A method, apparatus, and electronic equipment for segmenting point cloud at the bevel of a saddle-shaped weld in a thick plate.

Abstract

This application belongs to the field of weld recognition technology, specifically disclosing a point cloud segmentation method, device, and electronic equipment for thick plate saddle-shaped weld grooves. The method includes: performing dimensionality reduction processing on the acquired three-dimensional point cloud data of the saddle-shaped weld groove to obtain target plane point cloud data; assigning position numbers to each point in the target plane point cloud data, and determining multiple segmentation points from the numbered target plane point cloud data to determine multiple segmentation segments; iteratively optimizing the selection position of each segmentation point with the objective function of minimizing the objective function, until each optimal segmentation point is determined; the objective function is determined based on the endpoint fitting deviation and slope consistency deviation of each segmentation segment; and segmenting the three-dimensional point cloud data according to each optimal segmentation point. This application can significantly improve the point cloud segmentation accuracy for saddle-shaped weld grooves, accurately extracting point cloud data from different parts of the saddle-shaped weld groove.

Description

Technical Field

[0001] This application belongs to the field of weld seam recognition technology, and more specifically, relates to a method, device and electronic equipment for segmenting point cloud of saddle-shaped weld seam groove in thick plates. Background Technology

[0002] Currently, laser vision recognition mainly focuses on straight welds and fillet welds in single-layer, single-pass, and single-layer multi-pass flat welds. These weld bevels exhibit a V-shape or I-shape and have a symmetrical structure. In existing technologies, by setting the number of point clouds for each part and using a simple linear fitting method, the point cloud data of the aforementioned regular weld bevels can be effectively segmented to identify weld defects.

[0003] However, in actual industrial production, we often encounter saddle-shaped weld grooves on thick plates with multiple layers and varying cross-sections that intersect spatially. These weld grooves exhibit spatially varying curves, complex geometry, spatial diversity, and variable groove opening directions, resulting in irregular cross-sections. Due to the combined influence of these factors, the segmentation accuracy of existing segmentation algorithms for 3D point cloud data of saddle-shaped weld grooves is low, making effective segmentation difficult.

[0004] Therefore, how to effectively achieve point cloud segmentation of the saddle-shaped weld groove of thick plates has become a technical problem that the industry urgently needs to solve. Summary of the Invention

[0005] To address the shortcomings of existing technologies, the purpose of this application is to effectively segment the point cloud of saddle-shaped weld grooves in thick plates, aiming to solve the problem that existing segmentation algorithms have low segmentation accuracy and are difficult to achieve effective segmentation when segmenting 3D point cloud data of saddle-shaped weld grooves.

[0006] To achieve the above objectives, in a first aspect, this application provides a method for segmenting the point cloud of a saddle-shaped weld groove in a thick plate, comprising: The 3D point cloud data of the acquired saddle-shaped weld groove is reduced in dimension to obtain the target plane point cloud data. Each point in the target plane point cloud data is assigned a position number, and multiple segmentation points are determined from the numbered target plane point cloud data to determine multiple segmentation segments; With the objective function as the goal, the selection position of each segmentation point is iteratively optimized until the optimal segmentation point is determined; the objective function is determined based on the endpoint fitting deviation and slope consistency deviation of each segment. The 3D point cloud data is segmented based on each of the optimal segmentation points.

[0007] Optionally, the step of iteratively optimizing the selection position of each segmentation point with the objective function as the goal, until each optimal segmentation point is determined, includes: Step S101: Under the preset position constraint, the selected position of each of the segmentation points in the current iteration process is traversed and adjusted according to the preset displacement strategy, and the function value of the objective function corresponding to the selected position adjustment of all the segmentation points is calculated each time. Step S102: Determine the optimal displacement corresponding to the minimum objective function value among all the function values; Step S103: Determine each of the segmentation points in the next iteration process based on the current selected position of each segmentation point and the optimal displacement. Repeat steps S101 to S103 until the function value of the objective function satisfies the preset convergence condition, and determine each optimal split point.

[0008] Optionally, the preset displacement strategy is determined by equally dividing the step size function in the current iteration process; the step size function is determined based on the total number of points in the target plane point cloud data and the attenuation factor in the current iteration process, and the attenuation factor is determined based on the minimum objective function value in the previous iteration process and the initial function value of the objective function.

[0009] Optionally, before iteratively optimizing the selection position of each segmentation point with the objective function as the goal until each optimal segmentation point is determined, the method further includes: Based on the coordinate information of each segmentation point and the coordinate information of each point within each segment, the endpoint fitting deviation and slope consistency deviation of each segment are determined. The objective function is determined based on the endpoint fitting deviation and slope consistency deviation of each segment, as well as the preset segment length penalty term parameter.

[0010] Optionally, the step of determining the endpoint fitting deviation of each segment specifically includes: Based on the coordinate information of each segmentation point, a linear fitting function for each segment is determined; Determine the amplitude and the number of points within each segment; Based on the coordinate information of each point within each segment, the linear fitting function of each segment, the magnitude, and the number of points within the segment, the endpoint fitting deviation of each segment is determined.

[0011] Optionally, the step of determining the slope consistency deviation of each segment specifically includes: Based on the coordinate information of each segmentation point, the endpoint fitting slope of each segment is determined; Based on the coordinate information of each point within each segment, the least squares slope of each segment is determined; Based on the endpoint fitting slope and least squares slope of each segment, the slope consistency deviation of each segment is determined.

