripe and removed from the list of undiscriminated pixel coordinates, and then used as the starting point for the next search. If no other stripe center coordinate exists in its eight neighborhood directions, the laser stripe is considered to have reached its endpoint. The clustered laser stripes are sorted, and the pixel coordinates of each identified laser line are fitted using a quadratic curve. The lines are then sorted according to their distance from the image edge, corresponding one-to-one with the aforementioned laser plane equation.

[0050] S5. Three-dimensional wavefront reconstruction: Based on the internal and external parameters of the camera and the laser plane equation, the pixel coordinates of the center line of each laser stripe are converted into three-dimensional spatial coordinates, and two-dimensional interpolation is performed to obtain three-dimensional wavefront topography data.

[0051] This is the final step in transforming two-dimensional information into a three-dimensional shape. For each sorted laser stripe, its pixel coordinates are normalized to the camera coordinate system using the camera intrinsic parameters obtained in step S1. Then, the normalized coordinates are intersected with the corresponding calibrated laser plane equation to directly calculate the three-dimensional spatial coordinates (X, Y, Z) of that point. After transforming all the points on all laser stripes, a three-dimensional point cloud representing the wavefront morphology is obtained, such as... Figure 6 As shown. Finally, two-dimensional interpolation algorithms such as bilinear interpolation are used to interpolate these discrete three-dimensional points, generating a smooth and complete meshed three-dimensional wavefront, as shown. Figure 7 As shown, this allows for a direct and quantitative display of the transient three-dimensional morphology of water waves. This process, through a unique combination of a monocular camera and a known laser plane, achieves high-precision three-dimensional reconstruction without feature matching, making it particularly suitable for complex scenes such as breaking waves. Specifically: For each sorted laser stripe, its pixel coordinates are normalized to the camera coordinate system using the camera intrinsic parameter matrix K, as shown in equation (13): in , For normalized planar coordinates, , These are the pixel coordinates of the center of the laser stripe.

[0052] Then, the normalized coordinates are intersected with the corresponding laser plane equation (Equation (10)) to directly calculate the three-dimensional spatial coordinates (X, Y, Z) of the point, as shown in Equation (14): After converting all the points on the laser stripes, a three-dimensional point cloud representing the wavefront morphology is obtained. Figure 6 Finally, two-dimensional interpolation algorithms such as bilinear interpolation are used to interpolate these discrete points, generating a smooth and complete meshed three-dimensional wavefront. Figure 7 This allows for a direct and quantitative representation of the transient three-dimensional morphology of water waves.

[0053] This embodiment provides a water wave surface measurement system based on multi-line structured light and a monocular camera. When executed, this system implements the aforementioned water wave surface measurement method based on multi-line structured light and a monocular camera, including: The hardware subsystem includes a multi-line laser for projecting multiple parallel laser lines onto the water wave surface, a high-speed monocular camera for acquiring laser stripe images, and a workstation for controlling the camera and storing and processing data. The software subsystem, integrated into the workstation, is used to perform camera and laser plane calibration, image preprocessing, stripe center extraction and sorting, and three-dimensional wavefront reconstruction algorithms.

[0054] This embodiment provides an electronic device, including a memory and a processor. The memory stores a computer program, and the processor runs the computer program to enable the electronic device to perform the water wave surface measurement method based on multi-line structured light and a monocular camera.

[0055] In summary, this invention effectively overcomes the interference of broken wavefronts and avoids the feature matching problem by projecting multi-line structured light and utilizing monocular vision combined with known laser plane constraints. It achieves high-precision and high-robust measurement of the transient full-field three-dimensional shape of high-dynamic water waves. At the same time, the system has a simple structure and is environmentally friendly.

[0056] Embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0057] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0058] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0059] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0060] Contents not described in detail in this specification are prior art known to those skilled in the art. It is hereby indicated that the above description is intended to help those skilled in the art understand this invention, but does not limit the scope of protection of this invention. Any equivalent substitutions, modifications, improvements, or simplifications of the above descriptions that do not depart from the essential content of this invention fall within the scope of protection of this invention.

