An automatic welding method for wire mesh pallets based on visual analysis

Abstract

The present application relates to the field of welding positioning technology, and specifically to an automatic welding method for a wire mesh pallet based on visual analysis, the method comprising: obtaining horizontal and vertical overlapping images during the welding process of the wire mesh pallet; using a corner point detection algorithm to obtain suspected welding corner points; obtaining a welding positioning corner point set; confirming the relative degree of overlapping corner points between any two elements in the set; obtaining the overlapping corner point determination degree of each element in the set, and obtaining overlapping suspected corner points; obtaining an angle monitoring set for each overlapping suspected corner point, and marking the overlapping suspected corner points; combining a threshold segmentation algorithm to obtain overlapping corner points of horizontal and vertical overlapping images; using a clustering algorithm for all overlapping corner points of the horizontal and vertical overlapping images, and obtaining the best welding point of the wire mesh pallet based on the element distribution in the obtained clustering cluster. The present application aims to improve the accuracy of welding point positioning.

Description

Technical Field

[0001] The present application relates to the field of welding positioning technology, and in particular to an automatic welding method for a wire mesh pallet based on visual analysis. Background Art

[0002] The automatic welding machine plays a vital role in the production of wire mesh pallets. It can efficiently and accurately complete the welding of metal wires, ensure the quality and consistency of the welding points, and thus improve the strength and stability of the entire wire mesh. Through the automated welding process, the production process is more efficient, reducing the errors and instability in manual operation, while improving the consistency and precision of the product, reducing production costs, and providing solid technical support for large-scale, high-quality wire mesh production.

[0003] In the automated welding process, the longitudinal wire is transported by movement, and the transverse wire falls on the longitudinal wire at a fixed position, and then welded through multiple welding points. However, because the transverse wire is not completely straight, when it falls on the longitudinal wire, the welding point is offset from the center of the welding gun laser. This offset makes it impossible for the laser to accurately align with the center of the intersection of the transverse and longitudinal wires, resulting in a weak welding point, affecting the strength and stability of the overall welding. Summary of the invention

[0004] In view of the above, it is necessary to provide an automatic welding method for wire mesh pallets based on visual analysis to solve the above problems.

[0005] An embodiment of the present application provides a method for automatic welding of a wire mesh pallet based on visual analysis, the method comprising: Get each horizontal and vertical overlapping image during the welding process of the wire mesh pallet; A corner point detection algorithm is used for each horizontal and vertical overlapping image, and all the obtained corner points are used as suspected welding corner points. A welding positioning corner point set is determined based on all the suspected welding corner points in a preset neighborhood of each suspected welding corner point. Based on the distance measurement between any two elements in each welding positioning corner point set, determine the relative degree of overlapping corner points between the arbitrary two elements; analyze the relative degree of overlapping corner points between each element in each welding positioning corner point set and all other elements, obtain the overlapping corner point determination degree of each element, and obtain the overlapping suspected corner points; In each welding positioning corner point set, the position distribution characteristics of each overlapping suspected corner point and other elements are calculated based on the relative degree of the overlapping corner point to obtain the angle monitoring set of each overlapping suspected corner point; based on the element values ​​in the angle monitoring set, the overlapping suspected corner points are marked; according to the number of times each overlapping suspected corner point in the horizontal and vertical overlapping images is marked, combined with the threshold segmentation algorithm, the overlapping corner points of the horizontal and vertical overlapping images are obtained; A clustering algorithm is used for all overlapping corner points of the horizontally and vertically overlapping images, and based on the element distribution in the obtained clustering clusters, the optimal welding point of the wire mesh pallet is obtained, and welding is performed on the optimal welding point.

[0006] The process of obtaining the relative degree of the overlapping corner points between any two elements is as follows: For each welding point location corner point set of suspected welding corner points, calculate the distance metric between any two elements; When the distance metric is within a preset interval, a first preset value is used as the relative degree of the overlapping corner points between the arbitrary two elements; otherwise, a second preset value is used as the relative degree of the overlapping corner points between the arbitrary two elements; wherein the first preset value is greater than the second preset value.

