A weld seam extraction method and system based on three-dimensional geometric feature analysis

Abstract

This specification provides a weld extraction method based on three-dimensional geometric feature analysis. The method includes: acquiring a target three-dimensional model; obtaining a set of contour lines based on the geometric Boolean intersection of each independent entity in the target three-dimensional model; performing a screening process on the contour line set to obtain a weld set, the screening process including a first screening, a segmentation process, and a second screening; determining multiple vertical welds and multiple horizontal welds based on the coordinate differences of each weld in the weld set in a preset direction; determining the horizontal welding parameters of the multiple horizontal welds; and determining the vertical welding parameters based on the horizontal welding parameters, the multiple vertical welds, and the multiple horizontal welds' topological association.

Description

Technical Field

[0001] This specification relates to the field of automated welding technology, and in particular to a method and system for weld seam extraction based on three-dimensional geometric feature analysis. Background Technology

[0002] Existing weld extraction methods typically rely on pre-establishing weld information in the 3D model during the modeling phase. This involves manually annotating the weld location, type, or related parameters as model attributes using specific modeling standards. When parsing the model, these pre-defined weld attributes are directly read to obtain the weld information. However, these methods are highly dependent on the integrity of the modeling process and data, often requiring models from a unified source and consistent modeling standards. This limits their applicability and versatility.

[0003] In practical engineering applications, 3D models may originate from different design software or different historical projects. Some models may not have included weld information during the design phase, or weld attributes may be lost during model conversion, format compression, or data exchange, resulting in the absence of directly readable weld attributes in the model. Under these circumstances, existing weld extraction methods cannot effectively identify weld locations and related information, making it difficult to meet the application requirements of welding process planning, welding quality assessment, and automated welding.

[0004] Therefore, there is an urgent need for a weld extraction method and system based on three-dimensional geometric feature analysis, which can identify and extract welds based solely on the geometric features of the model itself without relying on pre-set weld attribute information in the three-dimensional model, in order to solve the problem that existing technologies cannot obtain weld information when the three-dimensional model does not contain weld information. Summary of the Invention

[0005] This specification provides one or more embodiments of a weld extraction method based on three-dimensional geometric feature analysis, comprising: acquiring a target three-dimensional model; obtaining a set of contour lines based on the geometric Boolean intersection of each independent entity in the target three-dimensional model; performing a screening process on the set of contour lines to obtain a set of welds, the screening process including a first screening, a segmentation process, and a second screening, wherein the first screening and the second screening correspond to different screening objects; determining multiple vertical welds and multiple horizontal welds based on the coordinate difference of each weld in the weld set in a preset direction; determining the horizontal welding parameters of the multiple horizontal welds; and determining the vertical welding parameters based on the horizontal welding parameters, the multiple vertical welds, and the multiple horizontal welds' topological association.

[0006] This specification provides one or more embodiments of a weld extraction system based on three-dimensional geometric feature analysis, comprising: an acquisition module configured to: acquire a target three-dimensional model, and obtain a set of contour lines based on the geometric Boolean intersection of each independent entity in the target three-dimensional model; a screening module configured to: perform screening processing on the set of contour lines to obtain a set of welds, the screening processing including a first screening, a segmentation process, and a second screening, wherein the first screening and the second screening correspond to different screening objects; and a parameter determination module configured to: determine multiple vertical welds and multiple horizontal welds based on the coordinate difference of each weld in the weld set in a preset direction; determine the horizontal welding parameters of the multiple horizontal welds; and determine the vertical welding parameters based on the horizontal welding parameters and the topological association of the multiple vertical welds and the multiple horizontal welds.

[0007] This specification provides one or more embodiments of a weld seam extraction device based on three-dimensional geometric feature analysis, including a processor for executing the above-described weld seam extraction method based on three-dimensional geometric feature analysis.

[0008] This specification provides one or more embodiments of a computer-readable storage medium that stores computer instructions. When a computer reads the computer instructions from the storage medium, the computer executes the above-described weld extraction method based on three-dimensional geometric feature analysis. Attached Figure Description

[0009] This specification will be further described by way of exemplary embodiments, which will be described in detail with reference to the accompanying drawings. These embodiments are not limiting; in these embodiments, the same reference numerals denote the same structures, wherein:

[0010] Figure 1 This is an exemplary block diagram of a weld extraction system based on three-dimensional geometric feature analysis, as shown in some embodiments of this specification;

[0011] Figure 2 This is an exemplary flowchart of a weld extraction method based on three-dimensional geometric feature analysis, as shown in some embodiments of this specification.

[0012] Figure 3 This is an exemplary flowchart illustrating the process of obtaining a weld set according to some embodiments of this specification;

[0013] Figure 4 This is an exemplary schematic diagram illustrating the determination of vertical welding parameters according to some embodiments of this specification;

[0014] Figure 5 This is an exemplary schematic diagram illustrating the determination of flat welding parameters according to some embodiments of this specification. Detailed Implementation

[0015] To more clearly illustrate the technical solutions of the embodiments in this specification, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are merely some examples or embodiments of this specification. For those skilled in the art, these drawings can be applied to other similar scenarios without creative effort. Unless obvious from the context or otherwise specified, the same reference numerals in the drawings represent the same structures or operations.

[0016] It should be understood that the terms “system,” “device,” “unit,” and / or “module” used herein are one way to distinguish different components, elements, parts, sections, or assemblies at different levels. However, if other terms can achieve the same purpose, they may be replaced by other expressions.

[0017] As indicated in this specification and claims, unless the context clearly indicates otherwise, the words "a," "an," "an," and / or "the" do not specifically refer to the singular and may also include the plural. Generally speaking, the terms "comprising" and "including" only indicate the inclusion of expressly identified steps and elements, which do not constitute an exclusive list, and the method or apparatus may also include other steps or elements.

[0018] Flowcharts are used in this specification to illustrate the operations performed by the system according to embodiments of this specification. It should be understood that the preceding or following operations are not necessarily performed in exact order. Instead, the steps can be processed in reverse order or simultaneously. Furthermore, other operations can be added to these processes, or one or more steps can be removed from them.

[0019] Figure 1 This is an exemplary block diagram of a weld extraction system based on three-dimensional geometric feature analysis, as shown in some embodiments of this specification.

[0020] This specification provides a weld extraction system based on three-dimensional geometric feature analysis (hereinafter referred to as the weld extraction system) through some embodiments. The weld extraction system 100 may include an acquisition module 110, a screening module 120, and a parameter determination module 130.

[0021] In some embodiments, the acquisition module 110, the screening module 120, and the parameter determination module 130 are all communicatively connected. Exemplary communication connection methods include, but are not limited to, Bluetooth, WIFI, 5G, etc.

[0022] In some embodiments, the acquisition module 110, the screening module 120, and the parameter determination module 130 may have their own processors or may share a single processor.

[0023] The processor can process data and / or information obtained from other devices or system components. Based on this data, information, and / or processing results, the processor can execute program instructions to perform one or more functions described in this application. In some embodiments, the processor may include one or more sub-processing devices (e.g., a single-core processing device or a multi-core multi-chip processing device). By way of example only, the processor may include a central processing unit, a controller, a microprocessor, or any combination thereof.

