A manual weld seam recognition method and a computer device

Abstract

The application provides a manual weld joint identification method and computer equipment, and relates to the technical field of welding control. On the basis of determining the original plane set of the workpiece to be detected, each original plane included in the original plane set is traversed, and a target plane constituting a concave dihedral angle with the traversed original plane in the original plane set is searched. Then, a theoretical intersection line segment between the target plane and the traversed original plane is divided into a plurality of locally distributed local line segments, and the plurality of local line segments are subjected to line segment distribution clustering according to the minimum point cloud projection distances of the target plane and the original plane on different local line segments. The manual weld joint distribution detection is performed according to the line segment clustering results of the target plane and the original plane on the plurality of local line segments, whether the manual weld joint exists between the target plane and the original plane is determined, and thus the accurate positioning and identification of all manual weld joints substantially existing on the workpiece to be detected are performed.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of welding control, in particular to a manual weld seam identification method and a computer device. BACKGROUND

[0002] With the rapid development of welding automation technology, various industries (for example, automobile industry, electronic assembly, metallurgical chemical industry, etc.) have higher requirements for welding precision, welding quality and welding safety. As the core link of the automatic welding system, the positioning result of the weld seam identification and positioning directly affects the final welding quality and production efficiency.

[0003] However, in the actual use process of the weld seam identification and positioning technology, there are often manual welding traces (i.e. manual weld seams) on the workpiece to be detected, which need to be accurately identified so that the subsequent automatic welding operation can avoid the existing manual weld seams, or the identified manual weld seams can be welded and repaired to improve the overall welding quality. It is worth noting that the existing manual weld seam identification method is constructed on the basis of regarding the space region between planes as a whole, and its essence is suitable for identifying manual weld seams with continuous welding characteristics (i.e. continuous distribution of welding traces), which leads to the omission of manual weld seams with non-continuous welding characteristics (i.e. segmented and spaced distribution of welding traces) in the corresponding identification result, and cannot accurately position and identify all manual weld seams actually existing on the workpiece to be detected. SUMMARY

[0004] Therefore, the purpose of the present application is to provide a manual weld seam identification method and a computer device, which can realize high-precision manual weld seam identification function on the basis of the plane intersection line segment segmentation and clustering result (which includes the line segment clustering result of each plane in the two intersecting planes determined according to the point cloud projection distance distribution details of itself under the same intersection line segment division result), combined with the manual weld seam distribution detection operation, and can accurately position and identify all manual weld seams actually existing on any workpiece to be detected (whether it involves continuous welding characteristics or non-continuous welding characteristics).

[0005] In order to achieve the above purpose, the technical solutions adopted by the embodiments of the present application are as follows:

[0006] In a first aspect, the present application provides a manual weld seam identification method, which comprises:

[0007] Performing plane detection on the surface point cloud data of the workpiece to be detected to obtain an original plane set of the workpiece to be detected;

[0008] Traversing each original plane included in the original plane set to find a target plane in the original plane set which forms a concave dihedral angle with the traversed original plane;

[0009] For each target plane found, the theoretical intersection segment between the target plane and the traversed original plane is divided into a plurality of locally distributed partial segments, and the plurality of partial segments are clustered according to the minimum point cloud projection distances of the target plane and the original plane on different partial segments, to obtain the segment clustering results of the target plane and the original plane on the plurality of partial segments;

[0010] According to the segment clustering results of the target plane and the original plane, artificial weld distribution detection is performed to determine whether there is an artificial weld between the target plane and the original plane.

[0011] In an optional embodiment, for each to-be-clustered plane in the traversed original plane and any one of the target planes, the plurality of partial segments are clustered according to the minimum point cloud projection distances of the to-be-clustered plane on the plurality of partial segments, to obtain the segment clustering results of the to-be-clustered plane on the plurality of partial segments, including:

[0012] The minimum point cloud projection distance of the to-be-clustered plane on any one partial segment that is less than a first preset distance threshold is set to zero, wherein the first preset distance threshold is a positive number close to 0;

[0013] According to the relative arrangement relationship between the plurality of partial segments, a target segment group is extracted from the plurality of partial segments; wherein the minimum point cloud projection distances of all partial segments in any one target segment group are greater than 0, and all partial segments in the same target segment group are arranged in sequence according to the relative arrangement relationship;

[0014] For each target segment group extracted, the segment clustering is performed according to the minimum point cloud projection distances of all partial segments included in the target segment group, to obtain at least one original clustering cluster matched with the target segment group; wherein all partial segments belonging to the same original clustering cluster are arranged in sequence according to the relative arrangement relationship, and the distance difference absolute value between the minimum point cloud projection distances of adjacent two partial segments in the same original clustering cluster is less than or equal to a preset distance difference threshold;

[0015] The original clustering cluster with a corresponding segment number greater than a preset segment number threshold and an average projection distance greater than a second preset distance threshold is taken as a target clustering cluster in the segment clustering results of the to-be-clustered plane; wherein the second preset distance threshold is greater than the first preset distance threshold.

[0016] In an optional embodiment, the step of performing artificial weld seam distribution detection according to the line segment clustering results of the target plane and the original plane to determine whether there is an artificial weld seam between the target plane and the original plane comprises:

[0017] detecting whether there is a target cluster in the line segment clustering results of the target plane and the original plane respectively, which corresponds to an average projection distance greater than a second preset distance threshold;

[0018] In the case where it is detected that there is no at least one target cluster in the line segment clustering results of the target plane and / or the original plane respectively, it is determined that there is no artificial weld seam between the target plane and the original plane.

[0019] In an optional embodiment, the step of performing artificial weld seam distribution detection according to the line segment clustering results of the target plane and the original plane to determine whether there is an artificial weld seam between the target plane and the original plane further comprises:

[0020] In the case where it is detected that there is at least one target cluster in the line segment clustering results of the target plane and the original plane respectively, all target clusters involved in the target plane and the original plane are combined and paired to obtain at least one set of effective cluster pairs between the target plane and the original plane.

[0021] For each set of effective cluster pairs, an intermediate plane matching the set of effective cluster pairs is constructed; wherein a first plane intersection segment between the intermediate plane and the original plane is located in the original plane, an actual distance between the first plane intersection segment and a corresponding theoretical intersection segment is an average projection distance of an effective cluster corresponding to the original plane in the set of effective cluster pairs, a second plane intersection segment between the intermediate plane and the target plane is located in the target plane, and an actual distance between the second plane intersection segment and a corresponding theoretical intersection segment is an average projection distance of an effective cluster corresponding to the target plane in the set of effective cluster pairs.

[0022] For each intermediate plane, whether there is an artificial weld seam structure feature near the intermediate plane is detected according to target point clouds near the intermediate plane on each reference plane of the original plane set except the target plane and the original plane.

[0023] When it is detected that there is at least one artificial weld seam structure feature near at least one intermediate plane, it is determined that there is an artificial weld seam between the target plane and the original plane.

[0024] When it is detected that there is no artificial weld seam structure feature near all intermediate planes, it is determined that there is no artificial weld seam between the target plane and the original plane.

[0025] In an optional embodiment, the step of combining and pairing all target cluster groups involved by the target plane and the original plane respectively to obtain at least one set of effective cluster group pairs between the target plane and the original plane comprises:

[0026] detecting whether the total number of target cluster groups corresponding to the target plane and the original plane respectively remains consistent;

[0027] when the total number of target cluster groups corresponding to the target plane and the original plane respectively remains consistent, traversing all target cluster groups involved by the original plane, taking each traversed target cluster group as one effective cluster group in a single set of effective cluster group pairs, and taking target cluster groups involved by the target plane which maintain the same arrangement order with the traversed target cluster group as the other effective cluster group in the single set of effective cluster group pairs; wherein the arrangement order between all target cluster groups involved by the target plane and the original plane respectively matches the relative arrangement relationship between the plurality of local line segments.

