Forming size detection method for additive manufacturing cladding channel feature point recognition based on all-morphology point cloud

Through the 3D visual full-morphological point cloud information method, the problems of depth of field error and breakpoint error in arc additive manufacturing of two-dimensional line laser sensors are solved, and high-precision cladding path size detection is achieved, suitable for the comparison of layers. In the case of high conditions, it provides data support for precise control of forming dimensionality and quality discrimination of arc additive manufacturing.

CN119006565AActive Publication Date: 2024-11-22UNIV OF ELECTRONICS SCI & TECH OF CHINA +1

Patent Information

Application Number
CN202411013575.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-26
Publication Date
2024-11-22
Estimated Expiration
2044-07-26

AI Technical Summary

Technical Problem

Existing two-dimensional line laser sensors have depth of field error and breakpoint error in arc additive manufacturing, which cannot effectively identify the real point cloud information in the incomplete fusion area between deposited layers, resulting in low detection accuracy, especially not suitable for high layers. The situation.

Method used

Using a method based on 3D visual full-morphological point cloud information, the three-dimensional point cloud data of each deposited layer is collected through a high-precision 3D vision scanner, and the noise points are filtered using statistical filtering algorithms to perform surface reconstruction and point cloud data uniform and fine , identify the characteristic points and bite points of welding toe, combine the Savitzky-Golay filter and gradient curvature threshold method, dimensionality reduction processing and heterogeneous region adaptive interpolation to realize high-precision dimensional detection of cladding channels.

Benefits of technology

Achieve high-precision cladding path width and height detection, avoid depth of field error and breakpoint error, and can effectively identify the bite points in the incomplete fusion area between layers. It is suitable for situations with high layers and provides arc additives. Manufactured shaping dimensions and data support for quality discrimination.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119006565A_ABST
    Figure CN119006565A_ABST
Patent Text Reader

Abstract

The invention provides an electric arc additive manufacturing cladding channel feature point recognition and forming size detection method based on 3D vision full-shape point cloud information, which comprises the following steps of: performing all-dimensional scanning on each layer of cladding channel by using a high-precision 3D scanner; performing preprocessing such as filtering, three-dimensional curved surface reconstruction, point cloud refinement and feature fitting alignment on the point cloud data to obtain effective point cloud data; dimension reduction processing is conducted on the effective point cloud, and a DW-GCT method is used for searching weld toe feature points of a first-layer cladding channel in a bidirectional approximation mode; and searching edge back-convex and non-edge back-convex regions by adopting a heterogeneous region adaptive interpolation method to obtain a cladding channel interpolation contour with small surface features reserved, and bidirectionally searching undercut points of incomplete fusion regions between adjacent layers by combining a layer intersection point dynamic positioning method. The method provided by the invention can effectively solve the problems of non-cladding channel side point cloud information, depth-of-field error and breakpoint error in the prior art, and provides a method for extracting top-layer cladding channel size data by using cladding channel full-shape three-dimensional point cloud contour information.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of metal additive manufacturing technology, and in particular to a background technology of a method for identifying feature points of an additive manufacturing cladding path and detecting forming dimensions based on 3D visual full-morphology point cloud information. Background Art

[0002] Wire Arc Additive Manufacturing (WAAM) is a technology based on the discrete-deposition principle, which uses an arc heat source to melt the end of a metal wire at high temperature, continuously feeds molten metal droplets into the molten pool through free transition or liquid bridge transition, and continuously deposits them layer by layer according to a pre-planned path for cladding and forming. Compared with traditional subtractive processing, wire arc additive manufacturing technology can quickly manufacture large-sized, special-shaped and complex structural parts. It has the advantages of rapid near-net forming, extremely high material utilization, high design flexibility, and high production efficiency. It has been widely used in aerospace, automotive, shipbuilding and other industrial fields.

[0003] However, due to the large heat input of the arc, the flow behavior and shape of the molten pool become more complex and difficult to control due to the influence of the heat input. The heat accumulation effect changes the thermal boundary conditions of the molten pool, affecting the morphology and dimensional consistency of the deposited layer, making it difficult to keep the height and width of the cladding path consistent, resulting in low robustness of the formed dimensions and poor process stability. Automatic forming control technology is the core way to improve the stability of the arc additive process and the accuracy of the cladding path forming dimensions by continuously correcting the deviation between the actual measured dimensions and the expected dimensions. The detection accuracy of the cladding path forming dimensions in the additive process is an important prerequisite for accurate control of the forming dimensions and judgment of the forming quality.

