Straight weld seam recognition method and device

Abstract

The application provides a straight weld joint identification method and device, and relates to the technical field of computers. The method comprises the following steps: calculating a first intersection line according to plane equations of a first initial plane and a second initial plane obtained by performing plane detection on a to-be-identified point cloud; determining a line segment from the first intersection line as an initial weld joint according to the distribution of points in a first point cloud corresponding to the first initial plane and a second point cloud corresponding to the second initial plane in the to-be-identified point cloud; and adjusting each initial endpoint of each initial weld joint according to a local plane in a local point cloud corresponding to the initial endpoint in the to-be-identified point cloud, so as to obtain a target endpoint corresponding to the initial endpoint, and two target endpoints corresponding to one initial weld joint are two endpoints of one target weld joint. In this way, the two endpoints of the initial weld joint determined based on the intersection line of the planes are adjusted, so that a more accurate straight weld joint position is obtained.

Description

Technical Field

[0001] This application relates to the field of computer technology, and more specifically, to a method and apparatus for identifying straight weld seams. Background Technology

[0002] Compared to manual welding, robotic welding offers advantages such as higher efficiency and precision. Therefore, with technological advancements, robotic welding has been increasingly widely applied across various fields. To ensure the welding quality of straight weld seams by robotic welding robots, it is crucial to accurately determine the start and end points of the weld seam. Consequently, improving the accuracy of detected straight weld seams has become a pressing technical challenge for those skilled in the art. Summary of the Invention

[0003] The purpose of this application includes, for example, providing a method, apparatus, computer device, and readable storage medium for identifying straight welds, which obtains a more accurate position of a straight weld by locally adjusting the two endpoints of an initial weld determined based on a plane intersection.

[0004] The embodiments of this application can be implemented as follows:

[0005] In a first aspect, embodiments of this application provide a method for identifying straight weld seams, the method comprising:

[0006] The first intersection line is calculated based on the plane equations of the first and second initial planes obtained by plane detection of the point cloud to be identified.

[0007] Based on the distribution of points in the first point cloud corresponding to the first initial plane and the second point cloud corresponding to the second initial plane in the point cloud to be identified, a line segment is determined from the first intersection line as the initial weld.

[0008] For each initial endpoint of an initial weld, the initial endpoint is adjusted according to the local plane in the local point cloud corresponding to the initial endpoint in the point cloud to be identified, so as to obtain the target endpoint corresponding to the initial endpoint. The two target endpoints corresponding to an initial weld are the two endpoints of a target weld, and the target weld is a straight weld.

[0009] Secondly, embodiments of this application provide a straight weld seam identification device, the device comprising:

[0010] The calculation module is used to calculate the first intersection line based on the plane equations of the first initial plane and the second initial plane obtained by plane detection of the point cloud to be identified;

[0011] The detection module is used to determine a line segment as the initial weld from the first intersection line based on the distribution of points in the first point cloud corresponding to the first initial plane and the second point cloud corresponding to the second initial plane in the point cloud to be identified.

[0012] The local point adjustment module is used to adjust each initial endpoint of each initial weld seam according to the local plane in the local point cloud corresponding to the initial endpoint in the point cloud to be identified, so as to obtain the target endpoint corresponding to the initial endpoint. The two target endpoints corresponding to an initial weld seam are the two endpoints of a target weld seam, and the target weld seam is a straight weld seam.

[0013] Thirdly, embodiments of this application provide a computer device including a processor and a memory, wherein the memory stores machine-executable instructions that can be executed by the processor, and the processor can execute the machine-executable instructions to implement the straight weld seam identification method described in the foregoing embodiments.

[0014] Fourthly, embodiments of this application provide a readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the straight weld seam identification method as described in the foregoing embodiments.

[0015] The method, apparatus, computer device, and readable storage medium for identifying straight welds provided in this application first calculate a first intersection line based on the equations of two initial planes obtained by planar detection of the point cloud to be identified. Then, based on the distribution of points in the point cloud corresponding to each of the two initial planes, a line segment is determined from the first intersection line as an initial weld. Finally, for each initial weld, for each of the two initial endpoints of the initial weld, the initial endpoint is adjusted according to a local plane in the local point cloud corresponding to that initial endpoint in the point cloud to be identified, and the resulting point is taken as the target endpoint corresponding to that initial endpoint. Here, the two target endpoints corresponding to an initial weld are the two endpoints of a target weld, and the target weld is a straight weld. Thus, by adjusting the local point positions of the two endpoints of the initial weld determined based on the planar intersection line, a more accurate straight weld position is obtained, avoiding the situation where the initial weld is significantly offset from the actual straight weld when directly using the initial weld as the identified straight weld. Attached Figure Description

[0016] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 A block diagram of a computer device provided in an embodiment of this application;

[0018] Figure 2 A schematic flowchart illustrating the straight weld seam identification method provided in this application embodiment;

[0019] Figure 3 This is a schematic diagram of line segment offset provided in an embodiment of this application;

[0020] Figure 4 for Figure 2 A flowchart illustrating the sub-steps included in step S200;

[0021] Figure 5 for Figure 4 A flowchart illustrating the sub-steps included in the neutron step S230;

[0022] Figure 6 for Figure 2 A flowchart illustrating the sub-steps included in step S300;

[0023] Figure 7 for Figure 6 A flowchart illustrating the sub-steps included in the neutron step S310;

[0024] Figure 8 A schematic diagram of a cylinder based on two reference points provided in an embodiment of this application;

[0025] Figure 9 for Figure 6 A flowchart illustrating the sub-steps included in the neutron step S330;

[0026] Figure 10 for Figure 9 A flowchart illustrating the sub-steps included in the neutron step S332;

[0027] Figure 11 A block diagram of a straight weld seam recognition device provided in an embodiment of this application.

[0028] Icons: 100 - Computer equipment; 110 - Memory; 120 - Processor; 130 - Communication unit; 200 - Straight weld seam identification device; 210 - Calculation module; 220 - Detection module; 230 - Local point adjustment module. Detailed Implementation

[0029] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. The components of the embodiments of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0030] Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of the application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0031] It should be noted that relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0032] The following detailed description of some embodiments of this application is provided in conjunction with the accompanying drawings. Unless otherwise specified, the following embodiments and features can be combined with each other.