[0028] In an optional embodiment, the step of combining and pairing all target cluster groups involved by the target plane and the original plane respectively to obtain at least one set of effective cluster group pairs between the target plane and the original plane further comprises:

[0029] when the total number of target cluster groups corresponding to the target plane and the original plane respectively does not remain consistent, determining a first plane with a smaller total number of target cluster groups and a second plane with a larger total number of target cluster groups in the target plane and the original plane;

[0030] traversing all first target cluster groups involved by the first plane, taking each traversed first target cluster group as one effective cluster group in a single set of effective cluster group pairs, and for each traversed first target cluster group, performing cluster group merging on a plurality of second target cluster groups involved by the second plane to take a corresponding merged cluster group as the other effective cluster group in the single set of effective cluster group pairs.

[0031] In an optional embodiment, for each traversed first target cluster group, the step of performing cluster group merging on a plurality of second target cluster groups involved by the second plane comprises:

[0032] constructing an initial cluster interval matching the traversed first target cluster group, wherein two interval boundary endpoints of the initial cluster interval remain consistent with two cluster group boundary endpoints of the traversed first target cluster group;

[0033] traversing all second target cluster groups involved by the second plane, and detecting whether each traversed second target cluster group is in a merged state;

[0034] If the second target cluster traversed is in the merged state, the next second target cluster is traversed, otherwise it is detected whether at least one cluster boundary endpoint of the second target cluster traversed is between the two interval boundary endpoints;

[0035] If neither of the two cluster boundary endpoints of the second target cluster traversed is between the two interval boundary endpoints, the next second target cluster is traversed, otherwise the second target cluster traversed is marked as in the merged state, and the target boundary endpoint of the two interval boundary endpoints close to the reference boundary endpoint of the second target cluster is updated according to the reference boundary endpoint of the second target cluster which is not between the two interval boundary endpoints;

[0036] After traversing all the second target clusters, the second target clusters covered by the initial clustering interval are clustered to obtain a merged cluster matching the first target cluster traversed.

[0037] In an optional embodiment, for each intermediate plane, the step of detecting whether there is a manual weld structure feature near the intermediate plane according to the target point cloud close to the intermediate plane on each reference plane in the original plane set except the target plane and the original plane, comprises:

[0038] A plurality of initial sampling points are generated uniformly in the intermediate plane, and the number of effective sampling points in the plurality of initial sampling points is counted based on all target point clouds close to the intermediate plane; wherein the average distance of each effective sampling point to a preset number of target discrete points adjacent to the effective sampling point in the target point cloud is less than a third preset distance threshold;

[0039] It is detected whether a first actual ratio between the total number of discrete points of all target point clouds and the maximum plane intersection segment length of the intermediate plane is greater than or equal to a first proportion threshold, and whether a second actual ratio between the number of effective sampling points and the total number of initial sampling points of the intermediate plane is greater than or equal to a second proportion threshold;

[0040] If it is detected that the first actual ratio is greater than or equal to the first proportion threshold, and the second actual ratio is greater than or equal to the second proportion threshold, it is determined that there is a manual weld structure feature near the intermediate plane, otherwise it is determined that there is no manual weld structure feature near the intermediate plane.

[0041] In an optional embodiment, the identification method further comprises:

[0042] In the process of traversing each original plane included in the original plane set, it is detected whether the original plane traversed belongs to a manual weld structure plane;

[0043] If the detected original plane belongs to the artificial weld structure plane, the next original plane is traversed, otherwise, the step of searching the target plane in the original plane set which forms a concave dihedral angle with the traversed original plane is performed, and whether the searched target plane belongs to the artificial weld structure plane is detected;

[0044] If the searched target plane belongs to the artificial weld structure plane, the step of searching the target plane in the original plane set which forms a concave dihedral angle with the traversed original plane is performed, otherwise, the step of dividing the theoretical intersection line segment between the target plane and the traversed original plane into a plurality of locally distributed local line segments is performed;

[0045] Wherein, for any one original plane in the original plane set, the step of detecting whether the original plane belongs to the artificial weld structure plane includes:

[0046] Counting whether the number of target discrete points on the original plane associated with the existing artificial weld structure features exceeds a preset discrete point number threshold;

[0047] If the number of target discrete points on the original plane exceeds the preset discrete point number threshold, it is determined that the original plane belongs to the artificial weld structure plane, otherwise, it is determined that the original plane does not belong to the artificial weld structure plane.

[0048] In a second aspect, the present application provides a computer device, comprising a processor and a memory, the memory stores a computer program executable by the processor, and the processor can execute the computer program to implement the artificial weld identification method of any one of the preceding embodiments.

[0049] In this case, the beneficial effects of the embodiments of the present application can include the following:

[0050] On the basis of determining the original plane set of the workpiece to be detected based on the surface point cloud data of the workpiece to be detected, each original plane included in the original plane set is traversed, and a target plane constituting a concave dihedral angle with the traversed original plane in the original plane set is searched, then for each target plane searched, a theoretical intersection line segment between the target plane and the traversed original plane is divided into a plurality of locally distributed local line segments, and the plurality of local line segments are clustered according to the minimum point cloud projection distances of the target plane and the original plane on different local line segments, to obtain the line segment clustering results of the target plane and the original plane on the plurality of local line segments, and then the line segment clustering results of the target plane and the original plane are used for artificial weld distribution detection to determine whether there is an artificial weld between the target plane and the original plane, so that on the basis of the plane intersection line segment segmentation and clustering results (which include the line segment clustering results of the two intersecting planes determined according to the point cloud projection distance distribution details of the two planes on the same intersection line segment division result), the high-precision artificial weld recognition function is realized by combining the artificial weld distribution detection operation (which involves hierarchical artificial weld structure feature detection on the three-dimensional point cloud data in the concave space range between the planes (i.e. the three-dimensional space range enclosed by the concave dihedral structure constituted by the two intersecting planes)), which can accurately locate and identify all artificial welds (whether involving continuous welding features or discontinuous welding features) actually existing on any workpiece to be detected.

[0051] In order to make the above objectives, characteristics and advantages of the present application more apparent, the following will describe a preferred embodiment in detail, and the accompanying drawings will be referred to, as follows. BRIEF DESCRIPTION OF DRAWINGS

[0052] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments, and it should be understood that the following drawings only show some embodiments of the present application, and therefore should not be regarded as a limitation to the scope, and for those skilled in the art, other related drawings can also be obtained without creative labor on the basis of these drawings.

[0053] Figure 1 The device composition schematic diagram of the computer device provided for the embodiments of the present application;

[0054] Figure 2 The flowchart of one of the artificial weld recognition methods provided for the embodiments of the present application;

[0055] Figure 3 The flowchart of the steps S230 included in the step S230 in Figure 2

[0056] Figure 4 ​​Figure 2 a flowchart of the steps of sub-step S240 in

[0057] Figure 5 a flowchart of the steps of sub-step S240 in Figure 4 a flowchart of the steps of sub-step S243 in

[0058] Figure 6 a flowchart of the steps of sub-step S245 in Figure 4 a flowchart of the steps of sub-step S245 in

[0059] Figure 7 a flowchart of the steps of sub-step S245 in

[0060] Fig. 10 is a schematic diagram of a computer device; Fig. 11 is a schematic diagram of a memory; Fig. 12 is a schematic diagram of a processor; and Fig. 13 is a schematic diagram of a communication unit. DETAILED DESCRIPTION

[0061] In order to make the objectives, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are some but not all of the embodiments of the present application. The components of the embodiments of the present application described and shown in the drawings can be arranged and designed in various different configurations.