[0004] At present, there are methods for measuring the size of cladding beads using visual sensors in the field of additive manufacturing. For example, Deng Junhao et al. [Deng Junhao. Research on arc additive forming morphology detection and predictive modeling based on visual sensing [D]. Shanghai Jiaotong University, 2019.] used a two-dimensional line laser sensor to vertically scan the cladding bead to construct three-dimensional point cloud information, and identified the weld toe through the first-order and second-order height difference methods to calculate the forming width size. This method can identify the weld toe feature points from the three-dimensional point cloud information containing the substrate and the deposited layer, but the feature point search method is only based on the height difference between adjacent points, and has poor robustness. In addition, the detection accuracy is restricted by the accuracy and scanning range of the line laser sensor. Therefore, the final detection result has a large error compared with the actual result.

[0005] Among the existing methods, most studies use two-dimensional line laser sensors to vertically illuminate the cladding track to obtain point cloud information. Since the two-dimensional line laser sensor cannot scan the side profile of the cladding wall and cannot identify the real point cloud information of the incomplete fusion area between the deposited layers, and the vertically irradiated line laser causes breakpoints on both sides of the cladding track, most studies directly use the breakpoints on both sides as feature points for calculating the width. However, due to the depth of field problem of the two-dimensional line laser, the position of the breakpoint is related to the laser height and the angle α, and when there are edge convex areas on both sides of the cladding track due to the liquid metal not spreading evenly to both sides before solidification, the breakpoint may not be a feature point reflecting the real width (such as Figure 1 As shown in Figure 2, the 2D line laser method is used to measure the width of each cladding track, which inevitably results in a certain depth of field error [1] and breakpoint error. The detection accuracy is limited by the accuracy and scanning range of the line laser sensor, which ultimately leads to a large error between the detected width and the actual width, causing the width detection data to be distorted. Therefore, it is not suitable for situations with a large number of layers. In addition, there is currently no method to directly use the 3D point cloud contour information of the full morphology of the cladding track to extract the size data of the top cladding track. Summary of the invention

[0006] In view of this, the purpose of the present invention is to overcome the problems of depth of field error and breakpoint error that always exist in existing two-dimensional line laser sensors, and propose a method for directly using the three-dimensional point cloud contour information of the full morphology of the cladding track to extract the size data of the top cladding track, and provide an arc additive manufacturing cladding track feature point recognition and forming size detection method based on 3D visual full morphology point cloud information, which is used to solve the problems that the existing method is not suitable for a large number of layers, cannot identify the real point cloud information of the bite points on both sides of the incomplete fusion zone between the deposition layers, and has depth of field error and breakpoint error.

[0007] In order to achieve the above objectives, the following technical solutions are proposed:

[0008] A forming size detection method for identifying characteristic points of an additive manufacturing cladding path based on a full-morphology point cloud, the method comprising the following steps:

[0009] S1, collects the original data of real 3D point cloud of each sedimentary layer;

[0010] S2, filtering noise points using a statistical filtering algorithm on the original three-dimensional point cloud data of each sedimentary layer described in S1 to obtain filtered point cloud data;

[0011] S3, performing surface three-dimensional reconstruction, uniform refinement of the point cloud data, and feature fitting and alignment of adjacent layer point clouds on the filtered point cloud data in S2 to obtain valid point cloud data;

[0012] S4, for the valid point cloud data described in S3, identifying the weld toe feature points and height feature points in the first layer of point cloud data, and calculating the width and height dimensions;

[0013] S5, for the valid point cloud data described in S3, identifying the undercut points and height feature points representing the incomplete fusion area between layers in the nth layer of point cloud data, where n>1, and calculating the undercut width, maximum width and height dimensions;

[0014] In a preferred embodiment, dimension reduction processing is performed on the effective point cloud data of S4 and S5, wherein the first layer weld toe feature points and the nth layer interlayer undercut points in S4 and S5 are extracted from the cross section perpendicular to the additive direction, and each cross section is separated by 1 mm along the additive direction.