[0012] Optionally, the step of performing dimensionality reduction processing on the acquired saddle-shaped weld groove three-dimensional point cloud data to obtain target plane point cloud data includes: The point cloud data is processed by decentralizing the point cloud coordinates to obtain decentralized three-dimensional point cloud data. Singular value decomposition is performed on the decentralized 3D point cloud data to determine the target rotation matrix; Based on the target rotation matrix, the three-dimensional point cloud data is projected onto the target plane to obtain the target plane point cloud data.

[0013] Secondly, this application provides a point cloud segmentation device for the bevel of a thick plate saddle-shaped weld, comprising: The dimension reduction module is used to perform dimension reduction processing on the acquired 3D point cloud data of the saddle-shaped weld groove to obtain the target plane point cloud data. The first processing module is used to assign a position number to each point in the target plane point cloud data, and to determine multiple segmentation points from the numbered target plane point cloud data to determine multiple segmentation segments; The second processing module is used to iteratively optimize the selection position of each segmentation point with the goal of minimizing the objective function, until the optimal segmentation point is determined; the objective function is determined based on the endpoint fitting deviation and slope consistency deviation of each segmentation segment. The segmentation module is used to segment the three-dimensional point cloud data according to each of the optimal segmentation points.

[0014] Thirdly, this application provides an electronic device, comprising: at least one memory for storing a program; and at least one processor for executing the program stored in the memory, wherein when the program stored in the memory is executed, the processor is configured to execute the method described in the first aspect or any possible implementation thereof.

[0015] Fourthly, this application provides a computer-readable storage medium storing a computer program that, when run on a processor, causes the processor to perform the method described in the first aspect or any possible implementation thereof.

[0016] Fifthly, this application provides a computer program product that, when run on a processor, causes the processor to perform the method described in the first aspect or any possible implementation thereof.

[0017] It is understood that the beneficial effects of the second to fifth aspects mentioned above can be found in the relevant descriptions in the first aspect mentioned above, and will not be repeated here.

[0018] Overall, the technical solutions conceived in this application have the following beneficial effects compared with the prior art: This application provides a point cloud segmentation method, device, and electronic equipment for thick plate saddle-shaped weld grooves. It utilizes three-dimensional point cloud data of spatially intersecting multi-layered, multi-channel variable cross-section saddle-shaped weld grooves obtained through frame-by-frame scanning using online laser vision. The point cloud data is then converted to a target plane for dimensionality reduction analysis. Based on the spatial variation curve characteristics of the saddle-shaped weld groove, the point cloud data is divided into multiple segments, and linear fitting is performed on each segment. Simultaneously, a composite objective function is constructed using a combination of deviation estimation of the fitted straight line at the endpoints of the segment and a slope consistency metric to achieve multi-objective optimization. This overcomes the problem of identifying irregular contours of the saddle-shaped weld groove cross-section, significantly improving the point cloud segmentation accuracy for saddle-shaped weld grooves and accurately extracting point cloud data from different parts of the saddle-shaped weld groove. Attached Figure Description

[0019] Figure 1 This is a flowchart illustrating the point cloud segmentation method for the bevel of a thick plate saddle-shaped weld provided in the embodiments of this application. Figure 2 This is a schematic diagram of the dimensionality reduction result of the point cloud data of the saddle-shaped weld groove provided in the embodiments of this application; Figure 3 (a) is a real-world schematic diagram of laser scanning point cloud data acquisition for thick plate saddle weld groove provided in the embodiment of this application, and (b) is a schematic diagram of feature extraction for point cloud data of thick plate saddle weld groove provided in the embodiment of this application. Figure 4 (a) in this application is a variable cross-section saddle-shaped weld groove provided in an embodiment of this application. x - z A schematic diagram of the segmentation results of point cloud data on a plane. (b) is another variable cross-section saddle-shaped weld groove provided in the embodiment of this application. x - z A schematic diagram of the segmentation results of point cloud data on a plane; Figure 5 This is a schematic diagram of the structure of the thick plate saddle-shaped weld groove dot cloud segmentation device provided in the embodiments of this application; Figure 6 This is a schematic diagram of the structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0021] The terms "first" and "second," etc., used in the specification and claims of this application are used to distinguish different objects, not to describe a specific order of objects. For example, "first processing module" and "second processing module," etc., are used to distinguish different processing modules, not to describe a specific order of processing modules.

[0022] In the embodiments of this application, the terms "exemplary" or "for example" are used to indicate that something is an example, illustration, or description. Any embodiment or design that is described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design. Specifically, the use of the terms "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0023] The embodiments of this application are described below with reference to the accompanying drawings.

[0024] Figure 1 This is a flowchart illustrating the point cloud segmentation method for the saddle-shaped weld groove of thick plates provided in the embodiments of this application, as shown below. Figure 1 As shown, it includes: Step S1: Dimensionality reduction processing is performed on the obtained 3D point cloud data of the saddle-shaped weld groove to obtain the target plane point cloud data. Step S2: Number the location of each point in the target plane point cloud data, and determine multiple segmentation points from the numbered target plane point cloud data to determine multiple segmentation segments; Step S3: With the goal of minimizing the objective function, iteratively optimize the selection position of each segmentation point until the optimal segmentation point is determined; the objective function is determined based on the endpoint fitting bias and slope consistency bias of each segment. Step S4: Segment the 3D point cloud data according to each optimal segmentation point.