[0007] The overlapping corner point determination degree of each element is obtained by accumulating the relative degrees of overlapping corner points between each element and all other elements in each welding positioning corner point set.

[0008] The overlapping suspected corner points are specifically elements in the welding positioning corner point set of each suspected welding corner point whose overlapping corner point certainty is greater than or equal to a third preset value.

[0009] The step of obtaining the angle monitoring set of each overlapping suspected corner point is specifically as follows: For each suspected overlapping corner point in each welding corner point location set, screening the remaining elements in the welding corner point location set based on the relative degree of the overlapping corner point; Connect each overlapping suspected corner point with the selected elements, count the angles between any two connecting lines, and obtain the angle monitoring set of each overlapping suspected corner point.

[0010] The remaining elements in the welding corner point positioning set are screened based on the relative degree of the overlapping corner points, specifically, the elements whose relative degree of the overlapping corner points with each overlapping suspected corner point is equal to the first preset value are screened.

[0011] The marking of the overlapping suspected corner points is specifically: if there are elements in the angle monitoring set that are within a preset angle interval, the overlapping suspected corner points corresponding to the angle monitoring set are marked.

[0012] The process of obtaining the overlapping corner points of the horizontally and vertically overlapping images is specifically as follows: The number of times all overlapping suspected corner points are marked is segmented using a threshold value to obtain a segmentation threshold value; and the overlapping suspected corner points whose number of times are marked is greater than or equal to the segmentation threshold value are recorded as overlapping corner points.

[0013] Wherein, the clustering algorithm is used for all overlapping corner points of the horizontally and vertically overlapping images, including: The coordinates of all overlapping corner points of the horizontally and vertically overlapping images are used as inputs of the DBSCAN clustering algorithm, the neighborhood radius is taken as the upper limit value of the preset interval; the minimum number is taken as the third preset value, and each cluster is obtained.

[0014] The optimal welding point of the wire mesh pallet is determined by the centroid of the actual coordinates of all elements in each cluster of the horizontal and vertical overlapping images on the wire mesh pallet.

[0015] This application has at least the following beneficial effects: The embodiment of the present application analyzes the horizontal and vertical overlapping images of the wire mesh pallet during the welding process, and adopts a corner detection algorithm according to the characteristics of the overlap of the two wires to obtain suspected welding corner points, which helps to narrow the scope of finding overlapping corner points and improve the search efficiency; with the help of the characteristics of the overlapping wires, a welding positioning corner point set is obtained for each suspected welding corner point, which helps to analyze the remaining suspected welding corner points within the preset neighborhood of each suspected welding corner point; further, all suspected welding corner points in the set are analyzed pairwise to obtain the relative degree of overlapping corner points and the confirmation degree of overlapping corner points, and the overlapping suspected corner points are obtained. The beneficial effect is that it is judged whether the corner point is caused by the overlap of the wires according to the distance, and the overlap of the overlapping corner points is further narrowed. Small range; in each welding positioning corner point set, based on the relative degree of the overlapping corner point, the position distribution characteristics of each overlapping suspected corner point and the remaining elements are compared to obtain an angle monitoring set of each overlapping suspected corner point, which has the beneficial effect of combining the distance characteristics and angle distribution characteristics of the corner points of the overlapping wires to make another judgment on the suspected overlapping corner points, so that the overlapping corner points obtained subsequently are more accurate; finally, the clustering algorithm is used to help cluster the corner points generated by the overlap of the same longitudinal wire and a transverse wire into one category, so that the optimal welding point of the wire mesh pallet is more accurate, which can significantly improve the accuracy and automation level of wire mesh welding, enhance welding quality control, and thus improve the efficiency and reliability of welding operations. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A flowchart of a method for automatic welding of wire mesh pallets based on visual analysis provided in this application; Figure 2 A schematic diagram of the overlap of a transverse metal wire and a plurality of longitudinal metal wires provided in the present application; Figure 3 Schematic diagram of overlapping corner points provided for this application. DETAILED DESCRIPTION

[0017] In the description of the embodiments of the present application, words such as "exemplary", "or", "for example" and the like are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary", "or", "for example" and the like is intended to present related concepts in a concrete manner.