[0024] The acquisition module 110 is a module used to acquire the target 3D model and contour set.

[0025] In some embodiments, the acquisition module is configured to acquire a target 3D model and obtain a set of contour lines based on the geometric Boolean intersection of each independent entity in the target 3D model.

[0026] The screening module 120 is a module used for screening and segmenting the contour line set.

[0027] In some embodiments, the screening module is configured to perform screening processing on the contour set to obtain a weld set, the screening processing including a first screening, a segmentation process and a second screening, wherein the screening objects corresponding to the first screening and the second screening are different.

[0028] In some embodiments, the screening module is further configured to: perform a first screening on the contour line set to obtain an initial line segment set; perform the following operations on each initial line segment in the initial line segment set: detect the intersection relationship between the initial line segment and the remaining initial line segments in the initial line segment set; when the intersection relationship is an interior point intersection, divide the initial line segment into multiple sub-line segments based on the intersection point; construct a segmented line segment set based on the initial line segment set and the multiple sub-line segments; and perform a second screening on the segmented line segment set to obtain a weld set.

[0029] In some embodiments, the screening module is further configured to: count the lengths of all contour lines in the contour line set and construct a length distribution map; determine a dynamic screening threshold based on the plate thickness noise peak and weld signal peak in the length distribution map; and remove contour lines with lengths less than the dynamic screening threshold from the contour line set to obtain an initial line segment set.

[0030] The parameter determination module 130 is a module used to determine the flat welding parameters and the vertical welding parameters.

[0031] In some embodiments, the parameter determination module is configured to: determine multiple vertical welds and multiple horizontal welds based on the coordinate differences of each weld in the weld set in a preset direction; determine the horizontal welding parameters of the multiple horizontal welds; and determine the vertical welding parameters based on the horizontal welding parameters and the topological association of the multiple vertical welds and the multiple horizontal welds.

[0032] In some embodiments, the parameter determination module is further configured to: for each of the plurality of flat welds, construct a vertical vector corresponding to the flat weld, the vertical vector passing through the midpoint of the flat weld; extend the vertical vector until the total length of the vertical vector exceeds a preset plate thickness; maintain the scanning direction of the flat weld in response to the vertical vector not intersecting with the parallel boundary line of the independent entity to which the flat weld belongs; and change the scanning direction of the flat weld in response to the vertical vector intersecting with the parallel boundary line of the independent entity to which the flat weld belongs.

[0033] For more information on the weld extraction system and its modules, please refer to [link / reference]. Figures 2-5 And its related descriptions.

[0034] It should be noted that the above description of the weld seam extraction system and its modules is for convenience only and should not be construed as limiting this specification to the scope of the illustrated embodiments. It is understood that those skilled in the art, after understanding the principles of the system, may arbitrarily combine the various modules or construct subsystems connected to other modules without departing from these principles. In some embodiments, Figure 1 The acquisition module 110, screening module 120, and parameter determination module 130 disclosed herein can be different modules within a single system, or a single module can implement the functions of two or more of the aforementioned modules. For example, the modules can share a single storage module, or each module can have its own separate storage module. Such variations are all within the scope of protection of this specification.

[0035] Figure 2 This is an exemplary flowchart illustrating a weld extraction method based on three-dimensional geometric feature analysis, according to some embodiments of this specification. Figure 2 As shown, process 200 includes steps 210-250 as described below. In some embodiments, process 200 may be performed by a weld extraction system.

[0036] Step 210: Obtain the target 3D model and obtain the contour line set based on the geometric Boolean intersection of each independent entity in the target 3D model.

[0037] In some embodiments, step 210 may be performed by the acquisition module. For more information about the acquisition module, see [link to relevant documentation]. Figure 1 And its related descriptions.

[0038] A target 3D model refers to a digital 3D model that represents the geometry, dimensional parameters, and spatial relationships of a welded component. For example, a computer-aided design (CAD) model of a welded component.

[0039] Welded components refer to an integral structure formed by welding together multiple components.

[0040] In some embodiments, the target 3D model can be a 3D model of an H-beam and its auxiliary components. The auxiliary components include members such as flanges, webs, and stiffening plates.

[0041] In some embodiments, the acquisition module can obtain the target 3D model by importing design files or 3D scanning data.

[0042] In some embodiments, the acquisition module can select three non-collinear points on the maximum plane in the target 3D model to determine the spatial position and normal direction of the maximum plane, and establish a 3D coordinate system based on the maximum plane, wherein the two mutually perpendicular directions in the maximum plane are respectively used as the X-axis and Y-axis, and the normal direction perpendicular to the maximum plane is used as the Z-axis; then, based on the established 3D coordinate system, the position and orientation of the target 3D model are uniformly transformed to facilitate subsequent geometric calculations.

[0043] An independent entity refers to a single component that constitutes a welded member in the target 3D model. For example, when the target 3D model is a 3D model of an H-beam and its associated components, each flange, web, and stiffener in the H-beam is an independent entity.

[0044] Geometric Boolean intersection refers to the common geometric portion obtained after performing a Boolean intersection operation on two or more independent entities in a target 3D model. In some embodiments, geometric Boolean intersection can be represented as a common line segment or set of common line segments formed at the points where two independent entities touch or intersect.

[0045] A contour set is a set of all contour lines in the common geometric part of a target 3D model, obtained by the geometric Boolean intersection of all independent entities. Contour lines are line segments.

[0046] In some embodiments, the acquisition module can perform geometric Boolean intersection operations on each pair of independent entities in the target 3D model to extract the contour lines of their common geometric parts, and then collect all the contour lines of the common geometric parts to form a contour line set.

[0047] Step 220: Perform a screening process on the contour set to obtain the weld set.

[0048] In some embodiments, step 220 may be performed by the screening module. For more information about the screening module, see [link to relevant documentation]. Figure 1 And its related descriptions.

[0049] Screening refers to the step-by-step screening and segmentation of each contour line in the contour line set to exclude non-weld contour lines and retain the contour lines corresponding to the weld.

[0050] In some embodiments, the screening process includes a first screening, a segmentation process, and a second screening, wherein the screening objects corresponding to the first screening and the second screening are different. A screening object refers to the set being screened.

[0051] For more information on the first and second screening processes, please see [link to relevant documentation]. Figure 3 And its related descriptions.

[0052] Segmentation refers to dividing a line segment into several continuous sub-segments.

[0053] The weld seam set refers to the set of line segments obtained after the contour line set has been screened.

[0054] In some embodiments, each line segment in the weld set is a contour line corresponding one-to-one with the actual welding position in the target 3D model; therefore, the line segments in the weld set are also referred to as welds. For more information on obtaining weld sets, see [link to relevant documentation]. Figure 3 And its related descriptions.

[0055] Step 230: Based on the coordinate difference of each weld in the weld set in the preset direction, determine multiple vertical welds and multiple horizontal welds.

[0056] In some embodiments, step 230 may be performed by the parameter determination module. For more information about the parameter determination module, see [link to relevant documentation]. Figure 1 And its related descriptions.

[0057] The preset direction refers to the reference direction that is pre-set in the target 3D model.