[0025] Specifically, the target plane point cloud data described in this application embodiment refers to the point cloud data obtained by projecting the three-dimensional point cloud data of the saddle-shaped weld groove onto the target plane. The target plane can be set according to the direction of the laser scan, ensuring that the target plane point cloud data includes the height information and overall shape of the saddle-shaped weld groove. This facilitates segmentation based on a single frame of laser scan point cloud data each time.

[0026] The segmentation points described in the embodiments of this application refer to the points in the target plane point cloud data that are numbered and used to segment the entire target plane point cloud data.

[0027] In other words, after determining each segmentation point, the entire target planar point cloud data can be segmented according to the position number of each segmentation point. Thus, the entire point cloud covered by the first point to the first segmentation point (inclusive) in the numbered target planar point cloud data can be considered the first segment; the entire point cloud covered by the first segmentation point to the second segmentation point (inclusive) can be considered the second segment; the entire point cloud covered by the second segmentation point to the third segmentation point (inclusive) can be considered the third segment, and so on, dividing the entire target planar point cloud data into multiple segments.

[0028] In embodiments of this application, each point in the point cloud data of the target plane can be numbered according to the saddle-shaped weld bevel along the weld extension direction on the target plane. For example... Figure 2 As shown, through spatial dimensionality reduction processing, the target plane point cloud data is projected into a 3D point cloud onto... x - z Point cloud data on a plane, point cloud in the plane y The values ​​are the same. All points in the target planar point cloud data can be classified according to... x Numbering is done along the axis direction, where for the same axis... x The value point, inside can be arranged along z By numbering along the axis, the location of each point in the target planar point cloud data can be assigned.

[0029] The objective function described in this application embodiment can be determined based on the endpoint fitting deviation and slope consistency deviation of each segment. The endpoint fitting deviation characterizes the average distance deviation from the fitted curve of the line connecting each point within each segment to the endpoint; the slope consistency deviation characterizes the slope deviation between the endpoint fitting slope of each segment and the least squares curve. Specifically, the endpoint fitting deviation and slope consistency deviation can be calculated based on the coordinate information of each segmentation point and the coordinate information of each point within each segment.

[0030] The optimal segmentation point described in the embodiments of this application refers to the segmentation point determined by minimizing the objective function.

[0031] In the embodiments of this application, in step S1, three-dimensional point cloud data of the saddle-shaped weld groove is acquired by three-dimensional laser scanning. Assume the three-dimensional point cloud dataset of the saddle-shaped weld groove is... It can be represented as: (1) In the formula, This represents the point cloud data of the saddle-shaped weld groove contour obtained from the laser visual scan of the j-th frame. Wherein, N represents the total number of points when scanning one frame of the contour, and M represents the total number of frames of the 3D point cloud of the saddle-shaped weld groove obtained by laser vision scanning.

[0032] By performing spatial dimension reduction processing, the above three-dimensional point cloud can be projected onto... x - z On a plane, obtaining the point cloud data of the target plane allows us to acquire the point cloud dataset of the saddle-shaped weld groove. .

[0033] Furthermore, in step S2, each point in the target planar point cloud data is assigned a location number, which can be done according to the aforementioned method. x The axis direction numbering method numbers each point from 1 to... N This allows for the determination of multiple initial segmentation points from the numbered target planar point cloud data. This determines multiple segmentation segments. The next goal is to find the optimal sequence of segmentation points. Divide the data into K = From a continuous sub-segment, we can obtain 4 segments. This process can be represented as: (2) (3) here, This represents the first point in the target plane point cloud data.

[0034] Among them, each segment is required The covered point cloud data satisfies the linear assumption: (4) It should be noted that the saddle-shaped weld bevel is asymmetrical, and its bevel profile can be composed of four parts: the outer wall of the branch pipe, the sidewall of the bevel, the root pass surface, and the outer wall of the main pipe. The surfaces of the branch pipe and the main pipe are circular pipe surfaces with varying degrees of ellipticity. After manual root pass and grinding, the root pass surface is uneven and lacks symmetry. Therefore, the number of dividing points k=1,2,3 can be set; the number of dividing segments... K =1,2,3,4.

[0035] Furthermore, in the embodiments of this application, in step S3, an objective function is pre-constructed by combining the endpoint fitting deviation and slope consistency deviation of each segment. Then, with the goal of minimizing the objective function, the selection position of each segment point is iteratively optimized until each optimal segment point is determined. Specifically, based on the statistical characteristics of point cloud differential geometry, the point cloud data is recursively divided into multiple segments. The point cloud data within each segment has similar differential geometric statistical characteristics. The data is segmented, and it is assumed that each segment has a linear relationship. The slope and fitting error of each segment are calculated by linear fitting using the constructed objective function. The selection position of each segment point is iteratively optimized, and the linear fitting parameters of adjacent segments are compared to identify change points. When the deviation exceeds a set threshold, a change point is determined. Finally, the position of the change point is returned, and the optimal segment point is determined.