[0018] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art in the present application. The terms used in the specification of the present application are only for the purpose of describing specific embodiments and are not intended to limit the present application.

[0019] It should also be noted that the terms "first" and "second" in this application and its drawings are used to distinguish similar objects, rather than to describe a specific order or sequence. The method disclosed in the embodiments of the present application or the method shown in the flow chart includes one or more steps for implementing the method. Without departing from the scope of protection of this application, the execution order of multiple steps can be interchanged with each other, and some steps can also be deleted.

[0020] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.

[0021] This application proposes an automatic welding method for wire mesh pallets based on visual analysis, which is applied to the field of welding positioning technology. Figure 1 , the method comprises the following steps: S1: Acquire each horizontal and vertical overlapping image during the welding process of the wire mesh pallet.

[0022] In the process of welding the wire mesh pallet, the longitudinal wire is accurately fed into the longitudinal groove of the welding machine and transmitted through precision devices such as bearing wheels. In the present application, the longitudinal wire stops once each time it advances a preset length so that a transverse wire can be dropped for laser welding; in this embodiment, the preset length is 15 cm; the specific advancement distance of the longitudinal wire can be adjusted by the implementer according to the actual situation. Furthermore, each time a transverse wire is dropped, the overlapping image of the longitudinal wire and the transverse wire is collected by a CCD camera.

[0023] In order to facilitate subsequent processing, the overlapping image is converted into a grayscale image by weighted averaging, and then the grayscale image is filtered by a non-local mean (NLM) filter, and the filtered image is used as the horizontal and vertical overlapping image. Among them, the weighted averaging method and NLM filtering are both well-known technologies, and the specific calculation process is not repeated here.

[0024] S2: A corner point detection algorithm is used for each horizontal and vertical overlapping image, and all the obtained corner points are used as suspected welding corner points. A welding positioning corner point set is determined based on all the suspected welding corner points in a preset neighborhood of each suspected welding corner point.

[0025] Since the wires on the wire mesh pallet are in the shape of round cylinders, during the laser welding process, the horizontal and vertical wires will form a cross. Each laser welding is performed on the same horizontal wire and multiple vertical wires. Therefore, multiple cross shapes will appear in the image captured by the camera. Since the wires have a certain width, the horizontal and vertical interlaced areas can be seen as the overlap of multiple thick line segments, presenting a cross-shaped structure. The schematic diagram of the overlap of a horizontal wire and multiple vertical wires is shown in the figure below. Figure 2 shown.

[0026] A corner point detection algorithm is used for each horizontal and vertical overlapping image. The edges of the horizontal and vertical metal wires will form multiple corner points, which is convenient for locating the overlapping position of the horizontal and vertical metal wires through the position of the corner points, that is, the welding position; in this embodiment, the Harris corner point detection algorithm is used, and the implementer may also use other well-known corner point detection algorithms, which are not limited in this application. All corner points in each horizontal and vertical overlapping image are regarded as suspected welding corner points.

[0027] Since the distance between the corner points of each wire is close to the diameter of the wire. Therefore, the area formed by taking each suspected welding corner point as the center point and a times the wire diameter as the radius is used as the neighborhood of each suspected welding corner point; all the suspected welding corner points in the neighborhood of each suspected welding corner point are obtained to form a welding positioning corner point set for each suspected welding corner point. It should be noted that since the wire mesh pallet considered in this embodiment has two vertically overlapping wires, for any two crossed wires, the maximum distance of the overlapping area is the wire diameter. times, so the neighborhood radius of each suspected welding corner point should be greater than Therefore, the value of a in this embodiment is 1.5, and the implementer can adjust it according to the actual situation.

[0028] S3: Based on the distance measurement between any two elements in each welding positioning corner point set, determine the relative degree of overlapping corner points between the any two elements; analyze the relative degree of overlapping corner points between each element in each welding positioning corner point set and all other elements, obtain the overlapping corner point determination degree of each element, and obtain the overlapping suspected corner points.