[0033] Please refer to Figure 1 , Figure 1 This is a block diagram of a computer device 100 provided in an embodiment of this application. The computer device 100 may be, but is not limited to, a computer, a server, a welding robot, etc. The computer device 100 includes a memory 110, a processor 120, and a communication unit 130. The memory 110, processor 120, and communication unit 130 are electrically connected to each other directly or indirectly to achieve data transmission or interaction. For example, these components can be electrically connected to each other through one or more communication buses or signal lines.

[0034] The memory 110 is used to store programs or data. The memory 110 may be, but is not limited to, random access memory (RAM), read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), etc.

[0035] The processor 120 is used to read / write data or programs stored in the memory 110 and execute corresponding functions. For example, the memory 110 stores a straight weld seam identification device 200, which includes at least one software function module that can be stored in the memory 110 in the form of software or firmware. The processor 120 executes various functional applications and data processing by running the software programs and modules stored in the memory 110, such as the straight weld seam identification device 200 in this embodiment, thereby implementing the straight weld seam identification method in this embodiment.

[0036] The communication unit 130 is used to establish a communication connection between the computer device 100 and other communication terminals through the network, and to send and receive data through the network.

[0037] It should be understood that, Figure 1 The structure shown is only a schematic diagram of the computer device 100. The computer device 100 may also include components that are larger than... Figure 1 The more or fewer components shown, or having the same Figure 1 The different configurations shown. Figure 1 The components shown can be implemented using hardware, software, or a combination thereof.

[0038] Please refer to Figure 2 , Figure 2 This is a flowchart illustrating a method for identifying straight weld seams provided in an embodiment of this application. The method can be applied to the aforementioned computer equipment. The specific flow of the method for identifying straight weld seams is described in detail below. In this embodiment, the method may include steps S100 to S300.

[0039] Step S100: Calculate the first intersection line based on the plane equations of the first initial plane and the second initial plane obtained by plane detection of the point cloud to be identified.

[0040] In this embodiment, the point cloud to be identified is the point cloud corresponding to the scenario where straight weld seam identification is required. For example, if it is necessary to identify straight weld seams on a workpiece for automatic welding, the point cloud of that workpiece can be collected as the point cloud to be identified. It is understood that the points in the point cloud to be identified are three-dimensional points.

[0041] The plane equations of the first initial plane and the second initial plane can be obtained. Then, based on these two plane equations, the equation of the first intersection line between the first initial plane and the second initial plane can be calculated. Thus, the first intersection line can be determined. It can be understood that the first intersection line is the plane intersection line between the first initial plane and the second initial plane, and this first intersection line is a straight line.

[0042] The equations of the first and second initial planes can be obtained by other devices performing plane detection on the point cloud to be identified and then sending it to the computer device, or they can be obtained by the computer device performing plane detection on the point cloud to be identified. That is, the first and second initial planes are planes obtained by performing plane detection on the point cloud to be identified. The plane detection method used can be, but is not limited to, the Random Hough Transform (RHT) algorithm, the Random Sample Consensus (RANSAC) algorithm, the Deterministic Voting algorithm, etc., and can be determined according to actual needs; no specific limitation is made here. The first and second initial planes are two intersecting planes determined based on the plane's normal vector. More than two planes may be detected from the point cloud to be identified. When more than two planes are detected, the specific method for determining the first and second initial planes from the more than two planes can be set according to actual needs.

[0043] Step S200: Based on the distribution of points in the first point cloud corresponding to the first initial plane and the second point cloud corresponding to the second initial plane in the point cloud to be identified, a line segment is determined from the first intersection line as the initial weld.

[0044] Since the first intersection line is an infinitely long straight line, and welding requires start and end points, a line segment can be determined from this first intersection line as the initial weld. To reduce misidentification of straight welds, a line segment can be determined from the first intersection line as the initial weld based on the distribution of points in the first point cloud and the distribution of points in the second point cloud. Specifically, the first point cloud is the point cloud in the point cloud to be identified that corresponds to the first initial plane; the second point cloud is the point cloud in the point cloud to be identified that corresponds to the second initial plane.

[0045] Determining the initial weld seam based on the distribution of points in the first and second point clouds can minimize the risk of including areas on the first intersection line that do not correspond to the first and second point clouds as part of the initial weld seam due to neglecting the distribution of the first and second point clouds.

[0046] It is worth noting that the number of initial welds that may be determined on the first intersection line could be 0, 1, or multiple, depending on the actual situation.

[0047] The inventors of this application have discovered that when determining the intersection line of planes obtained through a plane detection algorithm, and then identifying a line segment from the intersection line as the initial weld, the determined initial weld will deviate from the actual straight weld, resulting in a phenomenon where the initial weld does not completely coincide with the actual weld. For example... Figure 3 As shown, line segment 1 represents the initial weld seam obtained from the detected plane, and line segment 2 represents the actual weld seam. It can be seen that the left end of line segment 1 is offset downwards compared to the left end of line segment 2; and the right end of line segment 1 is offset upwards compared to the right end of line segment 2 (i.e., the actual weld seam). This occurs because point clouds generally contain some noise. When the actual object plane (e.g., the actual workpiece) constituting the straight weld seam has a large area, this object plane cannot actually be represented by a single plane equation. The plane obtained by the plane detection algorithm from the point cloud to be detected is only an average plane of the actual object plane. The average plane does not completely coincide with the actual object plane at the weld seam location. Therefore, the initial weld seam determined directly on the intersection line of the average plane will not match the actual weld seam.

[0048] Step S300: For each initial endpoint of each initial weld, adjust the initial endpoint according to the local plane in the local point cloud corresponding to the initial endpoint in the point cloud to be identified, so as to obtain the target endpoint corresponding to the initial endpoint.