[0058] In some embodiments, the preset direction may be preset by those skilled in the art.

[0059] In some embodiments, the preset direction can be the Z-axis direction of the three-dimensional coordinate system, that is, the normal direction of the largest plane in the target three-dimensional model.

[0060] Vertical weld refers to a weld whose welding direction is in a vertical plane.

[0061] A flat weld is a weld whose welding direction is in the horizontal plane.

[0062] In some embodiments, the parameter determination module can distinguish between vertical and horizontal welds based on the coordinate difference between the two endpoints of each weld in the weld set in a preset direction. For example, when the coordinate difference between the two endpoints of a weld in the Z-axis direction is greater than a preset threshold, the weld is determined to be a vertical weld; when the coordinate difference between the two endpoints of a weld in the Z-axis direction is less than or equal to the preset threshold, the weld is determined to be a horizontal weld. The preset threshold is manually set. For example, the preset threshold can be 50 mm.

[0063] Step 240: Determine the flat welding parameters for multiple flat welds.

[0064] In some embodiments, step 240 may be performed by the parameter determination module.

[0065] Flat welding parameters refer to the parameters used to guide welding equipment in performing flat welding.

[0066] In some embodiments, the flat welding parameters include the scanning direction and scanning angle of the flat weld.

[0067] The scanning direction refers to the direction in which the welding equipment moves along the flat weld seam. Welding equipment refers to automated or semi-automated devices used to perform welding operations, such as welding robots, welding robotic arms, or CNC welding equipment.

[0068] In some embodiments, the scanning direction can be the direction from the start point of the flat weld to the end point.

[0069] The scanning angle refers to the spatial orientation of the vision sensor relative to the independent entity to which the flat weld belongs when acquiring data from the flat weld. In some embodiments, the scanning angle is used to characterize the orientation of the vision sensor on which side of the independent entity to which the flat weld belongs. For example, the scanning angle can be located on the left or right side of the independent entity to which the flat weld belongs. The independent entity to which the flat weld belongs can be determined based on the target 3D model.

[0070] In some embodiments, for each flat weld, the parameter determination module can select one of the two endpoints of the flat weld as the starting point and the other as the ending point to determine the scanning direction. Since the welding equipment always maintains a fixed direction of movement in the scanning direction (e.g., from left to right), after the scanning direction is determined, the corresponding scanning angle is further determined according to the scanning direction and the independent entity to which the flat weld belongs, thereby obtaining the flat weld parameters.

[0071] For more information on determining flat welding parameters, please refer to [link / reference]. Figure 5 And its related descriptions.

[0072] Step 250: Determine the vertical welding parameters based on the flat welding parameters, multiple vertical welds, and the topological association of multiple flat welds.

[0073] In some embodiments, step 250 may be performed by the parameter determination module.

[0074] Topological association refers to the connection and positional relationships between all line segments in a target 3D model.

[0075] Connection relationship refers to whether the line segments are connected to each other.

[0076] Positional relationship refers to whether line segments are located on the same plane and whether they belong to the same independent entity.

[0077] In some embodiments, topological associations may be determined based on a target 3D model.

[0078] Vertical welding parameters refer to the parameters used to guide welding equipment in performing vertical welds.

[0079] In some embodiments, the vertical welding parameters include the scanning direction.

[0080] In some embodiments, assuming that a flat weld with a determined orientation exists on the plane containing each vertical weld, the parameter determination module can determine the vertical weld parameters based on the scanning direction of the flat weld with the determined orientation and its topological association with the vertical weld. For example, a vertical weld B has a flat weld A with a determined orientation on its plane. The scanning direction is → According to topological association, the flat weld A and the vertical weld B pass through line segment C ( , The connection sequence is as follows: → → → → → The scanning direction of the flat weld is now known to be... → The scanning direction of the vertical weld can be obtained. → This determined the vertical welding parameters for vertical weld B. , These are the two endpoints of the flat weld A; and These are the two endpoints of line segment C; and These are the two endpoints of vertical weld B.

[0081] For more information on how to determine vertical welding parameters when there are no multiple flat welds with a defined direction on the plane where the vertical weld is located, please refer to [link to relevant documentation]. Figure 4 And its related descriptions.

[0082] In some embodiments of this specification, through automated geometric feature extraction and topological correlation analysis, the weld position can be accurately identified from the target 3D model and the welding parameters of various welds can be planned without manual intervention, thereby improving the efficiency and accuracy of welding programming.

[0083] Figure 3 This is an exemplary flowchart illustrating the process of obtaining a weld set according to some embodiments of this specification. Figure 3 As shown, process 300 includes steps 310-340 as described below. In some embodiments, process 300 may be performed by a screening module. For more information about screening modules, see [link to relevant documentation]. Figure 1 And its related descriptions.

[0084] Step 310: Perform the first filtering on the contour line set to obtain the initial line segment set.

[0085] For more information on contour sets, see [link to relevant documentation]. Figure 2 And its related descriptions.

[0086] The first screening refers to the initial filtering of contour lines in the contour line set to remove contour lines that are too short.

[0087] In some embodiments, the first filter is a set of outlines.

[0088] The initial line segment set refers to the set of line segments obtained after the contour line set has undergone a first screening. In some embodiments, the initial line segment set includes multiple initial line segments.

[0089] In some embodiments, the screening module removes contour lines whose length is less than a preset length threshold from the contour line set to obtain an initial line segment set.

[0090] The preset length threshold refers to a reference threshold used to determine whether the outline should be retained when performing the first filtering. In some embodiments, the specific length of the preset length threshold can be preset manually.

[0091] Because the plate thickness parameters and modeling accuracy of individual entities may differ in different target 3D models, the length distribution characteristics of the contour lines will also change accordingly. If a fixed preset length threshold is used for the first screening, it is difficult to simultaneously adapt to target 3D models with different plate thicknesses or modeling errors. Especially when the target 3D model contains multiple plate thicknesses, a fixed threshold may incorrectly retain noise lines that should be removed, or misjudge the contour lines corresponding to welds as noise lines, thus affecting the screening effect. Therefore, in some embodiments, the screening module can perform the first screening based on a dynamic screening threshold.

[0092] Noise lines refer to contour lines that exist in the contour line set but do not correspond to actual welding positions in the target 3D model. For example, thickness lines. Thickness lines are contour lines formed by a component along its own thickness direction.

[0093] In some embodiments, in order to adapt to the changes in the length distribution of contour lines under different plate thicknesses and modeling errors, and to avoid missed or false weld detections caused by fixed threshold screening, the first screening includes: statistically analyzing the lengths of all contour lines in the contour line set and constructing a length distribution map; determining a dynamic screening threshold based on the plate thickness noise peak and weld signal peak in the length distribution map; and removing contour lines with lengths less than the dynamic screening threshold from the contour line set to obtain an initial line segment set.

[0094] In some embodiments, the screening module can calculate the Euclidean distance between the two endpoints of the contour line in the target 3D model, based on the coordinate values ​​of the two endpoints, and use this distance as the length of the contour line. For more information on the target 3D model, see [link to relevant documentation]. Figure 2 And its related descriptions.