[0015] S6, obtaining the cladding channel forming dimension data.

[0016] In a preferred embodiment, the step S4 specifically includes:

[0017] S4-1, determining two weld toe feature points in the cross section, the cladding track width of the current cross section being the distance between the two weld toe feature points;

[0018] S4-2, search for the point with the maximum value in the height direction in the section as the vertex of the cladding track in the current layer of this section, the vertex line of each section is the center line of the cladding track of the current layer, and the distance from the vertex to the line connecting the two weld toe feature points is taken as the height of the first layer of cladding track;

[0019] S4-3, record the cladding track height, maximum cladding track width and undercut width, and vertex coordinate data of each section in the first layer.

[0020] In a preferred embodiment, in step S4-1, since the point cloud data includes the substrate point cloud information, a dynamic sliding window based on a Savitzky-Golay filter and a gradient curvature threshold method are combined to perform a bidirectional approximation search to separate feature points representing weld toes from the substrate point cloud, as follows:

[0021]

[0022] Where N represents the set of all original data points in the current section, i represents the index of each point; x[c n +j] represents the center point c in the nth window. n The data point is shifted forward by j units; c j represents the coefficient of polynomial fitting for each point in the window, which is used to smooth the data points; m represents the width of the window, and its value changes dynamically and is related to the total number of point clouds in the current section; f[c n ] represents the center index c of each window after filtering nThe value of the data point can reduce the interference of noise and more accurately capture the trend of data changes; G(c n ,j) and C(c n ,j) represent the center index c n The gradient and curvature values ​​in this window; G tolerence and C tolerance They represent the tolerances of gradient and curvature respectively. When points that meet the tolerances of gradient and curvature are searched from both sides of the cladding track to the middle, the two points are recorded as weld toe feature points.

[0023] In a preferred embodiment, the step S5 specifically includes:

[0024] S5-1, extracting the point with the maximum value in the height direction in the cross section as the vertex of the cladding track in the current layer of the cross section, wherein the height of the layer is the difference between the height of the vertex in the current cross section and the height of the vertex in the previous layer, and the line connecting the vertices of each cross section is the center line of the cladding track of the current layer;

[0025] S5-2, identifying edge convex regions and non-edge convex regions, and performing heterogeneous region adaptive Hermite interpolation on the n-1th layer of point cloud data;

[0026] S5-3, identifying two points where the interpolation contours of the nth layer and the n-1th layer intersect by means of a layer intersection dynamic positioning method, extracting all point clouds between the two intersection points and storing them in an array, the width of the cladding track of the nth layer of the current section is the maximum distance between any two points in the array, and the undercut width on both sides of the cladding track of the nth layer of the current section is the distance between the two intersection points;

[0027] S5-4, recording the cladding track height, cladding track width, undercut width, and vertex coordinate data of each cross section in the n-1th layer;

[0028] S5-5, the intersection recognition and size detection of the current layer are completed, and the next layer of point cloud data processing is prepared.

[0029] In a preferred embodiment, in step S5-2, since the surface of the cladding track is not smooth due to oxidized welding slag, and the fitting method will ignore the small defect features in the contour surface of the point cloud data, an interpolation method is used to completely retain the defect features of the cladding track surface.

[0030] In a preferred embodiment, in step S5-2, since there may be small droplets on the edge of the cladding track contour that solidify without spreading evenly, that is, the edge convex area, this special area has multiple x values ​​in certain specific z values, so conventional interpolation cannot be performed. Therefore, the heterogeneous regional adaptive interpolation method is used to search for edge convex and non-edge convex areas, and parameter reverse Hermite interpolation and forward Hermite interpolation are performed respectively, and the formula is as follows:

[0031]

[0032] Wherein, N1 and N2 represent the point sets belonging to the edge convex and non-edge convex regions respectively; a and v a represents the x and z values ​​of the ath point in the N1 point set, u b and v b represents the x value and z value of the bth point in the N2 point set; g a and g b Respectively expressed in v a and u b The gradient value at ; h(t) and lk(z) represent the shape function and Lagrange basis function of the Hermite interpolation formula respectively.