[0062] Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the claimed present application, but only represents selected embodiments of the present application. All other embodiments obtained by a person of ordinary skill in the art without creative labor based on the embodiments in the present application are within the scope of protection of the present application.

[0063] It should be noted that: similar reference numerals and letters represent similar items in the following drawings, therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings.

[0064] In the description of the present application, it should be understood that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, or the orientation or positional relationship commonly understood by those skilled in the art, and are only for the convenience of describing the present application and simplifying the description, and are not intended to indicate or imply that the device or element referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation on the present application.

[0065] In the description of the application, it also needs to be explained that, unless otherwise explicitly specified and limited, the terms "set", "install", "connect", "connect" should be understood broadly, for example, can be fixedly connected, can also be detachably connected, or integrally connected; can be mechanically connected, can also be electrically connected; can be directly connected, can also be indirectly connected through an intermediate medium, can be the communication inside two elements. For those skilled in the art, the specific meaning of the above terms in the application can be understood according to the specific circumstances.

[0066] In addition, in the description of the application, it can be understood that the relationship terms such as "first" and "second" and the like are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between the entities or operations. Moreover, the terms "include", "contain" or any other variants thereof are intended to cover non-exclusive inclusion, so that the process, method, article or equipment including a series of elements not only includes those elements, but also includes other elements not explicitly listed or inherent to such process, method, article or equipment. Without more limitation, the element defined by the statement "including a" does not exclude the existence of other same elements in the process, method, article or equipment including the element. For those skilled in the art, the specific meaning of the above terms in the application can be understood according to the specific circumstances.

[0067] The applicant found through painstaking research that the existing artificial weld joint recognition scheme detects all workpiece planes using workpiece surface point cloud data, then traverses all the workpiece planes to detect whether any two intersecting planes meet the concave dihedral structure construction standard (which is used to determine whether the dihedral structure formed by the two intersecting planes presents a concave structure relative to the point cloud collection device), and performs edge detection on the two intersecting planes that meet the concave dihedral structure construction standard, respectively, to obtain edge point sets of the two intersecting planes near the intersection line segment of the planes. Then, straight line fitting is performed on the edge point sets of the two intersecting planes, respectively, to obtain two straight line segments (each straight line segment is individually in one of the two intersecting planes). At this time, the two straight line segments can directly construct a bounded plane. On this basis, the number of point clouds near the bounded plane can be counted from other workpiece planes or unallocated point clouds (i.e., point cloud data that has not been screened out to construct a workpiece plane), and the number of effective sampling points in the bounded plane can also be counted. Then, whether the number of point clouds near the bounded plane and the number of effective sampling points in the bounded plane meet the artificial weld joint discrimination requirements is detected. If both the number of point clouds and the number of effective sampling points meet the artificial weld joint discrimination requirements, it is directly determined that there is an artificial weld joint between the two intersecting planes.

[0068] However, it is worth noting that for artificial weld joints with non-continuous welding features, the edge points of each of the two intersecting planes in the local spatial region where the artificial weld joint exists are located on the edge of the artificial weld joint, and the edge points of each of the two intersecting planes in the local spatial region where the artificial weld joint does not exist are located on the intersection line of the planes. Therefore, when the existing artificial weld joint recognition scheme is used to process artificial weld joints with non-continuous welding features, the edge point sets of any two intersecting planes will have both edge points corresponding to the artificial weld joint and edge points corresponding to the intersection line of the planes, resulting in the two straight line segments fitted subsequently being more biased towards the intersection line of the planes, which cannot effectively represent the distribution of non-continuous artificial weld joints. In addition, because the existing artificial weld joint recognition scheme performs overall statistics on the concave spatial range between the planes in the point number statistics discrimination link, it does not consider the differences in the distribution characteristics of the nearby point clouds of continuous artificial weld joints and non-continuous artificial weld joints, resulting in the final artificial weld joint recognition result being unable to effectively count non-continuous artificial weld joints.

[0069] Therefore, the applicant developed a method for recognizing artificial weld joints and a computer device to accurately locate and identify all artificial weld joints (whether they involve continuous welding features or non-continuous welding features) on any workpiece to be detected, thereby solving the technical problems of the existing artificial weld joint recognition scheme.

[0070] Some embodiments of the present application will be described in detail with reference to the drawings. The following embodiments and features of the embodiments described below can be combined with each other without conflict.

[0071] Please refer to Figure 1 , Figure 1 is a device composition schematic diagram of a computer device 10 provided by the embodiments of the present application. In the embodiments of the present application, the computer device 10 can realize the manual weld seam identification function of different welding features (including continuous welding features and discontinuous welding features) for any to-be-detected workpiece which needs to identify the manual weld seam, and can effectively ensure the reliability of the final identification result, and realize the accurate positioning identification effect of the manual weld seam. Wherein, the computer device 10 can be a welding robot deployed with a visual perception system (which can be implemented by a depth camera, a laser radar and the like), or can be an independent electronic device in communication connection with the welding robot deployed with the visual perception system, wherein the independent electronic device can be, but is not limited to, a server, a personal computer, a notebook computer and the like.

[0072] In the embodiments of the present application, the computer device 10 can include a memory 11, a processor 12 and a communication unit 13. Wherein, the memory 11, the processor 12 and the communication unit 13 are directly or indirectly electrically connected with each other to realize the transmission or interaction of data. For example, the memory 11, the processor 12 and the communication unit 13 can be electrically connected with each other through one or more communication buses or signal lines.

[0073] In the embodiments of the present application, the memory 11 can be, but is not limited to, a random access memory (RAM), a read only memory (ROM), a programmable read only memory (PROM), an erasable programmable read only memory (EPROM), an electrically erasable programmable read only memory (EEPROM) and the like. Wherein, the memory 11 is used to store a computer program, and the processor 12 can execute the computer program accordingly after receiving an execution instruction.

[0074] In this embodiment, the processor 12 can be an integrated circuit chip with signal processing capabilities. The processor 12 can be a general-purpose processor, including at least one of a central processing unit (CPU), graphics processing unit (GPU), network processor (NP), digital signal processor (DSP), application-specific integrated circuit (ASIC), field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. The general-purpose processor can be a microprocessor or any conventional processor, capable of implementing or executing the methods, steps, and logic block diagrams disclosed in the embodiments of this application.

[0075] In this embodiment, the communication unit 13 is used to establish a communication connection between the computer device 10 and other electronic devices through a network, and to send and receive data through the network, wherein the network includes wired communication networks and wireless communication networks. For example, the computer device 10 can obtain surface point cloud data of the workpiece to be inspected (i.e., the real point cloud data of the surface of the workpiece to be inspected in three-dimensional space) from the visual perception system through the communication unit 13, and perform manual weld identification based on the surface point cloud data.

[0076] In this embodiment, the computer device 10 may pre-store a specific computer program related to the manual weld seam recognition function in the memory 11, and by driving the processor 12 to execute the specific computer program, a high-precision manual weld seam recognition function can be achieved through the organic combination of the planar intersecting line segment segmentation clustering operation and the manual weld seam distribution detection operation. It can accurately locate and identify all the manual weld seams that actually exist on any workpiece to be inspected (regardless of whether they involve continuous welding features or discontinuous welding features).