[0029] Since the metal wire mesh is coated with an anti-corrosion coating only after welding, and since the metal wire will undergo an oxidation-reduction reaction, there are often oxidized and reduced areas on the metal wire. The boundary points of this area may be detected as corner points. Therefore, it is necessary to judge the corner points to determine whether they are the corner points generated by the overlap of the horizontal and vertical metal wires.

[0030] For each welding point positioning corner point set of a suspected welding corner point, calculate the distance metric between any two elements. When the distance metric is within a preset interval, use the first preset value as the relative degree of the overlapping corner points between the two elements; otherwise, use the second preset value as the relative degree of the overlapping corner points between the two elements; wherein the first preset value is greater than the second preset value. In this embodiment, the distance between two corner points is measured using the Euclidean distance; the preset interval is , where d represents the diameter of the metal wire. It should be noted that there is a certain error in the thickness of the metal wire. In this embodiment, the error is ±1%; the implementer can adjust it according to the actual situation; in addition, in this embodiment, the first preset value is 1, and the second preset value is 0.

[0031] Furthermore, the cumulative sum of the relative degrees of overlapping corner points between each element in the welding positioning corner point set of each suspected welding corner point and all other elements is used as the overlapping corner point determination degree of each element.

[0032] It should be understood that if the distance between each element in the welding positioning corner point set of each suspected welding corner point and the other elements is closer to the diameter of the metal wire, it means that the element conforms to the distribution state of the overlapping corner point, and it is more likely to be an overlapping corner point. At the same time, the value of the overlapping corner point certainty of the element is larger; it should be noted that the overlapping corner points are specifically the four corner points formed by the intersection of the horizontal metal wire and the longitudinal metal wire, and its schematic diagram is as follows Figure 3 As shown; in the figure, A, B, C, and D are all overlapping corner points.

[0033] Ideally, each overlap position will usually form only four overlap corner points. The distance between adjacent overlap corner points is within a preset range, and there are usually only two overlap corner points. This is because the overlap corner points are formed by the intersection of the edges of the metal wires, and since the metal wire has a certain diameter, the distance between adjacent overlap corner points in the horizontal direction or vertical direction is usually close to the diameter of the metal wire.

[0034] S4: In each welding positioning corner point set, based on the relative degree of the overlapping corner point, each overlapping suspected corner point and the position distribution characteristics of the remaining elements are obtained to obtain the angle monitoring set of each overlapping suspected corner point; based on the element values ​​in the angle monitoring set, the overlapping suspected corner points are marked; according to the number of times each overlapping suspected corner point in the horizontal and vertical overlapping images is marked, combined with the threshold segmentation algorithm, the overlapping corner points of the horizontal and vertical overlapping images are obtained.

[0035] Due to the corrosive effect of the environment on the metal wire, multiple suspected welding corner points may appear around each overlapping corner point. In order to prevent them from being mistaken for overlapping corner points, each element is marked based on the overlapping corner point determination degree of each element and the distribution between the elements: elements with overlapping corner point determination degrees greater than or equal to the third preset value are recorded as overlapping suspected corner points; in each welding point positioning set, suspected welding corner points whose overlapping corner point relative degrees with each overlapping suspected corner point are equal to the first preset value are screened, each overlapping suspected corner point is connected with the screened suspected welding corner points, and the angles between any two connecting lines are counted to obtain the angle monitoring set of each overlapping suspected corner point; if there are elements in the angle monitoring set that are within the preset angle range, the overlapping suspected corner points corresponding to the angle monitoring set are marked.

[0036] In this embodiment, the third preset value is 2, and the preset angle interval is , implementers can adjust the angle range according to actual conditions.

[0037] Since the same overlapping suspected corner point may be in the welding positioning corner point set of different suspected welding corner points, the same overlapping suspected corner point may be marked multiple times. The number of times each suspected welding corner point in the horizontal and vertical overlapping images is marked as an overlapping suspected corner point is counted, and the number of times all overlapping suspected corner points are marked is subjected to threshold segmentation to obtain a segmentation threshold; overlapping suspected corner points whose number of marks is greater than or equal to the segmentation threshold are recorded as overlapping corner points; in this embodiment, the threshold segmentation method used is Otsu threshold segmentation, which is a known technology and is not limited in this application.