[0036] Furthermore, in the embodiments of this application, in step S4, by utilizing the three-dimensional spatial mapping relationship of each frame of laser scanning point cloud data, and based on each optimal segmentation point, the corresponding optimal segmentation points in each frame of laser scanning point cloud data can be determined. Then, through inter-frame combination, the corresponding optimal segmentation points in each frame of point cloud data are connected, thereby enabling accurate segmentation of the three-dimensional point cloud data of the saddle-shaped weld groove. This method can more stably and accurately process three-dimensional point cloud data of different saddle-shaped weld grooves under different scanning postures and different root pass grinding conditions, achieving effective point cloud segmentation and point cloud data extraction for four parts: the outer wall of the branch pipe, the side wall of the groove, the root pass surface of the weld, and the outer wall of the main pipe.

[0037] It should be noted that the form of each contour segment and the number of point clouds contained in the three-dimensional point cloud segmentation of the weld groove are related to the geometric contour features, spatial position and laser vision scanning angle of the saddle-shaped weld groove.

[0038] In actual production, the intersection weld of TKY type pipe joints differs significantly in geometry from traditional welds. Its bevel profile is not a regular straight line or circular curve, but rather a combination of straight lines, elliptical curves, and irregular curves. For example... Figure 3 As shown, Figure 3 (a) is a real-world schematic diagram of laser scanning point cloud data acquisition for a thick plate saddle-shaped weld groove provided in an embodiment of this application, and (b) is a schematic diagram of feature extraction from the point cloud data of the thick plate saddle-shaped weld groove provided in an embodiment of this application. The groove cross-sectional profile is composed of laser projections of the outer wall of the branch pipe, the side wall of the groove, the outer wall of the main pipe, and the root pass surface of the weld. When the laser irradiates the plane of the side wall of the groove, the projection is a straight profile; the profiles of the outer and inner walls of the branch pipe are elliptical profiles because the laser irradiates the circular pipe at a certain angle; and the profile of the root pass surface is irregular due to the irregularity of the root pass surface.

[0039] The point cloud segmentation method for thick plate saddle-shaped weld grooves in this application utilizes three-dimensional point cloud data of spatially intersecting multi-layered and multi-channel variable cross-section saddle-shaped weld grooves obtained by frame-by-frame scanning with online laser vision. The point cloud data is then converted to a target plane for dimensionality reduction analysis. Based on the spatial variation curve characteristics of the saddle-shaped weld groove, the point cloud data is divided into multiple segments, and linear fitting is performed on each segment. Simultaneously, a composite objective function is constructed using a combination of deviation estimation of the fitted straight line at the endpoints of the segment and slope consistency measurement to perform multi-objective optimization. This overcomes the problem of identifying irregular contours of the saddle-shaped weld groove cross-section, significantly improving the point cloud segmentation accuracy for saddle-shaped weld grooves and accurately extracting point cloud data from different parts of the saddle-shaped weld groove.

[0040] Based on the above embodiments, as an optional embodiment, step S1 involves dimensionality reduction processing of the acquired saddle-shaped weld groove three-dimensional point cloud data to obtain target plane point cloud data, including: The point cloud coordinates of the 3D point cloud data are decentralized to obtain the decentralized 3D point cloud data. Singular value decomposition is performed on the decentralized 3D point cloud data to determine the target rotation matrix; Based on the target rotation matrix, the 3D point cloud data is projected onto the target plane to obtain the target plane point cloud data.

[0041] Specifically, in the embodiments of this application, the obtained 3D point cloud data of the saddle-shaped weld groove is clean data after denoising and filtering the original point cloud data. When performing dimensionality reduction processing on the 3D point cloud data, singular value decomposition can be used to project the 3D point cloud data onto the target plane.

[0042] More specifically, the point cloud data of the saddle-shaped weld groove contour obtained from the laser visual scan of the j-th frame is acquired, assuming it is... Its corresponding centralized point cloud coordinates are ( Then we have: (5) If the 3D point cloud coordinates are decentralized, then: (6) Decentralized 3D point cloud data can be obtained. .

[0043] in .

[0044] Because the Singular Value Decomposition (SVD) method has advantages such as high stability and strong noise resistance in processing the distribution characteristics of point cloud data, it can be used to extract the plane normal vectors of the three-dimensional point cloud of the saddle-shaped weld groove, providing basic data for subsequent point cloud coordinate transformation.

[0045] Furthermore, the decentralized 3D point cloud Perform SVD decomposition, as shown in formula (7). The left singular vector corresponding to the smallest singular value is the normal vector of the plane, that is: (7) in, yes orthogonal matrix, yes diagonal matrix, yes The orthogonal matrix; then the plane normal vector According to the matrix Determined, that is .

[0046] Furthermore, calculate the axis of rotation. and rotation angle As shown in formulas (8) and (9).

[0047] (8) (9) in, .

[0048] The expression for the rotated 3D point cloud is shown in formulas (10)-(12). The target rotation matrix R is determined from this, and the target plane point cloud data is determined based on the target rotation matrix R. ,Right now: (10) (11) (12) The method in this application embodiment projects the three-dimensional point cloud onto the target plane using a rotation matrix after the original point cloud data has been denoised and filtered. This achieves the dimensionality reduction operation of the three-dimensional point cloud to two-dimensional space. It can reduce the feature analysis dimension while ensuring the preservation of the weld contour geometric information, and significantly reduce the processing difficulty and computational load.