[0077] Understandable Figure 1 The block diagram shown is only a schematic diagram of one configuration of the computer device 10. The computer device 10 may also include components such as... 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.

[0078] In this application, to ensure that the computer device 10 can accurately locate and identify all artificial welds that actually exist on any workpiece to be inspected, this application embodiment provides an artificial weld identification method to achieve the aforementioned objective. The artificial weld identification method provided by this application will be described in detail below.

[0079] Please refer to Figure 2 , Figure 2 is one of the flowcharts of the artificial weld joint recognition method provided in the embodiments of the present application. In the embodiments of the present application, Figure 2 The artificial weld joint recognition method shown in the figure can include steps S210-S250.

[0080] Step S210, performing plane detection on the surface point cloud data of the workpiece to be detected to obtain an original plane set of the workpiece to be detected.

[0081] In the embodiments, the original plane set of the workpiece to be detected is composed of a plurality of original planes of the workpiece to be detected, which are obtained by processing the surface point cloud data using a traditional plane detection algorithm (for example, a region growing algorithm, a RANSAC (Random Sample Consensus) algorithm, etc.). Wherein, a plane number (which can be used to represent the identity of the original plane) can be configured for each original plane in the original plane set, and all original planes are arranged in the original plane set in ascending order of the number according to the size of the plane number of each original plane.

[0082] Step S220, traversing each original plane included in the original plane set to find a target plane in the original plane set that forms a concave dihedral angle with the traversed original plane.

[0083] In the embodiments, for any one original plane in the original plane set, each target plane matched with the original plane is another original plane intersecting the original plane and corresponding to a concave dihedral structure; that is, for any one original plane, any one target plane matched with the original plane in the original plane set will form a concave dihedral angle with the original plane, and the two target planes forming the concave dihedral angle have a concave three-dimensional space required for artificial weld joint formation.

[0084] When the original plane set of the workpiece to be detected is obtained, each original plane can be traversed in turn according to the arrangement order of each original plane in the original plane set, and for each traversed original plane, the step of "finding a target plane in the original plane set that forms a concave dihedral angle with the traversed original plane" is performed. Wherein, for a certain traversed original plane, if any one target plane matched with the original plane cannot be found in the original plane set, or all target planes matched with the original plane have been found in the original plane set, the traversal operation of the original plane can be ended, and the traversal operation of the next original plane can be started.

[0085] In an implementation form of the embodiment, to improve the efficiency of the artificial weld joint recognition and avoid performing unnecessary target plane searching operations, the computer device 10 can introduce a non-repeated combination mechanism in the target plane searching process to achieve the foregoing purpose. Specifically, when a certain original plane in the original plane set is traversed, target plane searching can be performed in the remaining planes (i.e., all original planes corresponding to plane numbers greater than the current traversal plane number (i.e., the actual plane number of the original plane currently traversed)) of the original plane set following the original plane, so as to avoid repeated plane combination globally.

[0086] At step S230, for each target plane found, the theoretical intersection line segment between the target plane and the traversed original plane is divided into a plurality of locally distributed local line segments, and the plurality of local line segments are clustered according to the minimum point cloud projection distances of the target plane and the original plane on different local line segments, to obtain the line segment clustering results of the target plane and the original plane on the plurality of local line segments.

[0087] In the embodiment, for each traversed original plane, when any one target plane matched with the original plane is determined, the theoretical intersection line segment between the two workpiece planes (i.e., the traversed original plane and the target plane matched therewith) is calculated, and then the theoretical intersection line segment is divided into a plurality of locally distributed local line segments according to a preset line segment length (e.g., 1.5 mm). Then, for each workpiece plane of the two workpiece planes, the minimum point cloud projection distance of the local plane point cloud of the surface point cloud data on the corresponding workpiece plane when projected into each local line segment is counted (which actually represents the shortest distance from the plane boundary profile of the corresponding workpiece plane close to the theoretical intersection line segment to each local line segment of the theoretical intersection line segment). For any local line segment, if there is a minimum point cloud projection distance of a related workpiece plane at the local line segment belonging to a value close to 0, it indicates that the local three-dimensional space where the local line segment is located basically does not exist artificial weld joint for the workpiece plane; and if there is a minimum point cloud projection distance of a workpiece plane at the local line segment belonging to a larger value (e.g., greater than a preset number greater than 0), it indicates that the local three-dimensional space where the local line segment is located may exist artificial weld joint for the workpiece plane.

[0088] When the minimum point cloud projection distances of the two workpiece planes with respect to the same intersection line segment (i.e., the plurality of local line segments) are determined, the minimum point cloud projection distances of each workpiece plane on the relevant local line segments and the relative arrangement relationship between the local line segments (which describes the arrangement order of the plurality of local line segments) are used to perform line segment distribution clustering (which can be implemented by using an adaptive clustering algorithm) on all local line segments, so that all local line segments belonging to the same clustering cluster (each clustering cluster corresponds to a point cloud projection distance value interval) are arranged in sequence according to the relative arrangement relationship, and the absolute value of the distance difference between the minimum point cloud projection distances of adjacent local line segments in the same clustering cluster is less than or equal to a preset distance difference threshold, thereby obtaining the line segment clustering result of the workpiece plane at the plurality of local line segments (which is composed of clustering clusters corresponding to all point cloud projection distance value intervals).

[0089] Optionally, referring to Figure 3 , Figure 3 is Figure 2 Step S230 in the method 1000 includes the flowchart of the sub-steps. In the embodiment of the present application, in order to ensure that the line segment clustering result determined based on the line segment distribution clustering operation is as intuitive and accurate as possible to describe the possibility of the existence of the artificial weld, the present application filters out the clustering cluster in the corresponding line segment clustering result which obviously indicates the non-existence of the artificial weld by performing the specific step details (i.e., the sub-step S231 to the sub-step S234 in the method 1000) of the relevant step "performing line segment distribution clustering on the plurality of local line segments according to the minimum point cloud projection distances of the to-be-clustered plane at the plurality of local line segments, to obtain the line segment clustering result of the to-be-clustered plane at the plurality of local line segments" in step S230 for each to-be-clustered plane in the target plane of any one of the traversed original planes. Figure 3

[0090] The sub-step S231 performs zero processing on the minimum point cloud projection distance of the to-be-clustered plane at any one local line segment which is less than the first preset distance threshold.

[0091] In the embodiment, the first preset distance threshold is a positive number close to 0. The sub-step S231 is used to directly set the local line segment with the minimum point cloud projection distance close to 0 to the state of non-existence of the artificial weld.

[0092] The sub-step S232 extracts the target line segment group from the plurality of local line segments according to the relative arrangement relationship between the plurality of local line segments.

[0093] ​In the embodiment, the minimum point cloud projection distance of each local line segment in any one of the extracted target line segment groups is greater than 0, and all the local line segments in the same target line segment group are arranged in sequence according to the relative arrangement relationship. At this time, any one of the target line segment groups can substantially represent a suspected intersection line segment region of the artificial weld at the corresponding to-be-clustered plane.

[0094] In the sub-step S233, for each target line segment group extracted, line segment clustering is performed according to the minimum point cloud projection distance of each local line segment included in the target line segment group, to obtain at least one original clustering cluster matched with the target line segment group.

[0095] In the embodiment, different original clustering clusters in the same target line segment group each correspond to a point cloud projection distance numerical interval, part of the original clustering clusters involved by different target line segment groups can correspond to the same point cloud projection distance numerical interval, and the number of local line segments involved by different original clustering clusters can be the same or different; all the local line segments belonging to the same original clustering cluster are arranged in sequence according to the relative arrangement relationship, and the distance difference between the minimum point cloud projection distances of two adjacent local line segments in the same original clustering cluster is less than or equal to a preset distance difference threshold.