[0049] To mitigate the misalignment between the initial weld and the actual weld, once the initial welds are identified, local point adjustments can be made to each identified initial weld to reduce the misalignment. In this embodiment, for each initial weld, for each of its two initial endpoints, adjustments are made based on the local plane in the local point cloud corresponding to that initial endpoint in the point cloud to be identified. The adjusted point is then used as the target endpoint corresponding to that initial endpoint. This yields two target endpoints for an initial weld, which are the endpoints of a target weld, which is a straight weld. Thus, by locally adjusting the two endpoints of the initial welds determined based on the intersection of planes, a more accurate straight weld position is obtained. The two target endpoints of a target weld are its start and end points.

[0050] The aforementioned initial endpoints refer to the endpoints of the initial weld. The two initial endpoints of an initial weld are the start and end points of that initial weld. The local point cloud corresponding to an initial endpoint is the point cloud located near the initial endpoint in the point cloud to be identified. The local point cloud corresponding to an initial endpoint includes the initial endpoint. The specific determination method can be determined according to actual needs.

[0051] As one possible implementation, a first detection result can be obtained by performing plane detection on the point cloud to be identified. This first detection result includes planes detected from the point cloud. The planes in the first detection result can be combined, and based on the normal vectors of the planes in each combination, it can be determined whether two planes in each combination intersect. If intersection is determined, the two planes in the combination are used as the first initial plane and the second initial plane in step S100 to obtain the corresponding initial weld. Then, local point adjustments are made to the two endpoints of the initial weld to obtain the target weld corresponding to the initial weld. Through local point adjustments, the initial weld that has shifted can be corrected, resulting in a target weld with a high degree of conformity to the actual weld. In this way, straight welds (i.e., target welds) in the point cloud to be identified can be detected.

[0052] For example, variable i is used to iterate through the planes in the first detection result, where the value of i is 1 ≤ i < the number of planes in the first detection result - 1; variable j is used to iterate through the planes in the first detection result, where the value of j is 1 + i ≤ j < the number of planes in the first detection result. For all combinations of i and j, the plane intersection line is calculated, and then the initial weld is calculated on the obtained plane intersection line.

[0053] In cases where it is necessary to identify all target welds in the point cloud to be identified, after the initial weld determination process is completed based on all plane combinations, all the determined initial welds can be traversed, and the target welds can be obtained by adjusting the local points of the traversed initial welds; alternatively, after obtaining the initial welds corresponding to a plane combination, the local points of the initial welds can be adjusted, and then the next plane combination can be processed. The specific execution order can be determined according to actual needs and is not specifically limited here.

[0054] As one possible implementation method, it can be achieved through Figure 4 The initial weld seam is determined from a first intersection line as shown. Please refer to... Figure 4 , Figure 4 for Figure 2 A flowchart illustrating the sub-steps included in step S200. In this embodiment, step S200 may include sub-steps S210 to S230.

[0055] Sub-step S210: For the first point cloud, all points in the first point cloud located near the first intersection line are projected onto the first intersection line, and the obtained first projection points are clustered without knowing the number of clusters, to obtain at least one first projection point set.

[0056] Sub-step S220: For the second point cloud, all points in the second point cloud located near the first intersection line are projected onto the first intersection line, and the obtained second projection points are clustered without knowing the number of clusters, to obtain at least one second projection point set.

[0057] Sub-step S230: Analyze whether there is an intersection between the first line segment interval corresponding to each first projection point set and the second line segment interval corresponding to each second projection point set. If there is an intersection, determine the initial weld seam based on the existing intersection.

[0058] In this embodiment, for each point in the first point cloud, a third distance between that point and the first intersection line can be calculated, and it can be determined whether the third distance is less than a second preset distance. If it is less, the point is considered a point in the first point cloud located near the first intersection line; if it is not less, the point is not considered a point in the first point cloud located near the first intersection line. For each point in the first point cloud located near the first intersection line, the point is projected onto the first intersection line to obtain a first projection point.

[0059] Similarly, for each point in the second point cloud, a fourth distance between that point and the first intersection line can be calculated, and it can be determined whether the fourth distance is less than a second preset distance. If it is less, the point is considered a point in the second point cloud located near the first intersection line; if it is not less, the point is not considered a point in the second point cloud located near the first intersection line. For each point in the second point cloud located near the first intersection line, the point is projected onto the first intersection line to obtain a second projection point.

[0060] One possible implementation is to directly cluster all the obtained first projection points and all the second projection points separately, even when the number of clusters is uncertain. This yields clustering results for all first projection points and all second projection points. Each cluster in the clustering results for all first projection points corresponds to a set of first projection points, and each cluster in the clustering results for all second projection points corresponds to a set of second projection points. The clustering method used can be, but is not limited to, Euclidean clustering. The Euclidean clustering algorithm does not require pre-specifying the number of clusters.

[0061] As another possible implementation, we can first count the first number of all first projection points and the second number of all second projection points. Then, we determine whether both the first number and the second number are greater than a preset number. If so, we can perform cluster analysis to obtain at least one set of first projection points and at least one set of second projection points; if not, we can directly determine that the initial weld seam determined from the first intersection line is 0, that is, no initial weld seam can be determined from the first intersection line. In this case, we do not need to execute step S300. The preset number can be 0 or greater than 0, which can be determined according to the actual situation. In this way, unnecessary calculations can be reduced.

[0062] Since both the first and second projection points are located on the first intersection line, a corresponding first line segment interval can be determined based on each first projection point set, and a corresponding second line segment interval can be determined based on each second projection point set. The first line segment interval corresponding to a projection point set is determined by that projection point set. For example, if a projection point set includes points 1, 2, 3, and 4, with points 1-4 located on the first intersection line and points 2-4 located between points 1 and 4, then points 1 and 4 can be taken as the endpoints of the corresponding line segment interval. This line segment interval represents the line segment on the first intersection line whose endpoints are points 1 and 4.

[0063] Having determined the intervals of each first line segment and each interval of each second line segment, for each first line segment interval, it can be determined whether there is an intersection between it and each second line segment interval, and if an intersection is found, the intersection can be analyzed. It can be understood that the intersection represents the overlapping portion of a first line segment interval and a second line segment interval. For each intersection determined, it can be determined whether there is a corresponding initial weld, and sometimes the corresponding initial weld is determined; alternatively, if all intersections are determined, the determined intersections can be traversed to determine whether each traversed intersection has a corresponding initial weld, and sometimes the corresponding initial weld is determined.