[0033] In a preferred embodiment, in step S5-3, the layer intersection dynamic positioning method utilizes the heterogeneous region adaptive interpolation method and combines the binary search idea to perform a bidirectional search and compare the interpolation and coordinate values ​​of each current point and the next point. When a qualified point is found from the middle of the cladding track to both sides, the two points are recorded as the undercut feature points.

[0034] Compared with the shortcomings and deficiencies of the existing technology, the beneficial effects of the present invention are:

[0035] 1. The present invention adopts a high-precision 3D visual scanner to collect the full-morphology three-dimensional point cloud data of each deposition layer, and proposes a method for extracting the top layer size data from the full-morphology three-dimensional point cloud contour information, thereby avoiding the problem of distortion in the calculation of the cladding track width caused by the inability of the two-dimensional line laser sensor to collect the side contour of the cladding track and the insufficient collection accuracy.

[0036] 2. The present invention proposes a method for quickly identifying the characteristic points of the weld toe of the cladding track based on 3D visual full-morphology point cloud information. By performing dimensionality reduction processing on the high-precision three-dimensional point cloud data, the segmentation and extraction of the first layer of cladding track point cloud data from the substrate is effectively realized.

[0037] 3. The present invention proposes a method for quickly identifying the undercut points in the incomplete fusion zone between layers of the cladding track based on 3D visual full morphology point cloud information. By using the heterogeneous region adaptive interpolation method for the edge convex and non-edge convex area point clouds, the complete information of the surface contour of the cladding track can be effectively retained, and the undercut points in the incomplete fusion zone between layers can be effectively identified through the layer intersection dynamic positioning method. Finally, the continuously changing data of the morphology of each layer of the cladding track are obtained without manual detection. The problems of the existing method that the morphology data of the incomplete fusion zone between layers cannot be identified, the method is not suitable for the case of a large number of layers, and the existence of depth of field error and breakpoint error are solved, which provides data support for the precise control of the forming size and the judgment of the forming quality of arc additive manufacturing. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 is a schematic diagram of depth of field error and breakpoint error in a preferred embodiment of the present invention;

[0039] Figure 2 It is a flow chart of a method for identifying characteristic points of cladding tracks and detecting forming dimensions in arc additive manufacturing based on 3D visual full-morphology point cloud information in a preferred embodiment of the present invention;

[0040] Figure 3 is a diagram of a point cloud preprocessing example of a preferred embodiment of the present invention;

[0041] Figure 4 is an example diagram of a heterogeneous region adaptive interpolation method according to a preferred embodiment of the present invention;

[0042] Figure 5 This is an example diagram of the detection result of the width dimension of the first layer of cladding track of the preferred embodiment of the present invention;

[0043] Figure 6 This is an example diagram of the second layer cladding track width dimension detection result of the preferred embodiment of the present invention;

[0044] Figure 7 This is an example diagram of the width dimension detection result of the third layer cladding track of the preferred embodiment of the present invention;

[0045] Figure 8 This is an example diagram of the detection result of the width dimension of the 4th layer cladding track of the preferred embodiment of the present invention;

[0046] Fig. 9 This is an example diagram of the width dimension detection result of the 5th layer cladding track of the preferred embodiment of the present invention;

[0047] Fig.10 It is an example diagram of the detection result of the cladding track forming size according to the preferred embodiment of the present invention. DETAILED DESCRIPTION

[0048] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0049] like Figure 2 As shown, this embodiment proposes a method for arc additive manufacturing cladding track feature point recognition and forming size detection based on 3D visual full-morphology point cloud information, including the following steps:

[0050] S1, install a high-precision 3D visual mesh structured light scanner on the robotic arm, and perform a full-scale scan of each surface of the cladding track, collect the real 3D point cloud raw data of each deposition layer, and store it in PLY format;

[0051] S2, filtering noise points using a statistical filtering algorithm on the original three-dimensional point cloud data of each sedimentary layer described in S1 to obtain filtered point cloud data;

[0052] S3, performing surface 3D reconstruction, uniform refinement of the point cloud data, and feature fitting and alignment of adjacent layer point clouds on the filtered point cloud data described in S2 to obtain valid point cloud data, such as Figure 3 As shown;