[0096] In the sub-step S234, the original clustering cluster corresponding to the number of line segments greater than a preset line segment number threshold and the average projection distance greater than a second preset distance threshold is taken as one target clustering cluster in the line segment clustering result of the to-be-clustered plane.

[0097] In the embodiment, the average projection distance of any one original clustering cluster is the distance average value between the minimum point cloud projection distances of all the local line segments involved by the original clustering cluster at the same to-be-clustered plane (for example, the original plane traversed or the target plane corresponding thereto); the second preset distance threshold is greater than the first preset distance threshold, that is, the second preset distance threshold is used to preliminarily identify whether the local three-dimensional space involved by each clustering cluster is free of artificial weld or suspected to have artificial weld, that is, when there is at least one target clustering cluster in the line segment clustering result of any one to-be-clustered plane (for example, the original plane traversed), it is possible that there is an artificial weld in the concave space range between the to-be-clustered plane and another to-be-clustered plane (for example, any one target plane matched with the original plane traversed), but when there is no target clustering cluster in the line segment clustering result of the to-be-clustered plane, it is certain that there is no artificial weld in the concave space range between the to-be-clustered plane and another to-be-clustered plane.

[0098] Therefore, for each to-be-clustered plane in the original plane and any one of the target planes traversed, the line segment clustering result of each to-be-clustered plane can be ensured to be as intuitive and accurate as possible to describe the possibility of the existence of the artificial weld through the execution of the above sub-steps S231 to S234.

[0099] Step S240, according to the line segment clustering results of the target plane and the original plane respectively, artificial weld distribution detection is performed to determine whether there is an artificial weld between the target plane and the original plane.

[0100] In the embodiment, when the line segment clustering results of the target plane and any one of the target planes matched under the same intersection line segment division result are determined, the distribution state information about the target clustering cluster in the line segment clustering results of the two workpiece planes respectively (which can include the total number of target clustering clusters corresponding to a single workpiece plane, the arrangement order of all target clustering clusters belonging to the same workpiece plane on the corresponding theoretical intersection line segment, the clustering cluster coverage range of all target clustering clusters belonging to the same workpiece plane on the corresponding theoretical intersection line segment, etc.), combined with the specific distribution state of the local plane point cloud of the other original planes (which can be referred to as “reference planes”) except the aforementioned two workpiece planes (i.e. the original plane traversed and the target plane matched) in the original plane set, hierarchical artificial weld structure feature detection is performed in the concave space range between the aforementioned two workpiece planes, so as to realize the artificial weld distribution detection operation of the aforementioned two workpiece planes and effectively determine whether there is an artificial weld between the aforementioned two workpiece planes. Optionally, when it is determined that there is an artificial weld between the aforementioned two workpiece planes, a weld information array representing the artificial weld forming position can be constructed based on the plane numbers of the two workpiece planes respectively, so as to effectively determine the artificial weld distribution details on the workpiece to be detected through array reading.

[0101] Optionally, please refer to Figure 4 , Figure 4 is Figure 2 Step S240 in the flowchart of the sub-steps. In the embodiment, the step S240 can include sub-steps S241 to S247, so as to realize the detailed identification effect of the artificial weld through hierarchical artificial weld structure feature detection based on the plane intersection line segment division clustering result, and avoid the phenomenon of missing detection of the artificial weld.

[0102] Sub-step S241, detecting whether there is a target clustering cluster with a corresponding average projection distance greater than a second preset distance threshold in the line segment clustering results of the target plane and the original plane respectively.

[0103] In the embodiment, for the target plane traversed and any one of the target planes matched with the target plane, when at least one of the two workpiece planes (for example, the target plane traversed and / or the target plane matched with the original plane) does not involve any target cluster, it indicates that there is no possibility of the artificial weld seam between the two workpiece planes, and substep S242 is correspondingly executed; when at least one of the two workpiece planes (for example, the target plane traversed and / or the target plane matched with the original plane) involves one or more target clusters, it indicates that there is a possibility of the artificial weld seam between the two workpiece planes, and substep S243 is correspondingly executed.

[0104] Substep S242: determining that there is no artificial weld seam between the target plane and the original plane.

[0105] Substep S243: combining and pairing all target clusters involved in the target plane and the original plane respectively to obtain at least one set of effective cluster pairs between the target plane and the original plane.

[0106] In the embodiment, each set of effective cluster pairs includes an effective cluster corresponding to the original plane traversed and an effective cluster corresponding to the target plane found. Each set of effective cluster pairs corresponds to an artificial weld seam layer detection space between the two related workpiece planes.

[0107] Optionally, please refer to Figure 5 , Figure 5 is Figure 4 substep S243 in the step execution flow diagram. In the embodiment of the application, the substep S243 can include substep S243a to substep S243d to perform artificial weld seam layer detection space delineation between the two related workpiece planes based on the plane intersection line segment segmentation clustering result.

[0108] Substep S243a: detecting whether the total number of target clusters corresponding to the target plane and the original plane respectively is consistent.

[0109] In the embodiment, when the two related workpiece planes involved in the plane intersection line segment segmentation clustering result have the same total number of target clusters, it indicates that all target clusters involved in the two related workpiece planes can be paired one by one to form at least one set of effective cluster pairs, and substep S243b is correspondingly executed; when the total number of target clusters of the two related workpiece planes is different, it indicates that the number of effective clusters involved in the two related workpiece planes needs to be aligned through cluster fusion to form at least one set of effective cluster pairs, and substep S243c is correspondingly executed.

[0110] Sub-step S243b, traversing all target cluster groups involved in the original plane, taking each traversed target cluster group as one effective cluster group in a single set of effective cluster group pairs, and taking target cluster groups involved in the target plane and maintaining the same arrangement order as the traversed target cluster group as the other effective cluster group in the single set of effective cluster group pairs.

[0111] In the present embodiment, the arrangement order between all target cluster groups involved in each of the two relevant workpiece planes (i.e. the traversed original plane and its matched target plane) is matched with the relative arrangement relationship between the plurality of local line segments involved in the two relevant workpiece planes. For example, when the two relevant workpiece planes involve sequentially arranged local line segments 1-8, the target cluster groups involved in one workpiece plane are sequentially arranged as "{local line segment 1, local line segment 2}, {local line segment 5, local line segment 6} and {local line segment 7, local line segment 8}", and the target cluster groups involved in the other workpiece plane are sequentially arranged as "{local line segment 1, local line segment 2, local line segment 3}, {local line segment 4, local line segment 5} and {local line segment 6, local line segment 7, local line segment 8}", then the target cluster group {local line segment 1, local line segment 2} and the target cluster group {local line segment 1, local line segment 2, local line segment 3} form a set of effective cluster group pairs, the target cluster group {local line segment 5, local line segment 6} and the target cluster group {local line segment 4, local line segment 5} form a set of effective cluster group pairs, and the target cluster group {local line segment 7, local line segment 8} and the target cluster group {local line segment 6, local line segment 7, local line segment 8} form a set of effective cluster group pairs.

[0112] Sub-step S243c, determining a first plane with a smaller total number of target cluster groups and a second plane with a larger total number of target cluster groups in the target plane and the original plane.