[0064] Alternatively, each intersection can be directly used as the corresponding initial weld.

[0065] Alternatively, it can also be done through Figure 5 The method shown determines the initial weld seam based on intersection. Please refer to... Figure 5 , Figure 5 for Figure 4 A flowchart illustrating the sub-steps included in neutron step S230. In this embodiment, sub-step S230 may include sub-steps S231 to S234.

[0066] Sub-step S231: Determine the nearest point corresponding to each of the two intersection endpoints of the intersection from the first point cloud and the second point cloud, and calculate the second distance between each intersection endpoint and the corresponding nearest point.

[0067] Sub-step S232: Determine whether all the obtained second distances are less than or equal to the first preset distance.

[0068] If the calculated second distances are all less than or equal to the first preset distance, then execute sub-step S233.

[0069] Sub-step S233: The line segment indicated by the intersection is used as an initial weld.

[0070] If at least one second distance is greater than the first preset distance, then sub-step S234 is executed.

[0071] Sub-step S234: Do not use the line segment indicated by the intersection as an initial weld.

[0072] In this embodiment, an intersection represents a line segment, and the endpoints of the intersection can be called intersection endpoints. For an intersection, for each intersection endpoint, the nearest point of the intersection endpoint can be determined from the first point cloud, and the distance between the intersection endpoint and the nearest point can be calculated as a second distance; and the nearest point of the intersection endpoint can be determined from the second point cloud, and the distance between the intersection endpoint and the nearest point can be calculated as a second distance. In this way, four corresponding nearest points can be determined based on two intersection endpoints of an intersection.

[0073] Determine whether the four second distances corresponding to the four nearest points are all less than or equal to the first preset distance. If so, the line segment represented by the intersection can be used as an initial weld; otherwise, the line segment represented by the intersection is not used as an initial weld.

[0074] If at least one initial weld is determined through step S200, the at least one initial weld can be traversed to perform local point adjustments on each initial weld to obtain the corresponding target weld. As one possible implementation, this can be achieved through... Figure 6 The method shown is for local point adjustment. Please refer to... Figure 6 , Figure 6 for Figure 2 A flowchart illustrating the sub-steps included in step S300. In this embodiment, step S300 may include sub-steps S310 to S350.

[0075] Sub-step S310: For the initial endpoint, determine the local point cloud corresponding to the initial endpoint.

[0076] In this embodiment, a portion of the point cloud located near the initial endpoint can be determined from the point cloud to be identified as the local point cloud corresponding to the initial endpoint. Optionally, a bounding box can be set, and the corresponding local point cloud can be determined from the point cloud to be identified based on the bounding box.

[0077] As one possible implementation method, it can be achieved through Figure 7 The method shown determines the local point cloud corresponding to an initial endpoint. Please refer to... Figure 7 , Figure 7 for Figure 6 A flowchart illustrating the sub-steps included in neutron step S310. In this embodiment, sub-step S310 may include sub-steps S311 to S312.

[0078] Sub-step S311: Determine a cylinder based on the initial endpoint and the initial weld where the initial endpoint is located.

[0079] Sub-step S312: Using the determined cylinder, determine the local point cloud corresponding to the initial endpoint from the point cloud to be identified.

[0080] In this embodiment, the axis of the cylinder can be determined based on the initial endpoint and the initial weld where the initial endpoint is located, and combined with a preset radius, a cylindrical bounding box can be determined. The initial endpoint is located on the axis of the cylinder, and the axis of the cylinder is located on the corresponding first intersection line. After determining the cylinder, the point cloud within the cylinder in the point cloud to be identified can be used as the local point cloud corresponding to the initial endpoint.

[0081] As one possible implementation, the cylinder has only one initial endpoint on its axis. That is, the local point clouds corresponding to the two initial endpoints of an initial weld are different. This can prevent the poor adjustment effect caused by the local plane at the starting point and the local plane at the ending point not being completely consistent.

[0082] As another possible implementation, in order to speed up the adjustment, the cylinder has two initial endpoints on its axis, that is, the local point clouds corresponding to the two initial endpoints of an initial weld are the same. In this way, the two initial endpoints of the initial weld can be adjusted using the local point cloud when a local point cloud is determined.

[0083] As another possible implementation, in order to ensure the adjustment effect and adjustment speed, it can be determined whether to use a local point cloud or two different local point clouds when adjusting the two initial endpoints of the initial weld, based on the length of the current initial weld.

[0084] In this implementation, the weld length of the initial weld is first obtained, and then it is determined whether the weld length is greater than a preset length. The preset length can be set according to actual needs. If the weld length is greater than the preset length, it indicates that the initial weld is relatively long, and the local planes at the two initial endpoints of the initial weld may not be completely consistent. In this case, for each of the two initial endpoints of the initial weld, a point can be determined on the initial weld as a target point. Then, based on a preset radius, the initial endpoint, and the target point corresponding to the initial endpoint, a cylinder corresponding to the initial endpoint is determined. Based on the cylinder corresponding to the initial endpoint, the local point cloud corresponding to the initial endpoint is determined from the point cloud to be identified. The target point is located between the targeted initial endpoint and the other initial endpoint of the initial weld. In this way, when the weld length is long, the local point clouds corresponding to each of the two initial endpoints can be determined, and the corresponding initial endpoints can be adjusted based on the local planes in the local point clouds corresponding to each initial endpoint.

[0085] If the weld length is less than or equal to the preset length, it indicates that the initial weld length is not very long, and the local planes at the two initial endpoints of the initial weld may be consistent. In this case, a corresponding cylinder can be determined based on the preset radius and the two initial endpoints of the initial weld. Then, based on this cylinder, the same local point cloud corresponding to the two initial endpoints of the initial weld in the point cloud to be identified can be used to adjust the two initial endpoints of an initial weld when the weld length is short.

[0086] Optionally, if two distinct local point clouds need to be determined for an initial weld, one initial endpoint can be used as the center of one base of the corresponding cylinder, and the target point corresponding to that initial endpoint can be used as the center of the other base of the corresponding cylinder. Similarly, if only one local point cloud needs to be determined for an initial weld, one initial endpoint of the initial weld can be used as the center of one base of the corresponding cylinder, and the other initial endpoint of the initial weld can be used as the center of the other base of the corresponding cylinder.