[0053] S4, performing dimensionality reduction processing on the effective point cloud data described in S3, extracting point cloud data in each section of the effective point cloud along the section perpendicular to the additive direction, and each section is separated by 1 mm along the additive direction; identifying the weld toe feature points and height feature points in the first layer of point cloud data, and calculating the width and height dimensions, such as Figure 5 and Fig.10 As shown;

[0054] S4-1, based on the DW-GCT (dynamic sliding window-gradient curvature threshold) method, two weld toe feature points in the cross section are searched bidirectionally, and the cladding track width of the current cross section is the distance between the two weld toe feature points;

[0055] S4-2, searching for the point with the maximum value in the height direction in the section as the vertex of the cladding track in the section of the current layer, the line connecting the vertices of each section is the center line of the cladding track of the current layer, and the distance from the vertex to the line connecting the two weld toe feature points is used as the height of the first layer of cladding track;

[0056] S4-3, recording the cladding track height, maximum cladding track width, undercut width, and vertex coordinate data of each cross section in the first layer;

[0057] S4-4, the first layer of weld toe feature point recognition and size detection is completed, and the next layer of point cloud data processing is prepared.

[0058] According to the above-mentioned arc additive manufacturing cladding track feature point recognition and forming size detection method based on 3D visual full-morphology point cloud information, in the step S4-1, since the point cloud data contains the substrate point cloud information, the dynamic sliding window based on the Savitzky-Golay filter and the bidirectional approximation search combined with the GCT method are used to separate the feature points representing the weld toe from the substrate point cloud. The DW-GCT method formula is as follows:

[0059]

[0060] Where N represents the set of all original data points in the current section, i represents the index of each point; x[c n +j] represents the center point c in the nth window. n The data point is shifted forward by j units; c j represents the coefficient of polynomial fitting for each point in the window, which is used to smooth the data points; m represents the width of the window, and its value changes dynamically and is related to the total number of point clouds in the current section; f[c n ] represents the center index c of each window after filtering n The value of the data point can reduce the interference of noise and more accurately capture the trend of data changes; G(c n ,j) and C(c n ,j) represent the center index c n The gradient and curvature values ​​in this window; G tolerence and C tolerance They represent the tolerances of gradient and curvature respectively. When points that meet the tolerances of gradient and curvature are searched from both sides of the cladding track to the middle, the two points are recorded as weld toe feature points.

[0061] S5, performing feature point recognition on the cross-sectional point cloud after the dimensionality reduction processing described in S4, identifying the bite edge points and height feature points representing the incomplete fusion area between layers in the n-th layer (n>1) of the point cloud data, and calculating the bite edge width, maximum width and height dimensions;

[0062] S5-1, extracting the point with the maximum value in the height direction in the cross section as the vertex of the cladding track in the current layer of the cross section, wherein the height of the layer is the difference between the height of the vertex in the current cross section and the height of the vertex in the previous layer, and the line connecting the vertices of each cross section is the center line of the cladding track of the current layer;

[0063] S5-2, identifying edge convex regions and non-edge convex regions, and performing heterogeneous region adaptive Hermite interpolation on the n-1th layer (n>1) of point cloud data;

[0064] S5-3, identify the two points where the interpolation functions of the nth layer and the n-1th layer intersect through the layer intersection dynamic positioning method, extract all point clouds between the two intersections and store them in an array, the width of the cladding track of the nth layer in the current section is the maximum distance between any two points in the array, and the width of the undercuts on both sides of the cladding track of the nth layer in the current section is the distance between the two intersections.

[0065] S5-4, record the cladding track height, cladding track width, undercut width, and vertex coordinate data of each section in the nth (n>1) layer, such as Figure 6 , Figure 7 , Figure 8 , Fig. 9 , Fig.10 As shown;

[0066] S5-5, the intersection recognition and size detection of the current layer are completed, and the next layer of point cloud data processing is prepared.

[0067] According to the above-mentioned method for identifying characteristic points of arc additive manufacturing cladding tracks and detecting formed dimensions based on 3D visual full-morphology point cloud information, in step S5-2, since the surface of the cladding track is not smooth due to oxidized welding slag, and the fitting method will ignore the small defective features in the contour surface of the point cloud data, the interpolation method is used to completely retain the surface defect features of the cladding track.