[0113] Sub-step S243d, traversing all first target cluster groups involved in the first plane, taking each traversed first target cluster group as one effective cluster group in a single set of effective cluster group pairs, and performing cluster group merging on a plurality of second target cluster groups involved in the second plane for each traversed first target cluster group, so that the corresponding merged cluster group is the other effective cluster group in the single set of effective cluster group pairs.

[0114] In the present embodiment, the step "performing cluster group merging on a plurality of second target cluster groups involved in the second plane" in the above sub-step S243d can include sub-step A to sub-step E for each traversed first target cluster group, which are as follows.

[0115] Sub-step A: Construct initial clustering intervals that match the first target cluster encountered during traversal.

[0116] Wherein, the two boundary endpoints of the initial clustering interval on the corresponding theoretical intersection line segment are consistent with the two cluster boundary endpoints of the first target cluster traversed; wherein, the two cluster boundary endpoints of the first target cluster are respectively the two line segment endpoints of the first merged line segment of the first target cluster (which is formed by splicing together all the local line segments involved in the first target cluster).

[0117] Sub-step B: Traverse all second target clusters involved in the second plane and check whether the second target cluster encountered each time is in a merged state.

[0118] Sub-step C: If the second target cluster being traversed is in a merged state, then traverse the next second target cluster; otherwise, check whether at least one cluster boundary endpoint of the second target cluster being traversed is between the two interval boundary endpoints.

[0119] In this context, the two cluster boundary endpoints of any second target cluster are the two line segment endpoints of the second merged line segment of the second target cluster (which is formed by splicing together all the local line segments involved in the second target cluster).

[0120] Sub-step D: When the two cluster boundary endpoints of the traversed second target cluster are not located between the two interval boundary endpoints, traverse the next second target cluster; otherwise, mark the traversed second target cluster as merged, and update the target boundary endpoints of the two interval boundary endpoints that are closer to the reference boundary endpoint according to the reference boundary endpoint of the second target cluster that is not located between the two interval boundary endpoints.

[0121] Sub-step E: After traversing all the second target clusters, merge all the second target clusters covered by the initial clustering interval to obtain a merged cluster that matches the traversed first target clusters.

[0122] Therefore, this application can ensure the alignment of the number of effective clusters involved in each of the two related workpiece planes by performing the above sub-steps A to E and using cluster fusion methods.

[0123] This application can perform artificial weld layer detection spatial delineation between two related workpiece planes by executing the above sub-steps S243a to S243d, based on the results of planar intersecting line segmentation and clustering.

[0124] Sub-step S244: For each valid cluster pair, construct an intermediate plane that matches that valid cluster pair.

[0125] In the embodiment, for any one of the effective cluster pairs, the corresponding intermediate plane can be used to represent the central plane of the corresponding artificial weld layer detection space, the first plane intersection segment between the intermediate plane and the current traversed original plane is located in the original plane, and the actual distance between the first plane intersection segment and the related theoretical intersection segment is the average projection distance of the effective cluster corresponding to the original plane in the effective cluster pair; at the same time, the second plane intersection segment between the intermediate plane and the target plane is located in the target plane, and the actual distance between the second plane intersection segment and the corresponding theoretical intersection segment is the average projection distance of the effective cluster corresponding to the target plane in the effective cluster pair.

[0126] In the embodiment, for any one of the effective cluster pairs, the corresponding intermediate plane can be used to represent the central plane of the corresponding artificial weld layer detection space, the first plane intersection segment between the intermediate plane and the current traversed original plane is located in the original plane, and the actual distance between the first plane intersection segment and the related theoretical intersection segment is the average projection distance of the effective cluster corresponding to the original plane in the effective cluster pair; at the same time, the second plane intersection segment between the intermediate plane and the target plane is located in the target plane, and the actual distance between the second plane intersection segment and the corresponding theoretical intersection segment is the average projection distance of the effective cluster corresponding to the target plane in the effective cluster pair.

[0127] In the embodiment, for any one of the effective cluster pairs, the corresponding intermediate plane can be used to represent the central plane of the corresponding artificial weld layer detection space, the first plane intersection segment between the intermediate plane and the current traversed original plane is located in the original plane, and the actual distance between the first plane intersection segment and the related theoretical intersection segment is the average projection distance of the effective cluster corresponding to the original plane in the effective cluster pair; at the same time, the second plane intersection segment between the intermediate plane and the target plane is located in the target plane, and the actual distance between the second plane intersection segment and the corresponding theoretical intersection segment is the average projection distance of the effective cluster corresponding to the target plane in the effective cluster pair.

[0128] Optionally, please refer to Figure 6 , Figure 6 is Figure 4 the step execution flow diagram of the sub-step S245 in the embodiment. In the embodiment, for each intermediate plane, the above sub-step S245 can include sub-step S245a~sub-step S245d to detect the artificial weld structure feature for any one of the artificial weld layer detection spaces in the concave space range.

[0129] Sub-step S245a, a plurality of initial sampling points are uniformly generated in the intermediate plane, and the number of effective sampling points in the plurality of initial sampling points is counted based on all the target point clouds close to the intermediate plane.

[0130] In the embodiment, for each generated initial sampling point, a preset number (e.g., 8) of target discrete points closest to the initial sampling point can be found in the target point cloud of each reference plane near the corresponding intermediate plane, and then the average distance between the actual distances of the preset number of target discrete points to the initial sampling point is calculated, and then it is determined whether the initial sampling point belongs to the effective sampling point that may involve the artificial weld structure feature by judging whether the calculated average distance is less than a third preset distance threshold. When the average distance corresponding to a certain initial sampling point is less than the third preset distance threshold, it is determined that the initial sampling point belongs to the effective sampling point that may involve the artificial weld structure feature.

[0131] In the sub-step S245b, it is detected whether a first actual ratio between the total number of discrete points of all target point clouds and the maximum plane intersection segment length of the intermediate plane is greater than or equal to a first proportion threshold, and whether a second actual ratio between the number of effective sampling points and the total number of initial sampling points of the intermediate plane is greater than or equal to a second proportion threshold.

[0132] In the embodiment, the maximum plane intersection segment length of a single intermediate plane is the maximum value of the first plane intersection segment and the second plane intersection segment involved in the intermediate plane. When it is detected that the first actual ratio is greater than or equal to the first proportion threshold, and the second actual ratio is greater than or equal to the second proportion threshold, it indicates that there is essentially an artificial weld structure feature in the artificial weld layer detection space corresponding to the intermediate plane, and the sub-step S245c can be executed accordingly; when it is detected that the first actual ratio is less than the first proportion threshold, and / or the second actual ratio is less than the second proportion threshold, it indicates that there is essentially no artificial weld structure feature in the artificial weld layer detection space corresponding to the intermediate plane, and the sub-step S245d can be executed accordingly.

[0133] In the sub-step S245c, it is determined that there is an artificial weld structure feature near the intermediate plane.

[0134] In the embodiment, when it is determined that there is an artificial weld structure feature near a certain intermediate plane, the target discrete points of each reference plane in the corresponding target point cloud can be directly associated with the detected artificial weld structure feature.

[0135] In the sub-step S245d, it is determined that there is no artificial weld structure feature near the intermediate plane.

[0136] In the embodiment, when it is determined that there is no artificial weld structure feature near a certain intermediate plane, the target discrete points of each reference plane in the corresponding target point cloud cannot essentially establish an association relationship with the artificial weld structure feature.

[0137] Thus, the present application can perform the above-mentioned sub-step S245a to sub-step S245d to detect the artificial weld structure feature for any artificial weld layer detection space in the concave space range.

[0138] Sub-step S246, when detecting that there is an artificial weld structure feature near at least one intermediate plane, it is determined that there is an artificial weld between the target plane and the original plane.