[0049] Based on the above embodiments, as an optional embodiment, in step S3, before iteratively optimizing the selection position of each segmentation point with the objective function as the goal, until each optimal segmentation point is determined, the method further includes: Based on the coordinate information of each segmentation point and the coordinate information of each point within each segment, the endpoint fitting deviation and slope consistency deviation of each segment are determined. The objective function is determined based on the endpoint fitting bias and slope consistency bias of each segment, as well as the preset segment length penalty term parameter.

[0050] Specifically, in the embodiments of this application, in step S3, before iteratively optimizing the selection position of each segmentation point with the goal of minimizing the objective function, until each optimal segmentation point is determined, the objective function also needs to be constructed.

[0051] In the embodiments of this application, for complex weld grooves with spatially variable cross sections, a fitting method combining endpoint fitting deviation and slope consistency measurement is introduced. After initially determining each segmentation point from the target plane point cloud data after numbering, the endpoint fitting deviation and slope consistency deviation of each segment are determined by using the coordinate information of each segmentation point and the coordinate information of each point in each segment.

[0052] Based on the above embodiments, as an optional embodiment, the step of determining the endpoint fitting deviation of each segment specifically includes: Based on the coordinate information of each segmentation point, determine the linear fitting function for each segment; Determine the amplitude of each segment and the number of points within each segment; Based on the coordinate information of each point within each segment, the linear fitting function of each segment, the magnitude, and the number of points within each segment, the endpoint fitting deviation of each segment is determined.

[0053] Specifically, in the embodiments of this application, the specific implementation method for calculating the endpoint fitting deviation of each segment can be obtained by formula (13): For candidate segmentation points Define the endpoint fit bias of the Kth segment: (13) In the formula, Indicates the number of points within the segment; Indicates the magnitude of the segment; This represents the linear fitting function for the segmented data.

[0054] in, ; ; ; Therefore, the coordinate information of each segmentation point and the coordinate information of each point within each segment can be substituted into the above formulas to determine the function value, amplitude, and number of points within the linear fitting function of each segment. Then, the endpoint fitting deviation of each segment can be calculated using formula (13). .

[0055] The method in this application embodiment divides the target plane point cloud data into multiple segments according to the geometric features of the saddle-shaped weld groove. By using the coordinate information of each point in each segment, the linear fitting function of each segment, the amplitude, and the number of points in each segment, the endpoint fitting deviation of each segment is determined, ensuring the accuracy of the endpoint fitting deviation calculation.

[0056] Based on the above embodiments, as an optional embodiment, the step of determining the slope consistency deviation of each segment specifically includes: Based on the coordinate information of each segmentation point, the endpoint fitting slope of each segment is determined; Based on the coordinate information of each point within each segment, the least squares slope of each segment is determined; Based on the endpoint fitting slope and least squares slope of each segment, the slope consistency deviation of each segment is determined.

[0057] Specifically, in the embodiments of this application, the specific implementation method for calculating the slope consistency deviation of each segment can be obtained by formula (14): (14) In the formula: This represents the slope of the fitted endpoints of the segment. This represents the least squares slope of the segment.

[0058] in, ; ; Understandably, the endpoint fitting slope refers to the slope calculated by linear fitting using the two endpoints of the segment; the least squares slope refers to the slope calculated by least squares fitting a straight line using all point clouds within the segment.

[0059] Therefore, the coordinate information of each segmentation point and the coordinate information of each point within each segment can be substituted into the above formulas to determine the endpoint fitting slope and least squares slope of each segment. Then, the slope consistency deviation of each segment can be calculated using formula (14). .

[0060] The method in this application embodiment divides the target plane point cloud data into multiple segments according to the geometric features of the saddle-shaped weld groove. Using the coordinate information of each segment point and the coordinate information of each point within each segment, the endpoint fitting slope and least squares slope of each segment are determined. The slope consistency deviation of each segment is determined by combining the endpoint fitting slope and least squares slope, thus ensuring the accuracy of the slope consistency deviation calculation.

[0061] Furthermore, in the implementation of this application, a segment length penalty term can be introduced to construct an objective function based on the endpoint fitting deviation and slope consistency deviation of each segment, as well as the preset segment length penalty term parameters.

[0062] Specifically, the objective function after combining multiple objectives and introducing a piecewise length penalty term can be expressed as: (15) in, , represents the vector of the dividing points; , represents the weighting factor; , represents the length penalty coefficient; , representing the attenuation parameter Here, in this embodiment of the application, the quasi-convexity of the single-segment objective function is also proven, as described below.

[0063] fixed and Prove the function In the interval The upper part is convex.

[0064] The proof is as follows: make ,but: ; ; when As the derivative approaches the true dividing point, it changes sign and becomes zero at the true value.

[0065] The method in this application combines endpoint fitting deviation and slope consistency measurement to design a composite objective function with length penalty. It utilizes the quasi-convexity of the objective function to design an efficient optimization algorithm, which facilitates the rapid positioning and point cloud segmentation of the turning region of the saddle-shaped weld groove.