[0095] A length distribution chart is a histogram obtained by statistically analyzing the lengths of each contour line in a set of contour lines. In some embodiments, the horizontal axis of the length distribution chart represents the length range of the contour lines, and the vertical axis represents the number of contour lines in the corresponding length range.

[0096] In some embodiments, the screening module first obtains the length information of each contour line in the contour line set and divides the contour line length into multiple length intervals; then, each contour line is assigned to its corresponding length interval, and the number of contour lines in each length interval is counted to construct a length distribution histogram.

[0097] The contour set originates from all contour lines of the common geometric part formed by the geometric Boolean intersection of independent entities in the target 3D model. These contour lines differ significantly in their geometric origin and length scale. One portion of the contour lines corresponds to the contour lines formed by the intersection of independent entities along the plate thickness direction, i.e., plate thickness lines, whose length is constrained by the plate thickness and has a relatively small value. Another portion of the contour lines corresponds to the contour lines formed by the fit of independent entities along their extension directions, i.e., weld seams, whose length is determined by the overall size of the independent entities and is significantly greater than the length of the plate thickness lines. Therefore, when statistically analyzing the lengths of each contour line in the contour set, the length distribution histogram will exhibit two distinct peaks, a "bimodal" distribution characteristic.

[0098] The plate thickness noise peak refers to the peak located on the left side of the length distribution diagram.

[0099] In some embodiments, the outline of the region corresponding to the plate thickness noise peak is a noise line segment, and there is no weld.

[0100] The weld signal peak refers to the peak located on the right side of the length distribution diagram.

[0101] The dynamic filtering threshold is a reference value used to distinguish noise segments from welds in a set of contour lines.

[0102] In some implementations, the dynamic screening threshold is the trough value between the plate thickness noise peak and the weld signal peak in the length distribution diagram.

[0103] In some embodiments, the screening module may be based on the Otsu algorithm to calculate the dynamic screening threshold by setting the plate thickness noise peak as the foreground pixel and the weld signal peak as the background pixel.

[0104] In some embodiments, the screening module removes contour lines from the contour line set whose length is less than the dynamic screening threshold, and the remaining contour lines form the initial line segment set.

[0105] In some embodiments of this specification, a length distribution map is constructed by statistically analyzing the lengths of each contour line in the contour line set. A dynamic screening threshold is then determined based on the plate thickness noise peak and the weld signal peak, thereby overcoming the drawback of using a fixed threshold that makes it difficult to consider different plate thickness conditions or modeling errors. This method can dynamically adjust the screening criteria according to actual data characteristics, effectively distinguishing between weld lines and noise segments, and significantly reducing the missed detection rate and false detection rate during weld extraction.

[0106] In some embodiments, to avoid misjudging local welds when using a global length distribution histogram when the dimensions of H-beams vary significantly—for example, if one end of an H-beam has a large stiffener (20mm thick, 500mm long) and the other end has a small stiffener (6mm thick, 30mm long)—using a global length distribution histogram, the 30mm long weld of the small stiffener might be mistaken for noise from the 20mm thick large stiffener and rejected. Before the screening module constructs the length distribution histogram, the first screening also includes: marking connected components of the target 3D model based on a preset distance tolerance to obtain multiple sub-connected components of the target 3D model; for each of the multiple sub-connected components, constructing a length distribution sub-graph of the sub-connected component and calculating the corresponding local screening threshold.

[0107] The preset distance tolerance is the maximum Euclidean distance allowed when marking connected components, used to determine whether the endpoints of two line segments belong to the same sub-connected component. When the maximum Euclidean distance between the endpoints of two line segments is less than the preset distance tolerance, the screening module determines that the two line segments are connected and belong to the same sub-connected component.

[0108] In some embodiments, the preset distance tolerance can be set by those skilled in the art based on experience. For example, 0.1m, 0.5mm, 2mm, etc.

[0109] Connected component labeling refers to the process of labeling interconnected pixels or geometric elements as the same independent region in an image or geometric data.

[0110] A sub-connected region refers to multiple independent sub-regions obtained after connecting region labeling.

[0111] In some embodiments, the screening module treats a contour line in the target 3D model as an independent computational unit in the connected component labeling algorithm, and performs connectivity analysis on the contour line based on the connected component labeling. The determination of whether two computational units are connected is based on whether the Euclidean distance between the two computational units is less than or equal to a preset distance tolerance. If the Euclidean distance is less than or equal to the preset distance tolerance, they are determined to be connected; if the Euclidean distance is greater than the preset distance tolerance, they are determined to be disconnected.

[0112] In some embodiments, when the screening module performs connected component labeling on the target 3D model based on a preset distance tolerance, it adopts a spatial domain-based connected component labeling method, specifically including the following steps S1-S5.

[0113] Step S1: Initialize the data set and status. The screening module obtains the contour set from the acquisition module and sets an access status flag for each contour in the contour set. The access status flag is used to indicate whether the corresponding contour has been classified into a certain connected component. The access status flag is initially set to "False" and changes to "True" after being accessed.

[0114] Step S2, Seed Segment Selection. The screening module selects a contour line with an access status of False from the contour line set as the seed segment of the current connected component, initializes a sub-connected component to be constructed based on the seed segment, and marks the access status of the seed segment as "True".

[0115] Step S3, Tolerance Neighborhood Search. The screening module starts with the seed line segment and performs a tolerance neighborhood search based on spatial neighborhood relationships. Tolerance neighborhood search refers to calculating the Euclidean distance between the endpoints of the current seed line segment and the endpoints of other unvisited line segments. When the Euclidean distance is less than or equal to a preset distance tolerance, it is determined that the corresponding line segments have a connectivity relationship, and the line segments that meet the connectivity condition are incorporated into the current sub-connected domain. At the same time, the access status of the contour lines incorporated into the current sub-connected domain is marked as "True".

[0116] Step S4: Sub-connected component generation. The screening module continues to perform tolerance-based neighborhood search on the line segments to be incorporated into the current sub-connected component until there are no line segments that satisfy the condition that the distance is less than or equal to the preset distance tolerance, thereby completing the construction of a sub-connected component.

[0117] Step S5, Global Search. After completing the construction of the current sub-connected region, the screening module continues to select the next unvisited contour line from the contour line set, repeating the processing steps S2 to S4 until all contour lines in the contour line set have been classified, thereby obtaining multiple sub-connected regions of the target 3D model.

[0118] A length distribution subgraph is a histogram obtained by statistically analyzing the lengths of all line segments within each sub-connected region. The construction method of the length distribution subgraph is similar to that of the length distribution graph, and will not be elaborated here.

[0119] The local screening threshold is a reference value used to distinguish between noise segments and welds within each sub-connected domain.

[0120] The method for determining the local screening threshold is similar to that for determining the dynamic screening threshold, and will not be elaborated here.

[0121] In some embodiments, for each sub-connected region, the screening module calculates the local screening threshold corresponding to the sub-connected region, then removes the contour lines in the sub-connected region whose length is less than the local screening threshold, and retains the contour lines whose length is greater than or equal to the local screening threshold; then, the screening module gathers the contour lines left in each sub-connected region to construct an initial set of line segments.