[0038] S5: A clustering algorithm is used for all overlapping corner points of the horizontally and vertically overlapping images, and based on the element distribution in the obtained clustering clusters, the optimal welding point of the wire mesh pallet is obtained, and welding is performed on the optimal welding point.

[0039] Since the spacing between the longitudinal wires is much larger than the diameter of the wires, the distance between the corners formed by the edge of a transverse wire and the adjacent longitudinal wires is relatively far. The vertical and horizontal distances between the corners of a transverse wire and the edge of a longitudinal wire are roughly equal to the diameter of the wire. The DBSCAN clustering algorithm clusters by identifying core points, and can achieve better clustering effects for categories with local density differences. Therefore, the coordinate positions of all overlapping corner points in the horizontal and vertical overlapping images are used as the input of the DBSCAN clustering algorithm, and the necessary parameter neighborhood radius of the algorithm is taken as the upper limit of the preset interval, that is, The minimum number is set to the third preset value, that is, 2, and multiple clusters of the DBSCAN clustering algorithm are obtained. The calculation of the DBSCAN clustering algorithm is a well-known technology, and the specific calculation process is not repeated here.

[0040] Through the proportional relationship between the size of the horizontal and vertical overlapping images and the actual size of the corresponding wire mesh pallet, the elements in each cluster in the horizontal and vertical overlapping images are used to determine the actual position of the wire mesh pallet through the proportional relationship, and the centroid of the actual position coordinates corresponding to all elements in each cluster is used as the optimal welding point position. The welding gun corrects the laser beam by adjusting the angle of the angle controller to ensure that it is accurately aligned with the optimal welding point of the wire. During the welding process, the laser beam monitors the position and status of the welding point in real time, and dynamically adjusts it through a closed-loop control system to ensure that the laser beam always remains aligned with the optimal welding point. In this way, the center of the welding gun can accurately contact the welding point. During the laser welding process, the centroid temperature of the overlapping area of ​​the wire is the highest, and the temperature gradually diffuses to the surrounding area, thereby improving the welding effect of the horizontal wire and the longitudinal wire.

[0041] The present application provides an automatic welding method for a wire mesh pallet based on visual analysis, the method comprising: analyzing the horizontal and vertical overlapping images of the wire mesh pallet during the welding process, and using a corner detection algorithm according to the characteristics of the overlap of two wires to obtain suspected welding corner points, which helps to narrow the range of finding overlapping corner points and improve the search efficiency; using the characteristics of the overlapping wires to obtain a welding positioning corner point set for each suspected welding corner point, which helps to analyze the remaining suspected welding corner points within a preset neighborhood of each suspected welding corner point; further performing pairwise analysis on all suspected welding corner points in the set to obtain the relative degree of overlapping corner points and the confirmation degree of overlapping corner points, and obtaining overlapping suspected corner points, the beneficial effect of which is to judge whether the overlapping corner points are caused by the overlap of wires according to the distance. Corner points, further narrowing the range of overlapping corner points; in each welding positioning corner point set, based on the relative degree of the overlapping corner points, the position distribution characteristics of each overlapping suspected corner point and the remaining elements are compared to obtain an angle monitoring set of each overlapping suspected corner point. The beneficial effect is that the suspected overlapping corner points are judged again in combination with the distance characteristics and angle distribution characteristics of the corner points of the overlapping wires, so that the overlapping corner points obtained subsequently are more accurate; finally, the clustering algorithm is used to help cluster the corner points generated by the overlap of the same longitudinal wire and a transverse wire into one category, so that the optimal welding point of the wire mesh pallet is more accurate, which can significantly improve the accuracy and automation level of wire mesh welding, enhance welding quality control, and thus improve the efficiency and reliability of welding operations.