[0066] Based on the above embodiments, as an optional embodiment, step S3, with the objective of minimizing the objective function, iteratively optimizes the selection position of each segmentation point until the optimal segmentation points are determined, including: Step S101: Under the preset position constraints, the selected position of each segmentation point in the current iteration process is traversed and adjusted according to the preset displacement strategy, and the function value of the objective function corresponding to the selected position adjustment of all segmentation points is calculated each time. Step S102: Determine the optimal displacement corresponding to the minimum objective function value among all function values; Step S103: Based on the current selected position and optimal displacement of each segmentation point, determine each segmentation point in the next iteration process; Repeat steps S101 to S103 until the function value of the objective function meets the preset convergence condition, and determine each optimal split point.

[0067] Specifically, in the embodiments of this application, adaptive optimization is performed on each segmentation point. Specifically, for each segmentation point, several candidate positions (including the current point) are selected near the current point, the objective function value is calculated, and then the position that minimizes the objective function is selected as the new position of that segmentation point. This process is repeated until the objective function converges.

[0068] More specifically, let , which indicates the first n The split point vector of the next iteration Let the objective function be defined on the partition point space. Let represent a 1x3 positive integer vector. The algorithm solves the following optimization problem using the coordinate transformation method: (16) In the embodiments of this application, the positions of the selected segmentation points are first initialized. Specifically, the initial segmentation points can be set to a uniform distribution, i.e.: ; Based on the initial segmentation points, step S101 is executed, which involves traversing and adjusting the selected positions of each segmentation point in the current iteration process according to the preset displacement strategy under the preset position constraints, and calculating the function value of the objective function corresponding to the adjustment of the selected positions of all segmentation points each time.

[0069] Based on the above embodiments, as an optional embodiment, the preset displacement strategy is determined by equally dividing the step size function in the current iteration process; the step size function is determined based on the total number of points in the target plane point cloud data and the attenuation factor in the current iteration process, and the attenuation factor is determined based on the minimum objective function value in the previous iteration process and the initial function value of the objective function.

[0070] Specifically, in the embodiments of this application, an adaptive step size strategy is adopted, and an adaptive step size function is defined. for: (17) in, Indicates the attenuation factor. Adaptive step size parameters , N This represents the total number of points in the target plane point cloud data.

[0071] here, This represents the initial function value of the objective function, i.e. .

[0072] Furthermore, the preset displacement strategy is determined based on equally dividing the step size function in the current iteration process. Specifically, the preset displacement strategy is determined as follows: (18) in, The displacement represents the position of the current segmentation point. The adjustment range; It is a set of candidate displacements.

[0073] In the embodiments of this application, a coordinate rotation update strategy can be adopted. For each segmentation point t, the position of the segmentation point is tentatively adjusted using a preset displacement. Through discretized displacement probing, the optimal segmentation point position is found while ensuring topological constraints.

[0074] The method in this application avoids the step size sensitivity problem of continuous optimization methods (such as gradient descent) by adopting the above displacement strategy design, while maintaining low computational complexity, improving the algorithm optimization efficiency and accuracy, and realizing adaptive point cloud segmentation of saddle-shaped weld groove.

[0075] Furthermore, for each segmentation point, the preset positional constraints are determined as follows: (19) in, For minimum segment length constraints, .

[0076] Here, it should be noted that, It is an exploratory step size used to perform a local search traversal of the candidate displacement set near the current split point. For each displacement in, for each Candidate positions after displacement calculation are Based on the above-mentioned preset position constraints... This ensures that the selection of the segmentation point after displacement satisfies the constraint of the maximum number of change points, i.e. The value is not less than the position of the previous dividing point plus the minimum segment length, and not greater than the position of the next dividing point minus the minimum segment length.

[0077] By evaluating each The selection positions of each segmentation point are iterated and adjusted, and the function value of the objective function corresponding to each adjustment of the selection positions of all segmentation points is calculated. .

[0078] Furthermore, in step S102, the optimal displacement corresponding to the minimum objective function value among all function values ​​can be determined according to formula (20). : (20) in, , which represents the state of fixing other dividing points.

[0079] For example, the n Starting from the next iteration, optimize hour, That is, fixed , ,Sure Furthermore, optimization At that time, it needs to be fixed. , , determine Furthermore, optimize At that time, it needs to be fixed. , ,Sure .

[0080] Furthermore, in step S103, based on the current selected position and optimal displacement of each segmentation point, each segmentation point in the next iteration process can be determined using formula (21), thus completing the update of the segmentation points, i.e.: ;(twenty one) At the same time, according to the preset convergence conditions described in the following formulas (22) and (23), including the function value convergence condition and the function value gradient convergence condition, it is determined whether the function value of the current objective function has converged.

[0081] The convergence condition for the function value is: ;(twenty two) The convergence condition for the gradient of a function is: ;(twenty three) in: ; If the current objective function value does not meet the above-mentioned preset convergence conditions, then repeat steps S101 to S103 to perform the next iteration until the objective function value meets the above-mentioned function value convergence conditions and function value gradient convergence conditions. This determines the optimized position of each segmentation point in the current iteration, thus identifying each optimal segmentation point. .

[0082] Figure 4 (a) in this application is a variable cross-section saddle-shaped weld groove provided in an embodiment of this application. x - zA schematic diagram of the segmentation results of point cloud data on a plane. (b) is another variable cross-section saddle-shaped weld groove provided in the embodiment of this application. x - z A schematic diagram of the segmentation results of point cloud data on a plane. (Example) Figure 4 As can be seen, the outline shape and size of the variable cross-section saddle-shaped weld groove also change with the position. The method of this application embodiment can achieve accurate segmentation of point cloud data with different saddle-shaped weld groove outline shapes.