[0068] According to the above-mentioned arc additive manufacturing cladding track feature point recognition and forming size detection method based on 3D visual full-morphology point cloud information, in the step S5-2, since there may be small molten droplets at the edge of the cladding track contour that solidify without spreading evenly, that is, the edge convex area, this special area has multiple x values ​​in certain specific z values, so conventional interpolation cannot be performed, such as Figure 4 Therefore, the heterogeneous region adaptive interpolation method is used to search for edge convex and non-edge convex regions, and the parameter reverse Hermite interpolation and forward Hermite interpolation are performed respectively. The formula is as follows:

[0069]

[0070] Wherein, N1 and N2 represent the point sets belonging to the edge convex and non-edge convex regions respectively; a and v a represents the x and z values ​​of the ath point in the N1 point set, u b and v b represents the x value and z value of the bth point in the N2 point set; g a and g b Respectively expressed in v a and u b The gradient value at ; h(t) and lk(z) represent the shape function and Lagrange basis function of the Hermite interpolation formula respectively.

[0071] According to the above-mentioned method for identifying characteristic points of arc additive manufacturing cladding tracks and detecting formed dimensions based on 3D visual full-morphology point cloud information, in step S5-3, the layer intersection dynamic positioning method utilizes the heterogeneous region adaptive interpolation method and combines the idea of ​​dichotomy to bidirectionally search and compare the interpolation and coordinate values ​​of each current point and the next point. When a qualified point is bidirectionally searched from the middle of the cladding track to both sides, the two points are recorded as the biting edge feature points.

[0072] S6, obtaining the forming dimension data of the first layer of cladding track and the top layer of the cladding track.

[0073] The above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit the scope of protection of the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, a person of ordinary skill in the art should understand that any modification, equivalent substitution and improvement made to the technical solutions of the present invention without departing from the purpose and scope of the technical solutions of the present invention should be included in the scope of the claims of the present invention.

Claims

1. A forming size detection method for identifying characteristic points of an additive manufacturing cladding path based on a full-morphology point cloud, the method comprising the following steps: S1, collects the original data of real 3D point cloud of each sedimentary layer; S2, filtering noise points using a statistical filtering algorithm on the original three-dimensional point cloud data of each sedimentary layer described in S1 to obtain filtered point cloud data; S3, performing surface three-dimensional reconstruction, uniform refinement of the point cloud data, and feature fitting and alignment of adjacent layer point clouds on the filtered point cloud data in S2 to obtain valid point cloud data; S4, for the valid point cloud data described in S3, identifying the weld toe feature points and height feature points in the first layer of point cloud data, and calculating the width and height dimensions; S5, for the valid point cloud data described in S3, identifying the undercut points and height feature points representing the incomplete fusion area between layers in the nth layer of point cloud data, where n>1, and calculating the undercut width, maximum width and height dimensions; S6, obtaining the cladding channel forming dimension data.

2. A forming size detection method for identifying characteristic points of cladding tracks in additive manufacturing based on full morphology point cloud according to claim 1, characterized in that: The effective point cloud data of S4 and S5 are further processed for dimensionality reduction, wherein the first layer weld toe feature points and the nth layer interlayer undercut points in S4 and S5 are extracted from the cross section perpendicular to the additive direction, and each cross section is separated by 1 mm along the additive direction.

3. The forming size detection method of the additive manufacturing cladding track feature point recognition based on the full morphology point cloud according to claim 1 is characterized in that: The step S4 specifically includes: S4-1, determining two weld toe feature points in the cross section, the cladding track width of the current cross section being the distance between the two weld toe feature points; S4-2, search for the point with the maximum value in the height direction in the section as the vertex of the cladding track in the current layer of this section, the vertex line of each section is the center line of the cladding track of the current layer, and the distance from the vertex to the line connecting the two weld toe feature points is taken as the height of the first layer of cladding track; S4-3, record the cladding track height, maximum cladding track width and undercut width, and vertex coordinate data of each section in the first layer.