[0087] Optionally, two reference points can be set: p_in[0] and p_in[1]. One reference point represents an initial endpoint of an initial weld, and the other reference point represents another initial weld or target point of the initial weld. The reference point representing the initial endpoint can be adjusted a certain distance away from the other initial endpoint of the initial weld on the first intersection line, and the adjusted point is used as the center of one bottom surface of the corresponding cylinder, and the other unadjusted reference point is used as the center of the other bottom surface of the corresponding cylinder.

[0088] Alternatively, it can also be as follows Figure 8 As shown, the two reference points p_in[0] and p_in[1] are extended outward along the cylinder axis by a certain distance seg_l. The points obtained after the extension are then used as the centers of the upper and lower bottom surfaces of the cylinder, and combined with the preset radius seg_r, the cylinder is determined.

[0089] Sub-step S320: Perform planar detection on the local point cloud to obtain a second detection result.

[0090] Sub-step S330: Determine two local planes as target planes from the second detection result.

[0091] In this embodiment, upon obtaining a local point cloud, planar detection can be performed on the local point cloud, and the obtained result is used as a second detection result. The second detection result includes local planes, which are planes obtained from the planar detection of the local point cloud. Two local planes can be determined from the second detection result as target planes, i.e., two target planes are determined. One of the two target planes is similar to the first initial plane, and the other plane is similar to the second initial plane.

[0092] As one possible implementation method, it can be achieved through Figure 9 The two target planes are selected using the aforementioned method. Please refer to... Figure 9 , Figure 9 for Figure 6 A flowchart illustrating the sub-steps included in neutron step S330. In this embodiment, sub-step S330 may include sub-steps S331 to S333.

[0093] Sub-step S331: Determine whether the number of local planes in the second detection result is greater than 2.

[0094] In this embodiment, the second detection result is obtained based on the local point cloud where the initial endpoint is located. The initial endpoint is a point near the actual intersection of two planes. Therefore, the second detection result includes at least two local planes. The number of local planes in the second detection result can be determined first. Then, it is determined whether the number of planes is greater than 2.

[0095] If the number of planes is greater than 2, sub-step S332 can be executed.

[0096] Sub-step S332: Calculate the first similarity between each local plane and the first initial plane based on the plane information of each local plane in the second detection result and the plane information of the first initial plane, and select the local plane corresponding to the largest first similarity as the selected first target plane; and calculate the second similarity between each local plane and the second initial plane based on the plane information of each local plane and the plane information of the second initial plane, and select the local plane corresponding to the largest second similarity as the selected second target plane.

[0097] In this embodiment, when there are more than two local planes in the second detection result, the planar information of the local planes, the planar information of the first initial plane, and the planar information of the second initial plane can be combined to calculate the first similarity between each local plane and the first initial plane, and the second similarity between each local plane and the second initial plane. Then, the local plane corresponding to the largest first similarity is selected as the plane most similar to the first initial plane, i.e., the local plane corresponding to the largest first similarity is selected as the first target plane. Similarly, the local plane corresponding to the largest second similarity is selected as the plane most similar to the second initial plane, i.e., the local plane corresponding to the largest second similarity is selected as the second target plane. The aforementioned first target plane and second target plane are the two target planes determined from the second detection result.

[0098] The planar information described herein is used to describe the properties of the plane, and any information sufficient for calculating the similarity between planes is acceptable. It is understood that the plane described by the planar information is an infinitely large plane.

[0099] As one possible implementation, the plane information includes a normal vector n and a distance value ρ, where ρ represents the distance of the plane from the origin of the coordinate system. The distance value has a positive or negative sign, and the plane equation is: ax + by + cz + d = 0, where a, b, and c describe the normal vector n of the plane, and d = -ρ. The similarity between two planes is represented by a first distance; the greater the similarity, the smaller the first distance. The first distance can be determined by... Figure 10 The calculation was performed as shown. Please refer to... Figure 10 , Figure 10 for Figure 9 A flowchart illustrating the sub-steps included in neutron step S332. In this embodiment, sub-step S332 may include sub-steps S3321 to S3323.

[0100] Sub-step S3321 calculates the inner product of the first normal vector and the second normal vector.

[0101] Sub-step S3322: Obtain the interval distance value based on the first distance value corresponding to the first plane and the second distance value corresponding to the second plane.

[0102] Sub-step S3323: Calculate the product of the reciprocal of the inner product value and the interval distance value, and obtain the first distance based on the absolute value of the product.

[0103] In this embodiment, the two planes whose similarity needs to be calculated include a first plane and a second plane. The normal vector of the first plane is a first normal vector, and the distance value of the first plane is a first distance value. The normal vector of the second plane is a second normal vector, and the distance value of the second plane is a second plane value. The inner product of the first normal vector and the second normal vector can be calculated. Furthermore, the interval distance value is calculated based on the first distance value and the second distance value. The interval distance value describes the interval between the first distance value and the second distance value; that is, the interval distance value represents the similarity ρ between the first plane and the second plane. Then, the product of the reciprocal of the inner product value and the interval distance value can be calculated, and the first distance is obtained based on the absolute value of the product. Optionally, the absolute value of the product can be understood as the first distance. Thus, the first distance can be calculated by comprehensively considering the similarity between the normal vectors of the planes and the similarity between the distance values ​​ρ.

[0104] Specifically, it can be determined whether the inner product value is less than 0. If it is less than 0, the opposite of the first distance value is subtracted from the second distance value, and the result is used as the interval distance value. If it is not less than 0, the first distance value is subtracted from the second distance value, and the result is used as the spacing distance value.

[0105] The following explains how to use it. Figure 10 The method shown is used to select two target planes from two or more local planes for illustration.

[0106] Assuming the initial weld seam being targeted is line_i, two planes for line_i are calculated: the first initial plane and the second initial plane. The normal of the first initial plane is n1, and the parameter ρ is ρ1. The normal of the second initial plane is n2, and the parameter ρ2 is ρ2. Two values ​​are initialized: min_dis_1 = 1e10, min_dis_2 = 1e10. min_dis_1 represents the shortest distance from the detected local plane to the first initial plane, and min_dis_2 represents the shortest distance from the detected local plane to the second initial plane.