[0139] Sub-step S247, when detecting that there is no artificial weld structure feature near all intermediate planes, it is determined that there is no artificial weld between the target plane and the original plane.

[0140] Thus, the present application can perform the above-mentioned sub-step S241 to sub-step S247 to realize the detailed identification effect of the artificial weld by the hierarchical artificial weld structure feature detection based on the plane intersection line segment segmentation clustering result, and avoid the artificial weld missing detection phenomenon.

[0141] The present application can perform the above-mentioned step S210 to step S240 to realize the high-precision artificial weld identification function by the organic combination between the plane intersection line segment segmentation clustering operation and the artificial weld distribution detection operation, and can accurately position and identify all artificial welds actually existing on any to-be-detected workpiece.

[0142] Optionally, please refer to Figure 7 , Figure 7 is the second flowchart of the artificial weld identification method provided by the embodiment of the present application. In the embodiment of the present application, compared with the artificial weld identification method shown in Figure 2 , the difference between the artificial weld identification method shown in Figure 7 lies in that: Figure 7 the artificial weld identification method shown in Figure 2 , in the process of performing step S220, step S310 is performed before step S220a “finding a target plane in the original plane set which forms a concave dihedral angle with the traversed original plane”, and step S320 is inserted between the original step S220 and step S230, so as to verify the necessity of artificial weld identification for the traversed original plane and any one target plane found by step S310 and step S320, thereby improving the overall artificial weld identification efficiency.

[0143] Step S310, traversing each original plane included in the original plane set, detecting whether the traversed original plane belongs to the artificial weld structure plane.

[0144] In the embodiment, the artificial weld structure plane is used to represent the weld surface flattening result of the artificial weld; and the computer device 10, before performing step S220a, performs artificial weld structure plane detection on each original plane traversed, to determine whether it is necessary to continue performing the subsequent method step procedure (including step S220a, step S320, step S230 and step S240) on the current traversed original plane.

[0145] When it is detected that the current traversed original plane belongs to the artificial weld structure plane, it indicates that the original plane itself belongs to the part of the artificial weld, and there is no need to further identify the artificial weld, and the traversal operation on the original plane can be directly ended, and the traversal operation on the next original plane can be started (the next original plane can be traversed by re-executing step S310).

[0146] When it is detected that the current traversed original plane does not belong to the artificial weld structure plane, it indicates that the original plane itself does not belong to the part of the artificial weld, and there is a need to further identify the artificial weld, and step S220a “finding the target plane in the original plane set which forms a concave dihedral angle with the traversed original plane” can be continued.

[0147] In the embodiment, the step of detecting whether the original plane belongs to the artificial weld structure plane can include:

[0148] counting whether the number of target discrete points on the original plane associated with the existing artificial weld structure feature exceeds a preset discrete point threshold;

[0149] If the number of target discrete points on the original plane exceeds the preset discrete point threshold, it is determined that the original plane belongs to the artificial weld structure plane, otherwise it is determined that the original plane does not belong to the artificial weld structure plane.

[0150] Step S320, detecting whether each found target plane belongs to the artificial weld structure plane.

[0151] In the embodiment, when one or more target planes matching the original plane are found for any original plane which does not belong to the artificial weld structure plane, artificial weld structure plane detection is performed on each found target plane, to determine whether it is necessary to continue performing the subsequent method step procedure (including step S230 and step S240) on the current found target plane.

[0152] When it is detected that the current found target plane belongs to the artificial weld structure plane, it indicates that the target plane itself belongs to the part of the artificial weld, and there is no necessity for the target plane to perform the artificial weld identification with the matched original plane, at this time, the further processing of the target plane can be directly ended, and the searching operation of the next target plane is started (the next target plane matched with the original plane traversed can be searched by re-jumping to execute step S220a).

[0153] When it is detected that the current found target plane does not belong to the artificial weld structure plane, it indicates that the target plane itself does not belong to the part of the artificial weld, and there is the necessity for the target plane to perform the artificial weld identification with the matched original plane, at this time, the steps S230 and S240 can be continuously executed for the target plane and the original plane which do not belong to the artificial weld plane.

[0154] Therefore, the artificial weld identification method shown in the application can effectively improve the overall artificial weld identification efficiency in the process of realizing the accurate artificial weld positioning and identification effect for any to-be-detected workpiece. Figure 7

[0155] In the embodiments provided by the present application, it should be understood that the disclosed device and method can also be implemented in other ways. The device embodiments described above are only schematic. For example, the flowcharts and block diagrams in the accompanying drawings show the possible implementation architectures, functions and operation of the device, method and computer program product according to the embodiments of the present application. In this regard, each block in the flowcharts or block diagrams can represent a module, a program segment or a part of code, which contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions noted in the blocks can occur in a different order than that noted in the accompanying drawings. For example, two consecutive blocks can actually be executed substantially in parallel, and sometimes they can be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and the combination of blocks in the block diagrams and / or flowcharts, can be implemented by a dedicated hardware-based system for executing the specified functions or actions, or can be implemented by a combination of dedicated hardware and computer instructions.

[0156] ​In addition, each function module in each embodiment of the present application can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part. Each function provided by the present application, if realized in the form of a software function module and sold or used as an independent product, can be stored in a storage medium. Based on this understanding, the technical solutions of the present application or the parts that essentially contribute to the prior art or the parts of the technical solutions can be embodied in the form of a software product, which is stored in a readable storage medium and includes a plurality of instructions for causing a computer device (for example, a notebook computer, a welding robot deployed with a visual perception system, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned readable storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various media that can store program codes.

[0157] The above is only various embodiments of the present application, but the protection scope of the present application is not limited thereto, and any person skilled in the art can easily think of changes or replacements within the technical scope disclosed by the present application, which should be covered by the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method of artificial weld seam recognition, characterized in that The identification method comprises: plane detection is performed on surface point cloud data of a workpiece to be detected to obtain an original plane set of the workpiece to be detected; each original plane included in the original plane set is traversed to find a target plane in the original plane set that forms a concave dihedral angle with the traversed original plane; for each target plane found, a theoretical intersection line segment between the target plane and the traversed original plane is divided into a plurality of locally distributed partial line segments, and line segment distribution clustering is performed on the plurality of partial line segments according to minimum point cloud projection distances of the target plane and the original plane on different partial line segments to obtain line segment clustering results of the target plane and the original plane on the plurality of partial line segments; artificial weld distribution detection is performed according to the line segment clustering results of the target plane and the original plane to determine whether an artificial weld exists between the target plane and the original plane; wherein the step of performing artificial weld distribution detection according to the line segment clustering results of the target plane and the original plane to determine whether an artificial weld exists between the target plane and the original plane comprises: detecting whether there is a target clustering cluster with a corresponding average projection distance greater than a second preset distance threshold in the line segment clustering results of the target plane and / or the original plane, and determining that no artificial weld exists between the target plane and the original plane in a case where there is no at least one target clustering cluster in the line segment clustering results of the target plane and / or the original plane; wherein for each to-be-clustered plane in the target plane and any one target plane, line segment distribution clustering is performed on the plurality of partial line segments according to minimum point cloud projection distances of the to-be-clustered plane on the plurality of partial line segments to obtain line segment clustering results of the to-be-clustered plane on the plurality of partial line segments, the step comprising: zero processing is performed on the minimum point cloud projection distance of the to-be-clustered plane on any one partial line segment that is less than a first preset distance threshold, wherein the first preset distance threshold is a positive number close to 0; target line segment groups are extracted from the plurality of partial line segments according to a relative arrangement relationship between the plurality of partial line segments; wherein the minimum point cloud projection distances of all partial line segments in any one target line segment group extracted are all greater than 0, and all partial line segments in a same target line segment group are arranged in sequence according to the relative arrangement relationship; for each target line segment group extracted, line segment clustering is performed according to the minimum point cloud projection distances of all partial line segments included in the target line segment group to obtain at least one original clustering cluster matched with the target line segment group; wherein all partial line segments belonging to a same original clustering cluster are arranged in sequence according to the relative arrangement relationship, and a distance difference absolute value between the minimum point cloud projection distances of adjacent two partial line segments in a same original clustering cluster is less than or equal to a preset distance difference threshold. The original cluster corresponding to the number of line segments greater than the preset line segment threshold and the average projection distance greater than the second preset distance threshold is taken as a target cluster in the line segment clustering result of the to-be-clustered plane.