4. The forming size detection method of the additive manufacturing cladding track feature point recognition based on the full morphology point cloud according to claim 1 is characterized in that: The step S5 specifically includes: S5-1, extracting the point with the maximum value in the height direction in the cross section as the vertex of the cladding track in the current layer of the cross section, wherein the height of the layer is the difference between the height of the vertex in the current cross section and the height of the vertex in the previous layer, and the line connecting the vertices of each cross section is the center line of the cladding track of the current layer; S5-2, identifying edge convex regions and non-edge convex regions, and performing heterogeneous region adaptive Hermite interpolation on the n-1th layer of point cloud data; S5-3, identifying two points where the interpolation contours of the nth layer and the n-1th layer intersect by means of a layer intersection dynamic positioning method, extracting all point clouds between the two intersection points and storing them in an array, the width of the cladding track of the nth layer of the current section is the maximum distance between any two points in the array, and the undercut width on both sides of the cladding track of the nth layer of the current section is the distance between the two intersection points; S5-4, recording the cladding track height, cladding track width, undercut width, and vertex coordinate data of each cross section in the n-1th layer; S5-5, the intersection recognition and size detection of the current layer are completed, and the next layer of point cloud data processing is prepared.

5. The forming size detection method of the additive manufacturing cladding track feature point recognition based on the full morphology point cloud as claimed in claim 3, characterized in that: In step S4-1, since the point cloud data contains the substrate point cloud information, the dynamic sliding window of the Savitzky-Golay filter is combined with the gradient curvature threshold method to perform a bidirectional approximation search to separate the feature points representing the weld toe from the substrate point cloud. The method is as follows: Where N represents the set of all original data points in the current section, i represents the index of each point; x[c n +j] represents the center point c in the nth window. n The data point is shifted forward by j units; c j represents the coefficient of polynomial fitting for each point in the window, which is used to smooth the data points; m represents the width of the window, and its value changes dynamically and is related to the total number of point clouds in the current section; f[c n ] represents the center index c of each window after filtering n The value of the data point; G(c n ,j) and C(c n ,j) represent the center index c n The gradient and curvature values ​​in this window; G tolerence and C tolerance They represent the tolerances of gradient and curvature respectively. When points that meet the tolerances of gradient and curvature are searched from both sides of the cladding track to the middle, the two points are recorded as weld toe feature points.

6. The forming size detection method of the additive manufacturing cladding track feature point recognition based on the full morphology point cloud as claimed in claim 4, characterized in that: In step S5-2, since the surface of the cladding track is not smooth due to oxidized welding slag, and the fitting method will ignore the small defect features in the contour surface of the point cloud data, the interpolation method is used to completely retain the defect features of the cladding track surface.

7. The forming size detection method of the additive manufacturing cladding track feature point recognition based on the full morphology point cloud as claimed in claim 4, characterized in that: In step S5-2, the heterogeneous region adaptive interpolation method is used to search for edge convex and non-edge convex regions, and parameter reverse Hermite interpolation and forward Hermite interpolation are performed respectively, and the formula is as follows: Wherein, N1 and N2 represent the point sets belonging to the edge convex and non-edge convex regions respectively; a and v a represents the x and z values ​​of the ath point in the N1 point set, u b and v b represents the x value and z value of the bth point in the N2 point set; g a and g b Respectively expressed in v a and u b The gradient value at ; h(t) and lk(z) represent the shape function and Lagrange basis function of the Hermite interpolation formula respectively.

8. The forming size detection method of the additive manufacturing cladding track feature point recognition based on the full morphology point cloud as claimed in claim 4, characterized in that: In step S5-3, the layer intersection dynamic positioning method utilizes the heterogeneous region adaptive interpolation method and combines the binary search idea to bidirectionally search and compare the interpolation and coordinate values ​​of each current point and the next point. When a qualified point is bidirectionally searched from the middle of the cladding track to both sides, the two points are recorded as the undercut feature points.

Citation Information

Patent Citations

  • Connector pin size detection system and method based on three-dimensional machine vision

    CN109967374A

  • Method for detecting error of additive manufacturing cladding layer and correcting track based on point cloud information

    CN115546125A

  • Weld joint detection method based on line laser scanning

    CN117557492A

Cited By

  • Large-curvature thin-wall component cladding layer feature identification method based on three-dimensional point cloud information

    CN120953645A