[0107] By iterating through all local planes in the second detection result using variable i, the normal n of local plane i can be obtained. i and distance parameter ρ i Calculate n i ·n1, if the above inner product is less than 0, then let ρ i,1 =-ρ i Otherwise let ρ i,1 =ρ iThen, the distance from local plane i to the first initial plane (i.e., the first distance mentioned above) is calculated using the formula: |(ρi,1-ρ1) / (ni·n1)|. This distance formula takes into account both the similarity between the normals of the planes and the similarity between ρ. The closer the normals and ρ of two planes are, the closer the distance between them is. If the distance is less than min_dis_1, then min_dis_1 is set to equal the distance, and the label of the local plane is recorded. This label is used to obtain the plane equation corresponding to the local plane later, which facilitates the calculation of the second intersection line. Conversely, if the distance is greater than or equal to min_dis_1, then min_dis_1 is not updated.

[0108] Similarly, calculate n i ·n2, if the above inner product is less than 0, then let ρ i,2 =-ρ i Otherwise let ρ i,2 =ρ i Then, the distance from local plane i to the second initial plane (i.e., the first distance mentioned above) is calculated using the formula: |(ρi,2-ρ2) / (ni·n2)|. If the distance is less than min_dis_2, then min_dis_2 is set to equal the distance, and the label of the local plane is recorded. This label is used to obtain the plane equation corresponding to the local plane later, which facilitates the calculation of the second intersection line. Conversely, if the distance is greater than or equal to min_dis_2, then min_dis_2 is not updated.

[0109] In this way, the local plane closest to the first initial plane and the local plane closest to the second initial plane can be determined from the local plane of the second detection result.

[0110] If the number of local planes in the second detection result is no greater than 2, that is, if the number of planes is 2, then sub-step S333 can be executed.

[0111] Sub-step S332: The two local planes in the second detection result are directly used as the first target plane and the second target plane.

[0112] If two local planes are detected, the two detected local planes can be directly used as the first target plane and the second plane, without the need for the above similarity calculation.

[0113] Sub-step S340 calculates the second intersection line of the two target planes.

[0114] Sub-step S350: Project the initial endpoint onto the second intersection line, and take the obtained third projection point as the target endpoint corresponding to the initial endpoint.

[0115] After identifying two target planes, the equation of the second line of intersection of the two target planes can be calculated based on their plane equations. This second line of intersection is an infinitely long straight line. With the second line of intersection obtained, the initial endpoint corresponding to the local point cloud can be vertically projected onto the first line of intersection, resulting in a third projection point. This third projection point can be used as the target endpoint corresponding to the initial endpoint.

[0116] Based on the aforementioned descriptions of determining a local point cloud from two cylinders based on an initial weld seam, or determining a local point cloud from a single cylinder, it can be seen that, correspondingly, when determining two local point clouds based on two initial endpoints, the two initial endpoints are projected onto the second intersection line determined by the corresponding local point clouds to obtain the target endpoints corresponding to the two initial endpoints, thereby obtaining a target weld seam. This method requires calculating two second intersection lines and performing two projections. When determining a local point cloud based on two initial endpoints, both initial endpoints are projected onto the second intersection line determined based on that single local point cloud to obtain the target endpoints corresponding to the two initial endpoints. This method only requires calculating one second intersection line and performing one projection.

[0117] The following example illustrates how to make local point adjustments.

[0118] Iterate through all initial weld seams using the variable `i` to obtain the initial weld seam `line_i`. The start point of `line_i` is `line_i.p1_`, and the end point is `line_i.p2_`. The unit direction vector is `line_i.t_`, pointing from `line_i.p1_` to `line_i.p2_`. Calculate the length of `line_i` and determine if it exceeds a preset length. If it does, it indicates that the initial weld seam `line_i` has a relatively long length, and local point adjustments are made to the start and end points of `line_i`. If not, it indicates that the initial weld seam `line_i` has a relatively short length, and local point adjustments are made to both the start and end points.

[0119] The following section first introduces how to adjust the starting and ending points of the initial weld line_i locally when the length of the initial weld line_i is greater than the preset length.

[0120] Two reference points are set: p_in[0] = line_i.p1_, p_in[1] = line_i.p1_ + 100·line_i.t_, where p_in[0] is the starting point of the initial weld line_i; p_in[1] is the point that moves 100 units from the starting point of the initial weld line_i to the ending point of the initial weld line_i, which is the aforementioned target point. Based on the positions of the two points in p_in, a cylinder is constructed by combining the values ​​of seg_r and seg_l. The cylinder is used to extract the local point cloud from the point cloud to be identified. The point cloud inside the cylinder is the extracted local point cloud, which is the local point cloud corresponding to line_i.p1_. The extraction method is to use p_in[0] and p_in[1] to form the axis of the cylinder, seg_r as the radius of the cylinder, and seg_l as the distance extended outward along the axis of the cylinder. The cylinder is as follows: Figure 8 As shown.

[0121] Perform plane detection on the point cloud inside the cylinder (i.e., the local point cloud corresponding to the determined line_i.p1_). Then obtain the number of detected planes. If the number of detected planes is 2, no plane selection is needed, and these 2 planes are directly selected as the two target planes. If the number of detected planes is greater than 2, then perform the plane selection shown in sub-step S332 to select two planes as the two target planes.

[0122] After obtaining two target planes, p_in[0] is projected onto the intersection line formed by the two target planes to obtain the corresponding projection point p_out[0]. Let the adjusted line segment (i.e. the target weld) corresponding to line_i be line_i_ad, then the starting point of line_i_ad is line_i_ad.p1_=p_out[0]. The above process adjusts the local position of the starting point of line_i by detecting the plane in the local point cloud through the local point cloud of the starting point of line_i.

[0123] Next, the endpoint of the initial weld line_i is adjusted locally, in a similar manner.