[0122] In some embodiments of this specification, by dividing the target 3D model into connected components and independently constructing length distribution subgraphs and calculating local screening thresholds within each sub-connected component, the problem of the global dynamic screening threshold not being suitable for components of different scales is avoided. This prevents effective welds of small-sized components from being mistakenly rejected by plate thickness noise generated by large-sized components, thereby improving the accuracy and robustness of weld screening in complex models.

[0123] In some embodiments, the first screening further includes: in response to the difference between the length of the contour line and the dynamic screening threshold being within a preset range, determining a closed geometric loop of the contour line based on the target 3D model; obtaining the lengths of the remaining line segments within the closed geometric loop; and removing the contour line from the contour line set if the length of the contour line is the shortest line segment in the closed geometric loop.

[0124] The preset range refers to the range of positive and negative deviations allowed between the length of the outline and the dynamic filtering threshold.

[0125] In some embodiments, the preset range may be set by those skilled in the art based on experience.

[0126] In some embodiments, the preset range may be determined based on a length distribution map.

[0127] In some embodiments, the preset range is negatively correlated with the distance between the plate thickness noise peak and the weld signal peak in the length distribution diagram.

[0128] For example, when the distance between the plate thickness noise peak and the weld signal peak is large, it indicates that the noise line segment and the weld have a high degree of differentiation in length distribution. In this case, a smaller preset range can be set. When the distance between the plate thickness noise peak and the weld signal peak is small, it indicates that the noise line segment and the weld have a low degree of differentiation in length distribution. In this case, a larger preset range needs to be set so that more contour lines with similar length values ​​can enter the subsequent screening.

[0129] The distance between the plate thickness noise peak and the weld signal peak refers to the difference in length of the contour lines corresponding to the plate thickness noise peak and the weld signal peak, that is, the difference in the horizontal coordinates corresponding to the plate thickness noise peak and the weld signal peak.

[0130] In some embodiments of this specification, the preset range is dynamically adjusted through the length distribution map, which can reasonably allocate computing resources while ensuring the screening accuracy, and automatically enhance the inspection efforts for difficult scenarios with high data distribution overlap.

[0131] A closed geometric loop is a closed polygonal loop formed by connecting multiple contour lines end to end.

[0132] In some embodiments, the screening module can start from the endpoints of contour lines whose length difference from the dynamic screening threshold is within a preset range, and gradually trace other contour lines connected to their endpoints based on the line segment connection relationships in the target 3D model until a closed loop is formed, thereby determining a closed geometric loop.

[0133] In some embodiments, the screening module can calculate the Euclidean distance between the two endpoints of each line segment in the closed geometric loop based on the coordinate values ​​of the two endpoints in the three-dimensional coordinate system, and use this distance as the length of the line segment.

[0134] In some embodiments, in the structural design of H-beams, welds are used to connect and reinforce components. Their length is typically greater than the thickness of the component's plate. If the weld length is shorter than the plate thickness line, it is difficult for the weld to provide effective structural reinforcement. Based on the above structural design principle, the screening module compares the length relationships of the contour lines within a closed geometric loop. When a contour line is the shortest segment in the closed geometric loop, it is removed from the contour line set, thereby preventing the plate thickness line of the component from being mistakenly identified as a weld.

[0135] In some embodiments of this specification, by introducing a line segment length comparison mechanism within a closed geometric loop, when the outline length is close to the dynamic screening threshold, further discrimination is made by combining structural topological features, which effectively avoids identifying the plate thickness line of the component as a weld, thereby improving the accuracy and stability of the weld extraction results.

[0136] Step 320: Perform the following operations on each initial line segment in the initial line segment set.

[0137] In some embodiments, performing the following operations on each initial line segment in the initial line segment set includes performing steps 321 and 322 on each initial line segment in the initial line segment set. In some embodiments, steps 321 and 322 are performed by a screening module.

[0138] Step 321: Detect the intersection relationship between the initial line segment and the remaining initial line segments in the initial line segment set.

[0139] The initial line segment refers to the outline that is retained after the first screening.

[0140] Intersection relationship refers to the relative relationship between two line segments in space based on their geometric positions. In some embodiments, intersection relationship is used to characterize whether there is an intersection point between the line segments and the type of position of the intersection point.

[0141] In some embodiments, the intersection relationship includes parallelism, endpoint intersection, and interior point intersection. Parallelism means that the two line segments are oriented in the same or opposite directions in space and do not intersect in their extended directions. Interior point intersection means that the intersection point of the two line segments is located inside at least one of the line segments rather than at its endpoint. Endpoint intersection means that the intersection point of the two line segments is located at the endpoint of one of the line segments.

[0142] In some embodiments, the screening module can obtain the intersection relationship between each initial line segment based on the geometric properties (starting point coordinates, ending point coordinates, etc.) of each initial line segment in the target 3D model.

[0143] For example, for any two initial line segments A and B, the screening module obtains the start and end coordinates of initial line segments A and B in the three-dimensional coordinate system, respectively, denoted as A(P1,P2) for initial line segment A and B(Q1,Q2) for initial line segment B. The two initial line segments are vectorized to obtain direction vectors u=P1-P2, v=Q1-Q2; based on the direction vectors u and v, a squared distance function is constructed for the two line segments. ,in, and Let these be the parameter variables to be solved. Then, by calculating partial derivatives or vector dot products, the solution is obtained. To reach the minimum parameter value and ; in the solution of parameter values and After limiting it to the range [0,1], if The parameter corresponding to the minimum The difference between 0 and 0 is less than a preset difference threshold, and the parameter If the points are within the range (0,1), then they are considered to be intersecting interior points; if The parameter corresponding to the minimum If the difference between the original value and 0 is greater than a preset difference threshold, it is considered parallel; if... The parameter corresponding to the minimum It is 0 or 1, and the parameter If the difference between the two values ​​is less than a preset difference threshold, the endpoints are considered to intersect. The preset difference threshold is manually set.

[0144] Step 322: When the intersection relationship is interior point intersection, the initial line segment is divided into multiple sub-segments based on the intersection points.

[0145] A sub-segment is a segment formed by dividing an initial segment at its intersection point.

[0146] In some embodiments, the screening module uses the intersection point as the dividing point to divide the initial line segment where the interior points intersect into two sub-segments. One sub-segment is formed by the starting point of the initial line segment and the intersection point, and the other sub-segment is formed by the intersection point and the ending point of the initial line segment.

[0147] Step 330: Based on the initial set of line segments and multiple sub-line segments, construct a set of segmented line segments.

[0148] A segmented line segment set refers to a set of line segments consisting of the initial line segments that have not been segmented in the initial line segment set, and the sub-line segments obtained by segmenting the initial line segments that intersect at interior points.

[0149] In some embodiments, the screening module directly retains the initial line segments that have not been divided in the initial line segment set, and adds all the sub-line segments obtained by dividing based on the intersection points to form a divided line segment set.

[0150] Step 340: Perform a second filtering on the segment set to obtain the weld set.

[0151] The second screening refers to filtering the line segments in the segment set to remove those whose lengths do not meet the requirements.

[0152] In some embodiments, the filtering object of the second filter is a set of dividing line segments.

[0153] For the definition of weld sets, see [link to definition]. Figure 2 And its related descriptions.