[0083] Furthermore, based on the optimal split point The set of segmented points of each frame of the laser scanning point cloud of the 3D point cloud of the variable cross-section saddle-shaped weld groove can be determined as follows: ;(twenty four) Finally, through the spatial mapping relationship between the laser scanning point clouds of each frame, the three-dimensional point cloud data of the variable cross-section saddle-shaped weld groove can be accurately divided into four segments by the optimal segmentation points, which can be represented as follows: (25) The method in this application embodiment utilizes a composite objective function with length penalty to minimize the objective function, performs 3D point cloud segmentation point change detection to search for the optimal segmentation point position, minimizes the fitting deviation, and combines the constraint of the maximum number of change points to achieve rapid optimization and positioning of saddle-shaped weld groove point cloud segmentation points, effectively improving the accuracy of saddle-shaped weld groove point cloud data segmentation.

[0084] The following describes the point cloud segmentation device for the bevel of a thick plate saddle-shaped weld provided in this application. The point cloud segmentation device for the bevel of a thick plate saddle-shaped weld described below can be referred to in correspondence with the point cloud segmentation method for the bevel of a thick plate saddle-shaped weld described above.

[0085] Figure 5 This is a schematic diagram of the point cloud segmentation device for the thick plate saddle-shaped weld groove provided in the embodiments of this application, as shown below. Figure 5 As shown, it includes: Dimensionality reduction module 10 is used to perform dimensionality reduction processing on the acquired 3D point cloud data of the saddle-shaped weld groove to obtain the target plane point cloud data. The first processing module 20 is used to assign a position number to each point in the target plane point cloud data, and to determine multiple segmentation points from the numbered target plane point cloud data to determine multiple segmentation segments; The second processing module 30 is used to iteratively optimize the selection position of each segmentation point with the goal of minimizing the objective function until the optimal segmentation point is determined; the objective function is determined based on the endpoint fitting deviation and slope consistency deviation of each segment. The segmentation module 40 is used to segment the 3D point cloud data based on each optimal segmentation point.

[0086] It is understood that the detailed functional implementation of each of the above units / modules can be found in the description in the aforementioned method embodiments, and will not be repeated here.

[0087] It should be understood that the above-described device is used to execute the methods in the above embodiments. The implementation principle and technical effect of the corresponding program modules in the device are similar to those described in the above methods. The working process of the device can be referred to the corresponding process in the above methods, and will not be repeated here.

[0088] The point cloud segmentation device for thick plate saddle-shaped weld grooves in this embodiment utilizes three-dimensional point cloud data of spatially intersecting multi-layered and multi-channel variable cross-section saddle-shaped weld grooves obtained by frame-by-frame scanning with online laser vision. The point cloud data is then converted to a target plane for dimensionality reduction analysis. Based on the spatial variation curve characteristics of the saddle-shaped weld groove, the point cloud data is divided into multiple segments, and linear fitting is performed on each segment. Simultaneously, a composite objective function is constructed using a combination of deviation estimation of the fitted straight line at the endpoints of the segment and slope consistency measurement to perform multi-objective optimization. This overcomes the problem of identifying irregular contours of the saddle-shaped weld groove cross-section, significantly improving the point cloud segmentation accuracy for saddle-shaped weld grooves and accurately extracting point cloud data from different parts of the saddle-shaped weld groove.

[0089] Based on the methods in the above embodiments, this application provides an electronic device, such as... Figure 6 As shown, the electronic device may include a processor 610, a communications interface 620, a memory 630, and a communication bus 640, wherein the processor 610, the communications interface 620, and the memory 630 communicate with each other via the communication bus 640. The processor 610 can call logical instructions in the memory 630 to execute the methods in the above embodiments.

[0090] Furthermore, the logical instructions in the aforementioned memory 630 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application.

[0091] Based on the methods in the above embodiments, this application provides a computer-readable storage medium storing a computer program that, when run on a processor, causes the processor to execute the methods in the above embodiments.

[0092] Based on the methods in the above embodiments, this application provides a computer program product that, when run on a processor, causes the processor to execute the methods in the above embodiments.

[0093] It is understood that the processor in the embodiments of this application can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. A general-purpose processor can be a microprocessor or any conventional processor.

[0094] The method steps in this application embodiment can be implemented in hardware or by a processor executing software instructions. The software instructions can consist of corresponding software modules, which can be stored in random access memory (RAM), flash memory, read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), registers, hard disks, portable hard disks, CD-ROMs, or any other form of storage medium known in the art. An exemplary storage medium is coupled to the processor, enabling the processor to read information from and write information to the storage medium. Of course, the storage medium can also be a component of the processor. The processor and the storage medium can reside in an ASIC.

[0095] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially as a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted through the computer-readable storage medium. The computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state disk (SSD)).

[0096] It is understood that the various numerical designations used in the embodiments of this application are merely for the convenience of description and are not intended to limit the scope of the embodiments of this application.