2. The identification method according to claim 1, characterized in that, The step of performing manual weld distribution detection according to the line segment clustering results of the target plane and the original plane to determine whether there is a manual weld between the target plane and the original plane further comprises: In the case where it is detected that there is at least one target cluster in the line segment clustering results of the target plane and the original plane respectively, all target clusters involved in the target plane and the original plane are combined and paired to obtain at least one group of effective cluster pairs between the target plane and the original plane. For each group of effective cluster pairs, an intermediate plane matching the group of effective cluster pairs is constructed; wherein a first plane intersection line segment between the intermediate plane and the original plane is located in the original plane, an actual distance between the first plane intersection line segment and a corresponding theoretical intersection line segment is an average projection distance of an effective cluster corresponding to the original plane in the group of effective cluster pairs, a second plane intersection line segment between the intermediate plane and the target plane is located in the target plane, and an actual distance between the second plane intersection line segment and a corresponding theoretical intersection line segment is an average projection distance of an effective cluster corresponding to the target plane in the group of effective cluster pairs. For each intermediate plane, whether there is a manual weld structure feature near the intermediate plane is detected according to target point clouds near the intermediate plane on each reference plane of the original plane set except the target plane and the original plane. When it is detected that there is at least one manual weld structure feature near the intermediate plane, it is determined that there is a manual weld between the target plane and the original plane. When it is detected that there is no manual weld structure feature near all intermediate planes, it is determined that there is no manual weld between the target plane and the original plane.

3. The identification method according to claim 2, characterized in that, The step of combining and pairing all target clusters involved in the target plane and the original plane to obtain at least one group of effective cluster pairs between the target plane and the original plane comprises: Detect whether the total number of target clusters corresponding to the target plane and the original plane respectively remains consistent; When the total number of target clusters corresponding to the target plane and the original plane respectively remains consistent, all target clusters involved in the original plane are traversed, each time the target cluster traversed is taken as an effective cluster in a single group of effective cluster pairs, and the target cluster of the target plane involved and maintaining the same arrangement order as the target cluster traversed is taken as another effective cluster in the single group of effective cluster pairs; wherein the arrangement order between all target clusters involved in the target plane and the original plane matches the relative arrangement relationship between the plurality of local line segments.

4. The identification method according to claim 3, characterized in that, The step of combining and pairing all target clusters involved in the target plane and the original plane to obtain at least one group of effective cluster pairs between the target plane and the original plane further comprises: when the total number of target cluster groups corresponding to the target plane and the original plane respectively does not remain consistent, determining a first plane with a smaller total number of target cluster groups and a second plane with a larger total number of target cluster groups in the target plane and the original plane; traversing all first target cluster groups involved in the first plane, taking each traversed first target cluster group as one effective cluster group in a single set of effective cluster group pairs, and performing cluster group merging on a plurality of second target cluster groups involved in the second plane for each traversed first target cluster group, so as to take a corresponding merged cluster group as the other effective cluster group in the single set of effective cluster group pairs.

5. The identification method according to claim 4, characterized in that, For each traversed first target cluster group, the step of performing cluster group merging on the plurality of second target cluster groups involved in the second plane comprises: constructing an initial cluster interval matched with the traversed first target cluster group, wherein two interval boundary endpoints of the initial cluster interval remain consistent with two cluster group boundary endpoints of the traversed first target cluster group; traversing all second target cluster groups involved in the second plane, and detecting whether each traversed second target cluster group is in a merged state; if the traversed second target cluster group is in the merged state, traversing the next second target cluster group, otherwise detecting whether at least one cluster group boundary endpoint of the traversed second target cluster group is between the two interval boundary endpoints; when neither of the two cluster group boundary endpoints of the traversed second target cluster group is between the two interval boundary endpoints, traversing the next second target cluster group, otherwise marking the traversed second target cluster group as in the merged state, and updating a target boundary endpoint of the two interval boundary endpoints close to a reference boundary endpoint of the second target cluster group which is not between the two interval boundary endpoints; after traversing all the second target cluster groups, performing cluster group merging on all second target cluster groups covered by the initial cluster interval to obtain a merged cluster group matched with the traversed first target cluster group.

6. The identification method according to any one of claims 2 to 5, characterized in that, For each intermediate plane, the step of detecting whether there is an artificial weld structure feature near the intermediate plane according to target point clouds close to the intermediate plane on each reference plane except the target plane and the original plane from the original plane set comprises: generating a plurality of initial sampling points uniformly in the intermediate plane, and counting a number of effective sampling points in the plurality of initial sampling points based on all target point clouds close to the intermediate plane; wherein an average distance from each effective sampling point to a preset number of target discrete points adjacent to the effective sampling point in the target point cloud is less than a third preset distance threshold; detecting whether a first actual ratio between a total number of discrete points of the all target point clouds and a maximum plane intersection segment length of the intermediate plane is greater than or equal to a first proportion threshold, and whether a second actual ratio between the number of effective sampling points and a total number of initial sampling points of the intermediate plane is greater than or equal to a second proportion threshold; If the first actual ratio is greater than or equal to a first ratio threshold and the second actual ratio is greater than or equal to a second ratio threshold, it is determined that the artificial weld structure feature exists near the intermediate plane, otherwise it is determined that the artificial weld structure feature does not exist near the intermediate plane.

7. The identification method according to claim 6, characterized in that, The identification method further comprises: In the process of traversing each original plane included in the original plane set, it is detected whether the traversed original plane belongs to an artificial weld structure plane; If the traversed original plane belongs to an artificial weld structure plane, the next original plane is traversed, otherwise the step of searching for a target plane in the original plane set which forms a concave dihedral angle with the traversed original plane is performed, and it is detected whether each searched target plane belongs to an artificial weld structure plane; If it is detected that the searched target plane belongs to an artificial weld structure plane, the step of searching for a target plane in the original plane set which forms a concave dihedral angle with the traversed original plane is continuously performed, otherwise the step of dividing a theoretical intersection line segment between the target plane and the traversed original plane into a plurality of continuously distributed local line segments is performed; For any one original plane in the original plane set, the step of detecting whether the original plane belongs to an artificial weld structure plane comprises: It is detected whether the number of target discrete points on the original plane which are associated with the existing artificial weld structure feature exceeds a preset discrete point number threshold; If the number of target discrete points on the original plane exceeds the preset discrete point number threshold, it is determined that the original plane belongs to an artificial weld structure plane, otherwise it is determined that the original plane does not belong to an artificial weld structure plane.

8. A computer device, comprising: The processor and the memory, the memory stores a computer program which can be executed by the processor, and the processor can execute the computer program to realize the artificial weld identification method in any one of claims 1-7.

Images