[0154] In some embodiments, the screening module removes initial line segments from the segment set whose length is less than a preset length threshold, thus obtaining a weld set. For more information on the preset length threshold, see step 310 and its related description.

[0155] In some embodiments of this specification, after the first screening, the intersection relationship between each initial line segment in the initial line segment set and the other initial line segments is detected. When the intersection relationship is an interior point intersection, the initial line segment is segmented based on the intersection point. Then, the segmented sub-line segments are combined to construct a segmented line segment set and a second screening is performed, thus achieving refined processing of the line segment topology. Therefore, when a long weld seam intersects with a plate thickness line, segmenting the initial line segment based on the intersection relationship can first eliminate the topological ambiguity introduced by the intersection position. Then, the subsequent second screening effectively removes line segments with a length less than a preset length threshold from the segmented line segment set, avoiding incorrect cutting or misidentification of the weld seam, thereby improving the purity and stability of the weld seam extraction results.

[0156] It should be noted that the above descriptions of processes 200 and 300 are for illustrative purposes only and do not limit the scope of this specification. Those skilled in the art can make various modifications and changes to processes 200 and 300 under the guidance of this specification. However, these modifications and changes remain within the scope of this specification.

[0157] Figure 4 This is an exemplary schematic diagram illustrating the determination of vertical welding parameters according to some embodiments of this specification.

[0158] In some embodiments, for each of the plurality of vertical welds, in response to the existence of a plurality of flat welds with a determined direction in the plane where the vertical weld is located, vertical welding parameters 420 are determined based on topological association 410; in response to the absence of a plurality of flat welds with a determined direction in the plane where the vertical weld is located, a first detection region 430 and a second detection region 440 are constructed with the two endpoints of the vertical weld as centers respectively; a first feature value 450 and a second feature value 460 are determined based on the number of independent planes and solid angles within the first detection region 430 and the second detection region 440; and vertical welding parameters 420 are determined based on the first feature value 450 and the second feature value 460.

[0159] For more information on topological association, vertical welding parameters, and how to determine vertical welding parameters based on topological association when there are multiple flat welds with a defined direction on the plane containing the vertical weld, please refer to [link to relevant documentation]. Figure 2 And its related descriptions.

[0160] The first detection area refers to a spherical detection area constructed with a predetermined radius, centered at one end of the vertical weld.

[0161] The second detection area refers to a spherical detection area constructed with the other end of the vertical weld as the center and a preset radius.

[0162] In some embodiments, the preset radius is preset by those skilled in the art based on experience.

[0163] An independent plane is a flat surface region in three-dimensional space enclosed by a closed boundary and possessing a specific normal vector. This specific normal vector is one that is not parallel to the normal vectors of any other plane.

[0164] The number of independent planes refers to the number of independent planes within the first detection area or the second detection area.

[0165] In some embodiments, the parameter determination module can obtain all planes involved in the first detection area or the second detection area, and extract the normal vector of each plane; for multiple planes with the same or opposite normal vector directions, only one plane is retained and the rest are discarded; the retained planes are counted to obtain the number of independent planes.

[0166] A solid angle is the projected area of ​​a unit sphere constructed with the center of the detection area as its vertex, onto which the portion of the detection area not obscured by other independent entities is projected. In some embodiments, a larger solid angle indicates a more open space available for welding operations at that endpoint; a smaller solid angle indicates a stronger spatial constraint at that endpoint.

[0167] The first eigenvalue refers to a numerical index that characterizes the degree of spatial constraint of the first detection area corresponding to the endpoint of a vertical weld.

[0168] The second eigenvalue refers to a numerical index that characterizes the degree of spatial constraint in the second detection area corresponding to the endpoint of a vertical weld.

[0169] In some embodiments, the parameter determination module can normalize the number of independent planes and solid angles within the first detection area corresponding to a certain vertical weld endpoint, and then perform a weighted summation of the normalized results according to preset weights to obtain a first feature value corresponding to a certain vertical weld endpoint. The weighting coefficients can be set empirically to balance the influence of different spatial constraints on the feature value.

[0170] The method for determining the second eigenvalue is the same as that for determining the first eigenvalue, and will not be repeated here.

[0171] In some embodiments, based on the first and second characteristic values, the parameter determination module can determine the vertical welding parameters in various ways. For example, the parameter determination module can determine the endpoint of the vertical weld corresponding to the characteristic value with the larger value as the starting point of the vertical weld, and the endpoint of the vertical weld corresponding to the characteristic value with the smaller value as the ending point of the vertical weld, according to the relative magnitude relationship between the first and second characteristic values. The parameter determination module uses the direction from the starting point to the ending point of the vertical weld as the scanning direction of the vertical welding parameters.

[0172] In some embodiments, in response to the difference between the first feature value and the second feature value being less than a preset difference threshold, the first empty distance and the second empty distance corresponding to the vertical weld are calculated. The first empty distance corresponds to the first scanning direction, and the second empty distance corresponds to the second scanning direction. The first scanning direction is different from the second scanning direction. Based on the first scanning direction, the second scanning direction, the first empty distance and the second empty distance, the vertical weld parameters are determined.

[0173] A preset difference threshold refers to the boundary characterizing the degree of difference between a first characteristic value and a second characteristic value. In some embodiments, the preset difference threshold can be set based on experience or determined after statistical analysis based on historical welding data.

[0174] In some embodiments, when the difference between the first feature value and the second feature value is less than a preset difference threshold, it indicates that the spatial constraints of the two ends of the vertical weld are similar and difficult to distinguish.

[0175] The idle distance refers to the path length that welding equipment travels from its current position or the end of the previous weld to the starting point of the target vertical weld without performing welding operations.

[0176] The first empty stroke distance refers to the empty stroke distance when the scanning direction of the vertical weld is the first scanning direction.

[0177] The second empty stroke distance refers to the empty stroke distance when the scanning direction of the vertical weld is the second scanning direction.

[0178] In some embodiments, the parameter determination module can obtain the nearest adjacent weld with a determined welding direction that is closest to the spatial position of the vertical weld, and obtain the welding end coordinates of the adjacent weld; calculate the Euclidean distance between the coordinates of the adjacent welding end and the corresponding starting coordinates of the vertical weld in the first scanning direction and the second scanning direction, respectively, as the first travel distance and the second travel distance.

[0179] The first scanning direction refers to the direction from one end of the vertical weld to the other end of the vertical weld.

[0180] The second scanning direction refers to the direction opposite to the first scanning direction.

[0181] In some embodiments, the parameter determination module determines the scanning direction corresponding to the smaller of the first and second idle distances by comparing the first and second idle distances, and uses this scanning direction as the scanning direction of the vertical weld, thereby determining the vertical weld parameters.

[0182] In some embodiments of this specification, when the difference between the first feature value and the second feature value is less than a preset difference threshold, the first idle distance and the second idle distance corresponding to the vertical weld in the first scanning direction and the second scanning direction are calculated respectively, and the scanning direction is selected based on the idle distance, so that the welding direction of the vertical weld is more reasonably connected with the path of the adjacent weld, thereby effectively shortening the movement path of the welding equipment in the non-welding state, reducing the invalid travel time, improving the continuity of welding path planning and the overall production efficiency of welding operation.