[0097] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for segmenting point clouds at the bevel of a saddle-shaped weld in a thick plate, characterized in that, include: The 3D point cloud data of the acquired saddle-shaped weld groove is reduced in dimension to obtain the target plane point cloud data. Each point in the target plane point cloud data is assigned a position number, and multiple segmentation points are determined from the numbered target plane point cloud data to determine multiple segmentation segments; With the objective function as the goal, the selection position of each segmentation point is iteratively optimized until the optimal segmentation point is determined; the objective function is determined based on the endpoint fitting deviation and slope consistency deviation of each segment. The three-dimensional point cloud data is segmented according to each of the optimal segmentation points; The method further includes, prior to iteratively optimizing the selection position of each segmentation point with the objective function as the minimization target until the optimal segmentation points are determined: Based on the coordinate information of each segmentation point and the coordinate information of each point within each segment, the endpoint fitting deviation and slope consistency deviation of each segment are determined. The objective function is determined based on the endpoint fitting deviation and slope consistency deviation of each segment, as well as the preset segment length penalty term parameter. The steps for determining the endpoint fitting deviation of each segment specifically include: Based on the coordinate information of each segmentation point, a linear fitting function for each segment is determined; Determine the amplitude and the number of points within each segment; Based on the coordinate information of each point within each segment, the linear fitting function of each segment, the magnitude, and the number of points within each segment, the endpoint fitting deviation of each segment is determined. The steps for determining the slope consistency deviation of each segment specifically include: Based on the coordinate information of each segmentation point, the endpoint fitting slope of each segment is determined; Based on the coordinate information of each point within each segment, the least squares slope of each segment is determined; Based on the endpoint fitting slope and least squares slope of each segment, the slope consistency deviation of each segment is determined.

2. The method for segmenting the point cloud of a saddle-shaped weld groove in a thick plate according to claim 1, characterized in that, The step of iteratively optimizing the selection position of each segmentation point with the objective of minimizing the objective function until the optimal segmentation points are determined includes: Step S101: Under the preset position constraint, the selected position of each of the segmentation points in the current iteration process is traversed and adjusted according to the preset displacement strategy, and the function value of the objective function corresponding to the selected position adjustment of all the segmentation points is calculated each time. Step S102: Determine the optimal displacement corresponding to the minimum objective function value among all the function values; Step S103: Determine each of the segmentation points in the next iteration process based on the current selected position of each segmentation point and the optimal displacement. Repeat steps S101 to S103 until the function value of the objective function satisfies the preset convergence condition, and determine each optimal split point.

3. The method for segmenting the point cloud of the saddle-shaped weld groove in thick plates according to claim 2, characterized in that, The preset displacement strategy is determined by equally dividing the step size function in the current iteration process; the step size function is determined based on the total number of points in the target plane point cloud data and the attenuation factor in the current iteration process, and the attenuation factor is determined based on the minimum objective function value in the previous iteration process and the initial function value of the objective function.

4. The method for segmenting the point cloud of a saddle-shaped weld groove in a thick plate according to any one of claims 1-3, characterized in that, The dimensionality reduction processing of the acquired saddle-shaped weld groove three-dimensional point cloud data to obtain target plane point cloud data includes: The point cloud data is processed by decentralizing the point cloud coordinates to obtain decentralized three-dimensional point cloud data. Singular value decomposition is performed on the decentralized 3D point cloud data to determine the target rotation matrix; Based on the target rotation matrix, the three-dimensional point cloud data is projected onto the target plane to obtain the target plane point cloud data.

5. A point cloud segmentation device for the bevel of a thick plate saddle-shaped weld, characterized in that, include: The dimension reduction module is used to perform dimension reduction processing on the acquired 3D point cloud data of the saddle-shaped weld groove to obtain the target plane point cloud data. The first processing module is used to assign a position number to each point in the target plane point cloud data, and to determine multiple segmentation points from the numbered target plane point cloud data to determine multiple segmentation segments; The second processing module is used to iteratively optimize the selection position of each segmentation point with the goal of minimizing the objective function, until the optimal segmentation point is determined; the objective function is determined based on the endpoint fitting deviation and slope consistency deviation of each segmentation segment. The segmentation module is used to segment the three-dimensional point cloud data according to each of the optimal segmentation points; Before iteratively optimizing the selection position of each segmentation point with the objective of minimizing the objective function until the optimal segmentation point is determined, the process further includes: Based on the coordinate information of each segmentation point and the coordinate information of each point within each segment, the endpoint fitting deviation and slope consistency deviation of each segment are determined. The objective function is determined based on the endpoint fitting deviation and slope consistency deviation of each segment, as well as the preset segment length penalty term parameter. The steps for determining the endpoint fitting deviation of each segment specifically include: Based on the coordinate information of each segmentation point, a linear fitting function for each segment is determined; Determine the amplitude and the number of points within each segment; Based on the coordinate information of each point within each segment, the linear fitting function of each segment, the magnitude, and the number of points within each segment, the endpoint fitting deviation of each segment is determined. The steps for determining the slope consistency deviation of each segment specifically include: Based on the coordinate information of each segmentation point, the endpoint fitting slope of each segment is determined; Based on the coordinate information of each point within each segment, the least squares slope of each segment is determined; Based on the endpoint fitting slope and least squares slope of each segment, the slope consistency deviation of each segment is determined.

6. An electronic device, characterized in that, include: At least one memory for storing computer programs; At least one processor is configured to execute a program stored in the memory, wherein when the program stored in the memory is executed, the processor is configured to perform the method as described in any one of claims 1-4.

7. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is run on the processor, it causes the processor to perform the method as described in any one of claims 1-4.

Images