Claims

1. A method for measuring the wavefront of water waves based on multi-line structured light and a monocular camera, characterized in that, Includes the following steps: S1. System setup and calibration: Deploy a multi-line laser that projects multiple parallel laser lines and a high-speed monocular camera to form a measurement hardware system. Use a checkerboard target to take pictures at different spatial positions in the measured area to calibrate the internal and external parameters of the camera and fit the laser plane equation corresponding to each laser line. S2. Water wave image acquisition: Remove the target, project the laser line emitted by the multi-line laser onto the surface of the water wave to be measured, and acquire the laser stripe image reflected from the water wave surface through the monocular camera; S3. Image preprocessing: The acquired original image of laser stripes is preprocessed to enhance stripe features and suppress interference. The preprocessing includes contrast enhancement, background suppression, response normalization, and disconnection of lines. S4. Stripe Center Extraction and Sorting: Extract the pixel coordinates of the center line of each laser stripe from the preprocessed image with sub-pixel precision, and cluster and sort the pixel coordinates belonging to the same laser line to establish a correspondence with the laser plane equation. S5. Three-dimensional wavefront reconstruction: Based on the internal and external parameters of the camera and the laser plane equation, the pixel coordinates of the center line of each laser stripe are converted into three-dimensional spatial coordinates, and two-dimensional interpolation is performed to obtain three-dimensional wavefront topography data.

2. The method for measuring water wave surface based on multi-line structured light and a monocular camera according to claim 1, characterized in that, The calibration of camera internal and external parameters described in S1 specifically includes: A checkerboard target was placed in the test area, and target images were acquired at multiple different positions and orientations, with the laser off and the laser on respectively. The checkerboard corner points are identified from the target image with the laser off. The camera's internal parameter matrix and distortion coefficients are calculated based on Zhang Zhengyou's calibration method, as well as the target's external parameter matrix at different positions. Extract the center subpixel coordinates of each laser line projected onto the target plane from the target image with the laser turned on; Based on the external parameter matrix, the center sub-pixel coordinates of each laser line are transformed to the world coordinate system to obtain a three-dimensional spatial point set; By performing plane fitting on the three-dimensional spatial point set generated by the same laser line at different target positions, the laser plane equation corresponding to each laser line is obtained.

3. The method for measuring water wave surface based on multi-line structured light and a monocular camera according to claim 2, characterized in that, The extraction process for the center sub-pixel coordinates of each laser line is as follows: using the Steger algorithm based on the Hessian matrix, the normal direction of the stripes is determined by calculating the second derivative of the image, and sub-pixel positioning is performed in this direction to obtain the accurate center point coordinates.

4. The method for measuring water wave surface based on multi-line structured light and a monocular camera according to claim 1, characterized in that, The background suppression described in S3 uses morphological opening operations to extract non-uniform background images and subtracts the original image from the background image to eliminate the influence of uneven illumination.

5. The method for measuring water wave surface based on multi-line structured light and a monocular camera according to claim 1, characterized in that, The response normalization described in S3 specifically includes: Calculate the Hessian response matrix of the image; Gaussian filtering is applied to the Hessian response matrix to obtain its smoothed result; The Hessian response matrix and its smoothing result are divided at the pixel level to enhance the stripe feature.

6. The method for measuring water wave surface based on multi-line structured light and a monocular camera according to claim 1, characterized in that, The pixel coordinates of the center line of each laser stripe mentioned in S4 are extracted as follows: the Steger algorithm based on the Hessian matrix is ​​used to determine the normal direction of the stripe by calculating the second derivative of the image, and sub-pixel positioning is performed in this direction to obtain the accurate center point coordinates.

7. The method for measuring water wave surface based on multi-line structured light and a monocular camera according to claim 1, characterized in that, The clustering described in S4 uses a neighborhood search algorithm to merge connected points into the same laser stripe by searching the eight neighborhoods of each central pixel.

8. The method for measuring water wave surface based on multi-line structured light and a monocular camera according to claim 1, characterized in that, The sorting described in S4 involves fitting the pixel coordinates of each laser stripe with a quadratic curve and sorting them according to the relative position of the fitted curve and the image edge to match the corresponding laser plane equation.

9. A water wave wavefront measurement system based on multi-line structured light and a monocular camera, characterized in that, When executed, this system implements the water wave surface measurement method based on multi-line structured light and a monocular camera as described in any one of claims 1-8, comprising: The hardware subsystem includes a multi-line laser for projecting multiple parallel laser lines onto the water wave surface, a high-speed monocular camera for acquiring laser stripe images, and a workstation for controlling the camera and storing and processing data. The software subsystem, integrated into the workstation, is used to perform camera and laser plane calibration, image preprocessing, stripe center extraction and sorting, and three-dimensional wavefront reconstruction algorithms.

10. An electronic device, characterized in that: It includes a memory and a processor, the memory being used to store computer programs, and the processor running the computer programs to enable the electronic device to perform the water wave surface measurement method based on multi-line structured light and a monocular camera as described in any one of claims 1-8.

Images