[0183] Figure 5 This is an exemplary schematic diagram illustrating the determination of flat welding parameters according to some embodiments of this specification.

[0184] In some embodiments, for each of the plurality of flat welds, a vertical vector corresponding to the flat weld is constructed, the vertical vector passing through the midpoint of the flat weld; the vertical vector is extended until the total length of the vertical vector exceeds a preset plate thickness; in response to the vertical vector not intersecting with the parallel boundary line of the independent entity to which the flat weld belongs, the scanning direction of the flat weld is maintained; in response to the vertical vector intersecting with the parallel boundary line of the independent entity to which the flat weld belongs, the scanning direction of the flat weld is changed.

[0185] A perpendicular vector is a vector located within the contact surface of a flat weld, originating from the midpoint of the flat weld and perpendicular to the weld.

[0186] The contact surface where the flat weld is located refers to the contact surface that includes the flat weld among the multiple contact surfaces of the independent entity to which the flat weld belongs and other independent entities.

[0187] Multiple contact surfaces between the independent entity to which the flat weld belongs and other independent entities can be obtained through the target 3D model.

[0188] In some embodiments, the direction of the vertical vector is the same as the scanning angle. For example, the scanning angle is to the left of the independent entity to which the flat weld belongs, and the vertical vector corresponding to the flat weld points to the left within the contact surface.

[0189] The preset plate thickness refers to the thickness of the independent entity to which the flat weld belongs. In some embodiments, the preset plate thickness can be set based on the structural parameters or empirical data of the target welded component.

[0190] The parallel boundary line of an independent entity refers to the outline line located within the contact surface of the flat weld and parallel to the flat weld.

[0191] In some embodiments, the parameter determination module can determine the independent entity to which the flat weld belongs based on the target three-dimensional model, extract the contact surface between the independent entity and other independent entities, and screen out the contour lines parallel to the direction of the flat weld in the contact surface as the parallel boundary lines of the independent entity to which the flat weld belongs, for subsequent detection of the intersection relationship with the perpendicular vector.

[0192] In some embodiments, the parameter determination module can determine whether the current scanning angle points towards the interior of an independent entity by detecting whether the perpendicular vector intersects with the parallel boundary line. If the perpendicular vector intersects with the parallel boundary line of the independent entity to which the flat weld belongs, it indicates that the current scanning angle points towards the interior of the independent entity. Welding equipment performing welding from this scanning angle may collide with the steel plate, therefore the scanning direction of the flat weld needs to be changed. If the perpendicular vector does not intersect with the parallel boundary line of the independent entity to which the flat weld belongs, it indicates that the current scanning angle does not point towards the interior of the independent entity, and there is no need to change the scanning direction of the flat weld.

[0193] In some embodiments, changing the scanning direction of the flat weld seam involves exchanging the start and end points of the flat weld seam, thereby causing the welding equipment to scan the flat weld seam in a direction opposite to the original scanning direction. Since the welding equipment maintains a fixed direction of movement (e.g., from left to right) in the weld seam scanning direction, once the scanning direction is determined, the corresponding scanning angle (i.e., the direction of the perpendicular vector) is also determined. When the perpendicular vector of the flat weld seam intersects the parallel boundary line of the independent entity to which the flat weld seam belongs in the current scanning direction, it indicates that the welding equipment, at that scanning angle, will pass through the independent entity from the side excluding the flat weld seam and perform welding on the other side of the independent entity, posing a risk of collision, which does not conform to actual welding conditions. Therefore, the scanning direction is changed by exchanging the start and end points of the flat weld seam. After the scanning direction is changed, while the welding equipment maintains the same fixed direction of movement, the scanning angle corresponding to the flat weld seam is flipped relative to the independent entity, causing the perpendicular vector to point outwards from the flat weld seam and no longer intersect the parallel boundary line, thus achieving safe welding.

[0194] In some embodiments of this specification, by constructing a perpendicular vector at the midpoint of the flat weld and detecting its intersection with the parallel boundary line of the independent entity to which it belongs, the scanning direction of the flat weld can be automatically corrected based on the geometric position relationship. This avoids the welding equipment scanning in a direction pointing into the interior of the independent entity, thereby ensuring that the welding equipment has a reasonable spatial orientation when performing flat weld welding, reducing the risk of interference or collision, and improving the stability and reliability of the welding operation.

[0195] The basic concepts have been described above. Obviously, for those skilled in the art, the detailed disclosure above is merely illustrative and does not constitute a limitation of this specification. Although not explicitly stated herein, those skilled in the art may make various modifications, improvements, and corrections to this specification. Such modifications, improvements, and corrections are suggested in this specification and therefore remain within the spirit and scope of the exemplary embodiments described herein.

[0196] Furthermore, this specification uses specific terms to describe embodiments thereof. For example, "an embodiment," "one embodiment," and / or "some embodiments" refer to a particular feature, structure, or characteristic associated with at least one embodiment of this specification. Therefore, it should be emphasized and noted that references to "an embodiment," "one embodiment," or "an alternative embodiment" in different locations throughout this specification do not necessarily refer to the same embodiment. Moreover, certain features, structures, or characteristics in one or more embodiments of this specification can be appropriately combined.

[0197] Furthermore, unless expressly stated in the claims, the order of processing elements and sequences, the use of numbers and letters, or other names described in this specification are not intended to limit the order of the processes and methods described herein. Although various examples have been discussed in the foregoing disclosure of some embodiments of the invention that are currently considered useful, it should be understood that such details are for illustrative purposes only, and the appended claims are not limited to the disclosed embodiments; rather, the claims are intended to cover all modifications and equivalent combinations that conform to the spirit and scope of the embodiments described herein. For example, while the system components described above can be implemented using hardware devices, they can also be implemented solely using software solutions, such as installing the described system on existing servers or mobile devices.

[0198] Similarly, it should be noted that, in order to simplify the description disclosed herein and thus aid in the understanding of one or more embodiments of the invention, the foregoing description of embodiments in this specification may sometimes combine multiple features into a single embodiment, drawing, or description thereof. However, this method of disclosure does not imply that the subject matter of this specification requires more features than those mentioned in the claims. In fact, the embodiments contain fewer features than all the features of a single embodiment disclosed above.

[0199] In some embodiments, numbers describing the quantity of components and attributes are used. It should be understood that such numbers used in the description of embodiments are modified in some examples with the terms "approximately," "approximately," or "generally." Unless otherwise stated, "approximately," "approximately," or "generally" indicates that the numbers are allowed to vary by ±20%. Accordingly, in some embodiments, the numerical parameters used in the specification and claims are approximate values, which may be changed depending on the characteristics required by individual embodiments. In some embodiments, numerical parameters should take into account specified significant digits and employ a general method of digit reservation. Although the numerical ranges and parameters used to confirm their breadth of range in some embodiments of this specification are approximate values, in specific embodiments, such values ​​are set as precisely as feasible.