[0124] Two reference points are set: p_in[0] = line_i.p2_-100·line_i.t_ and p_in[1] = line_i.p2_. p_in[1] is the endpoint of the initial weld line_i, and p_in[0] is the point 100 units away from the endpoint of the initial weld line_i towards its starting point. Based on the positions of the two points in p_in, a cylinder is constructed using the values ​​of seg_r and seg_l. This cylinder is then used to extract the local point cloud. The method for constructing the cylinder is the same as before.

[0125] Perform plane detection on the point cloud inside the cylinder (i.e., the local point cloud corresponding to the determined line_i.p2_). Obtain the number of detected planes. If the number of detected planes is 2, directly select these 2 planes as the two target planes. If the number of detected planes is greater than 2, then perform the plane selection shown in sub-step S332 to select two target planes.

[0126] Project p_in[1] onto the intersection line formed by the two target planes to obtain the corresponding projection point p_out[1]. The endpoint of line_i_ad is line_i_ad.p2_=p_out[1]. The above process adjusts the local point position of the endpoint of line_i by detecting the plane in the local point cloud through the local point cloud. In this way, the starting and ending positions of the initial weld line_i are adjusted.

[0127] The following section describes how to simultaneously adjust the start and end points of the initial weld line_i when the length of the initial weld line_i is less than or equal to the preset length.

[0128] The process of simultaneously adjusting the local position of the starting point and ending point of the initial weld line_i is the same as the above process type, except that the starting point and ending point of the initial weld line_i are adjusted locally at the same time. Two reference points are set, p_in[0] = line_i.p1_, p_in[1] = line_i.p2_, and the starting point and ending point of the initial weld line_i are directly used as reference points. Local point cloud is obtained by using p_in, seg_r, and seg_l. Then, plane detection is performed on the local point cloud, and two target planes are determined from the detected planes. After that, p_in (i.e., p_in[0], p_in[1]) is projected onto the intersection line of the two target planes to obtain p_out, then the starting point of the adjusted initial weld (i.e. the target weld) is line_i_ad.p1_ = p_out[0], and the ending point of the adjusted initial weld (i.e. the target weld) is line_i_ad.p2_ = p_out[1].

[0129] By adjusting the starting and ending points of the initial weld using the local point adjustment method provided in this embodiment, more precise starting and ending point positions can be obtained. Because a local point cloud near the starting and ending points is used, and a plane is detected within this local point cloud, this plane is closer to the actual object plane constituting the weld. Projecting the starting and ending points onto the intersection line formed by the local planes yields accurate projection points. In this embodiment, the weld length is also considered during local point adjustment. When the weld length is long, local point adjustments are performed separately for the starting and ending points to prevent inconsistencies between the local planes at the starting and ending points. When the weld length is short, local point adjustments are performed simultaneously for both the starting and ending points.

[0130] To perform the corresponding steps in the above embodiments and various possible methods, an implementation of the straight weld seam identification device 200 is given below. Optionally, the straight weld seam identification device 200 can adopt the above-described... Figure 1 The diagram shows the device structure of a computer. Further, please refer to... Figure 11 , Figure 11 This is a block diagram of the straight weld seam recognition device 200 provided in this embodiment. It should be noted that the basic principle and technical effects of the straight weld seam recognition device 200 provided in this embodiment are the same as those in the above embodiments. For the sake of brevity, any parts not mentioned in this embodiment can be referred to the corresponding content in the above embodiments. In this embodiment, the straight weld seam recognition device 200 may include: a calculation module 210, a detection module 220, and a local point adjustment module 230.

[0131] The calculation module 210 is used to calculate the first intersection line based on the plane equations of the first initial plane and the second initial plane obtained by plane detection of the point cloud to be identified.

[0132] The detection module 220 is used to determine a line segment as an initial weld seam from the first intersection line based on the distribution of points in the first point cloud corresponding to the first initial plane and the second point cloud corresponding to the second initial plane in the point cloud to be identified.

[0133] The local point adjustment module 230 is used to adjust each initial endpoint of each initial weld seam according to a local plane in the local point cloud corresponding to the initial endpoint in the point cloud to be identified, so as to obtain the target endpoint corresponding to the initial endpoint. The two target endpoints corresponding to one initial weld seam are the two endpoints of one target weld seam, and the target weld seam is a straight weld seam.

[0134] Optionally, the above modules can be stored in the form of software or firmware. Figure 1The memory 110 shown is either stored in or embedded in the operating system (OS) of the computer device 100, and can be used by... Figure 1 The processor 120 executes the program. Meanwhile, the data and program code required to execute the above modules can be stored in the memory 110.

[0135] This application also provides a readable storage medium storing a computer program thereon, which, when executed by a processor, implements the straight weld seam identification method described above.

[0136] In summary, this application provides a method, apparatus, computer device, and readable storage medium for identifying straight weld seams. First, a first intersection line is calculated based on the equations of two initial planes obtained through planar detection of the point cloud to be identified. Then, based on the distribution of points in the point cloud corresponding to each of the two initial planes, a line segment is determined from the first intersection line as an initial weld seam. Finally, for each initial weld seam, for each of its two initial endpoints, the initial endpoint is adjusted according to a local plane in the local point cloud corresponding to that initial endpoint in the point cloud to be identified, and the resulting point is taken as the target endpoint corresponding to that initial endpoint. Here, the two target endpoints corresponding to one initial weld seam are the two endpoints of a target weld seam, and the target weld seam is a straight weld seam. Thus, by adjusting the local point positions of the two endpoints of the initial weld seam determined based on the planar intersection line, a more accurate straight weld seam position is obtained, avoiding the situation where the initial weld seam is significantly offset from the actual straight weld seam when directly using the initial weld seam as the identified straight weld seam.