[0200] For each patent, patent application, patent application publication, and other material such as articles, books, specifications, publications, and documents referenced in this specification, the entire contents of which are incorporated herein by reference. This excludes historical application documents that are inconsistent with or conflict with the content of this specification, as well as documents that limit the broadest scope of the claims in this specification (currently or subsequently appended to this specification). It should be noted that in the event of any inconsistency or conflict between the descriptions, definitions, and / or terminology used in the supplementary materials to this specification and the content of this specification, the descriptions, definitions, and / or terminology used in this specification shall prevail.

[0201] Finally, it should be understood that the embodiments described in this specification are merely illustrative of the principles of the embodiments described herein. Other variations may also fall within the scope of this specification. Therefore, alternative configurations of the embodiments described herein are intended to be illustrative rather than limiting, and should be considered consistent with the teachings of this specification. Accordingly, the embodiments described herein are not limited to those explicitly introduced and described herein.

Claims

1. A weld extraction method based on three-dimensional geometric feature analysis, comprising: Obtain the target 3D model, and obtain the contour line set based on the geometric Boolean intersection of each independent entity in the target 3D model; A screening process is performed on the contour line set to obtain a weld set. The screening process includes a first screening, a segmentation process, and a second screening, wherein the screening objects corresponding to the first screening and the second screening are different. The step of performing a screening process on the contour set to obtain the weld set includes: The first filtering is performed on the set of contour lines to obtain an initial set of line segments; the first filtering includes: Calculate the lengths of all contour lines in the contour line set and construct a length distribution map; Based on the plate thickness noise peak and weld signal peak in the length distribution map, a dynamic screening threshold is determined. The initial line segment set is obtained by removing the contour lines whose length is less than the dynamic filtering threshold from the contour line set. Perform the following operations on each initial line segment in the initial line segment set: Detect the intersection relationship between the initial line segment and the remaining initial line segments in the set of initial line segments; When the intersection relationship is an interior point intersection, the initial line segment is divided into multiple sub-line segments based on the intersection points; Based on the initial set of line segments and the multiple sub-line segments, construct a set of segmented line segments; The second filtering is performed on the set of segmented lines to obtain the set of welds; Based on the coordinate difference of each weld in the weld set in a preset direction, multiple vertical welds and multiple horizontal welds are determined. Determine the flat welding parameters for the plurality of flat welds; Based on the flat welding parameters, the plurality of vertical welds, and the topological association of the plurality of flat welds, the vertical welding parameters are determined, including: For each of the multiple vertical welds: In response to the existence of multiple flat welds with a determined direction in the plane where the vertical weld is located, the vertical weld parameters are determined based on topological association. In response to the absence of multiple flat welds in the determined direction on the plane where the vertical weld is located, a first detection area and a second detection area are constructed with the two endpoints of the vertical weld as centers, respectively; based on the number of independent planes and solid angles within the first and second detection areas, a first feature value and a second feature value are determined; in response to the difference between the first feature value and the second feature value being less than a preset difference threshold, a first empty distance and a second empty distance corresponding to the vertical weld are calculated, wherein the first empty distance corresponds to a first scanning direction and the second empty distance corresponds to a second scanning direction, and the first scanning direction is different from the second scanning direction; based on the first scanning direction, the second scanning direction, the first empty distance, and the second empty distance, the vertical weld parameters are determined.

2. The method of claim 1, wherein, Determining the flat welding parameters of the plurality of flat welds includes: For each of the plurality of flat welds Construct a perpendicular vector corresponding to the flat weld, the perpendicular vector passing through the midpoint of the flat weld; The total length extending from the perpendicular vector to the perpendicular vector exceeds the preset plate thickness; In response to the fact that the vertical vector does not intersect the parallel boundary line of the independent entity to which the flat weld belongs, the scanning direction of the flat weld is maintained; In response to the intersection of the vertical vector with the parallel boundary line of the independent entity to which the flat weld belongs, the scanning direction of the flat weld is reversed.

3. A weld seam extraction system based on three-dimensional geometric feature analysis, characterized in that, include: The acquisition module is configured as follows: Obtain the target 3D model, and obtain the contour line set based on the geometric Boolean intersection of each independent entity in the target 3D model; The screening module is configured as follows: A screening process is performed on the contour line set to obtain a weld set. The screening process includes a first screening, a segmentation process, and a second screening, wherein the screening objects corresponding to the first screening and the second screening are different. The first filtering is performed on the set of contour lines to obtain an initial set of line segments; the first filtering includes: Calculate the lengths of all contour lines in the contour line set and construct a length distribution map; Based on the plate thickness noise peak and weld signal peak in the length distribution map, a dynamic screening threshold is determined. The initial line segment set is obtained by removing the contour lines whose length is less than the dynamic filtering threshold from the contour line set. Perform the following operations on each initial line segment in the initial line segment set: Detect the intersection relationship between the initial line segment and the remaining initial line segments in the set of initial line segments; When the intersection relationship is an interior point intersection, the initial line segment is divided into multiple sub-line segments based on the intersection points; Based on the initial set of line segments and the multiple sub-line segments, construct a set of segmented line segments; The second filtering is performed on the set of segmented lines to obtain the set of welds; The parameter determination module is configured as follows: Based on the coordinate difference of each weld in the weld set in a preset direction, multiple vertical welds and multiple horizontal welds are determined. Determine the flat welding parameters for the plurality of flat welds; Based on the flat welding parameters, the topological association between the plurality of vertical welds and the plurality of flat welds, the vertical welding parameters are determined, including: For each of the multiple vertical welds: In response to the existence of multiple flat welds with a determined direction in the plane where the vertical weld is located, the vertical weld parameters are determined based on topological association. In response to the absence of multiple flat welds in the determined direction on the plane where the vertical weld is located, a first detection area and a second detection area are constructed with the two endpoints of the vertical weld as centers, respectively; based on the number of independent planes and solid angles within the first and second detection areas, a first feature value and a second feature value are determined; in response to the difference between the first feature value and the second feature value being less than a preset difference threshold, a first empty distance and a second empty distance corresponding to the vertical weld are calculated, wherein the first empty distance corresponds to a first scanning direction and the second empty distance corresponds to a second scanning direction, and the first scanning direction is different from the second scanning direction; based on the first scanning direction, the second scanning direction, the first empty distance, and the second empty distance, the vertical weld parameters are determined.

4. The system according to claim 3, characterized in that, The parameter determination module is further configured to: For each of the plurality of flat welds Construct a perpendicular vector corresponding to the flat weld, the perpendicular vector passing through the midpoint of the flat weld; The total length extending from the perpendicular vector to the perpendicular vector exceeds the preset plate thickness; In response to the fact that the vertical vector does not intersect the parallel boundary line of the independent entity to which the flat weld belongs, the scanning direction of the flat weld is maintained; In response to the intersection of the vertical vector with the parallel boundary line of the independent entity to which the flat weld belongs, the scanning direction of the flat weld is reversed.

5. A weld extraction device based on three-dimensional geometric feature analysis, comprising a processor, the processor being used to execute the weld extraction method based on three-dimensional geometric feature analysis as described in any one of claims 1 to 2.

6. A computer-readable storage medium storing computer instructions, wherein when a computer reads the computer instructions in the storage medium, the computer executes the weld extraction method based on three-dimensional geometric feature analysis as described in any one of claims 1 to 2.

Images