[0137] The above description is merely an embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for identifying straight weld seams, characterized in that, The method includes: The first intersection line is calculated based on the plane equations of the first and second initial planes obtained by plane detection of the point cloud to be identified. Based on the distribution of points in the first point cloud corresponding to the first initial plane and the second point cloud corresponding to the second initial plane in the point cloud to be identified, a line segment is determined from the first intersection line as the initial weld. For each initial endpoint of an initial weld, the initial endpoint is adjusted according to the local plane in the local point cloud corresponding to the initial endpoint in the point cloud to be identified, so as to obtain the target endpoint corresponding to the initial endpoint. The two target endpoints corresponding to an initial weld are the two endpoints of a target weld, and the target weld is a straight weld. Specifically, for each initial endpoint of an initial weld, adjusting the initial endpoint based on a local plane in the local point cloud corresponding to the initial endpoint in the point cloud to be identified, to obtain a target endpoint corresponding to the initial endpoint, includes: determining the local point cloud corresponding to the initial endpoint; performing plane detection on the local point cloud to obtain a second detection result; determining two local planes as target planes from the second detection result, wherein one of the two target planes is similar to the first initial plane and the other is similar to the second initial plane; calculating a second intersection line of the two target planes; projecting the initial endpoint onto the second intersection line, and using the obtained third projection point as the target endpoint corresponding to the initial endpoint.

2. The method according to claim 1, characterized in that, The step of determining two local planes as target planes from the second detection result includes: Determine whether the number of local planes in the second detection result is greater than 2; When the number of planes is greater than 2, the first similarity between each local plane and the first initial plane is calculated based on the plane information of each local plane in the second detection result and the plane information of the first initial plane, and the local plane corresponding to the largest first similarity is selected as the selected first target plane; and the second similarity between each local plane and the second initial plane is calculated based on the plane information of each local plane and the plane information of the second initial plane, and the local plane corresponding to the largest second similarity is selected as the selected second target plane.

3. The method according to claim 2, characterized in that, The planar information includes a normal vector and a distance value, wherein the distance value represents the distance of the plane from the origin of the coordinate system, and the distance value has a positive or negative sign; the similarity between two planes is represented by a first distance, the greater the similarity, the smaller the first distance, and the first distance is calculated in the following way: The inner product of the first normal vector and the second normal vector is calculated, wherein the first normal vector is the normal vector of the first plane among the two planes, and the second normal vector is the normal vector of the second plane among the two planes; An interval distance value is obtained based on the first distance value corresponding to the first plane and the second distance value corresponding to the second plane, wherein the interval distance value is used to describe the interval between the first distance value and the second distance value; The product of the reciprocal of the inner product value and the interval distance value is calculated, and the first distance is obtained based on the absolute value of the product.

4. The method according to claim 2, characterized in that, The step of determining two local planes as target planes from the second detection result further includes: When the number of planes is 2, the two local planes in the second detection result are directly used as the first target plane and the second target plane.

5. The method according to claim 1, characterized in that, The step of determining the local point cloud corresponding to the initial endpoint includes: Based on the initial endpoint and the initial weld where the initial endpoint is located, a cylinder is determined, wherein the initial endpoint is located on the axis of the cylinder, and the axis of the cylinder is located on the corresponding first intersection line; The local point cloud corresponding to the initial endpoint is determined from the point cloud to be identified using the determined cylinder.

6. The method according to claim 5, characterized in that, The step of determining a cylinder based on the initial endpoint and the initial weld where the initial endpoint is located includes: When the length of the initial weld is greater than the preset length, a point is determined on the initial weld as a target point for the initial endpoint, and a corresponding cylinder is determined according to the preset radius, the initial endpoint and the target point. The target point is located between the initial endpoint and the other initial endpoint of the initial weld, and the local point clouds corresponding to the two initial endpoints of the initial weld are different. When the weld length is less than or equal to the preset length, a corresponding cylinder is determined based on the preset radius and the two initial endpoints of the initial weld. The two initial endpoints of the initial weld correspond to the same local point cloud, and the two target endpoints of the initial weld are obtained by projecting the two initial endpoints of the initial weld onto the same second intersection line.

7. The method according to any one of claims 1-6, characterized in that, The step of determining a line segment as the initial weld from the first intersection line based on the distribution of points in the first point cloud corresponding to the first initial plane and the second point cloud corresponding to the second initial plane in the point cloud to be identified includes: For the first point cloud, all points in the first point cloud located near the first intersection line are projected onto the first intersection line, and the obtained first projection points are clustered without knowing the number of clusters, to obtain at least one first projection point set. For the second point cloud, all points in the second point cloud located near the first intersection line are projected onto the first intersection line, and the obtained second projection points are clustered without knowing the number of clusters, to obtain at least one second projection point set; Analyze whether there is an intersection between the first line segment interval corresponding to each first projection point set and the second line segment interval corresponding to each second projection point set. If there is an intersection, determine the initial weld seam based on the existing intersection.

8. The method according to claim 7, characterized in that, Determining the initial weld seam based on the existing intersection includes: The nearest points corresponding to the two intersection endpoints of the intersection are determined from the first point cloud and the second point cloud, and the second distance between each intersection endpoint and its corresponding nearest point is calculated. When the calculated second distances are all less than or equal to the first preset distance, the line segment indicated by the intersection is taken as an initial weld. If at least one second distance is greater than the first preset distance, the line segment indicated by the intersection is not considered as an initial weld.

9. A straight weld seam identification device, characterized in that, The device includes: The calculation module is used to calculate the first intersection line based on the plane equations of the first initial plane and the second initial plane obtained by plane detection of the point cloud to be identified; The detection module is used to determine a line segment as the initial weld from the first intersection line based on the distribution of points in the first point cloud corresponding to the first initial plane and the second point cloud corresponding to the second initial plane in the point cloud to be identified. The local point adjustment module is used to adjust each initial endpoint of each initial weld seam according to the local plane in the local point cloud corresponding to the initial endpoint in the point cloud to be identified, so as to obtain the target endpoint corresponding to the initial endpoint. The two target endpoints corresponding to an initial weld seam are the two endpoints of a target weld seam, and the target weld seam is a straight weld seam. Specifically, the local point adjustment module is used to: determine the local point cloud corresponding to the initial endpoint; perform planar detection on the local point cloud to obtain a second detection result; determine two local planes as target planes from the second detection result, wherein one of the two target planes is similar to the first initial plane and the other plane is similar to the second initial plane; calculate the second intersection line of the two target planes; project the initial endpoint onto the second intersection line, and use the obtained third projection point as the target endpoint corresponding to the initial endpoint.

Images