[0042] The flowchart and block diagram in the accompanying drawings show the possible architecture, function and operation of the system, method and computer program product according to the embodiment of the present application. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of the code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. In some alternative implementations, the function marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two continuous boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, which can depend on the functions involved. In the description corresponding to the flowchart and the block diagram in the accompanying drawings, the operations or steps corresponding to different boxes can also occur in a different order from the order disclosed in the description, and sometimes there is no specific order between different operations or steps. For example, two continuous operations or steps can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, which can depend on the functions involved. Each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may be implemented by a dedicated hardware-based system that performs the specified functions or actions, or may be implemented by a combination of dedicated hardware and computer instructions.

[0043] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A method for automatic welding of wire mesh pallets based on visual analysis, characterized in that: The method comprises the following steps: Get each horizontal and vertical overlapping image during the welding process of the wire mesh pallet; A corner point detection algorithm is used for each horizontal and vertical overlapping image, and all the obtained corner points are used as suspected welding corner points. A welding positioning corner point set is determined based on all the suspected welding corner points in a preset neighborhood of each suspected welding corner point. Based on the distance measurement between any two elements in each welding positioning corner point set, determine the relative degree of overlapping corner points between the arbitrary two elements; analyze the relative degree of overlapping corner points between each element in each welding positioning corner point set and all other elements, obtain the overlapping corner point determination degree of each element, and obtain the overlapping suspected corner points; In each welding positioning corner point set, the position distribution characteristics of each overlapping suspected corner point and other elements are calculated based on the relative degree of the overlapping corner point to obtain the angle monitoring set of each overlapping suspected corner point; based on the element values ​​in the angle monitoring set, the overlapping suspected corner points are marked; according to the number of times each overlapping suspected corner point in the horizontal and vertical overlapping images is marked, combined with the threshold segmentation algorithm, the overlapping corner points of the horizontal and vertical overlapping images are obtained; A clustering algorithm is used for all overlapping corner points of the horizontally and vertically overlapping images, and based on the element distribution in the obtained clustering clusters, the optimal welding point of the wire mesh pallet is obtained, and welding is performed on the optimal welding point.

2. The automatic welding method for wire mesh pallets based on visual analysis as claimed in claim 1, characterized in that: The process of obtaining the relative degree of the overlapping corner points between any two elements is as follows: For each welding point location corner point set of suspected welding corner points, calculate the distance metric between any two elements; When the distance metric is within a preset interval, a first preset value is used as the relative degree of the overlapping corner points between the arbitrary two elements; otherwise, a second preset value is used as the relative degree of the overlapping corner points between the arbitrary two elements; wherein the first preset value is greater than the second preset value.

3. The automatic welding method for wire mesh pallets based on visual analysis as claimed in claim 1, characterized in that: The overlapping corner point determination degree of each element is obtained by accumulating the relative degrees of overlapping corner points between each element and all other elements in each welding positioning corner point set.

4. The automatic welding method for wire mesh pallets based on visual analysis as claimed in claim 1, characterized in that: The overlapping suspected corner points are specifically elements in the welding positioning corner point set of each suspected welding corner point whose overlapping corner point certainty is greater than or equal to a third preset value.

5. The automatic welding method for wire mesh pallets based on visual analysis as claimed in claim 2, characterized in that: The angle monitoring set of each overlapping suspected corner point is obtained as follows: For each suspected overlapping corner point in each welding corner point location set, screening the remaining elements in the welding corner point location set based on the relative degree of the overlapping corner point; Connect each overlapping suspected corner point with the selected elements, count the angles between any two connecting lines, and obtain the angle monitoring set of each overlapping suspected corner point.

6. The automatic welding method for wire mesh pallets based on visual analysis as claimed in claim 2, characterized in that: The remaining elements in the welding corner point positioning set are screened based on the overlapping corner point relative degree, specifically, the elements whose overlapping corner point relative degree with each overlapping suspected corner point is equal to the first preset value are screened.

7. The automatic welding method for wire mesh pallets based on visual analysis as claimed in claim 1, characterized in that: The marking of the overlapping suspected corner points is specifically: if there are elements in the angle monitoring set that are within a preset angle interval, the overlapping suspected corner points corresponding to the angle monitoring set are marked.

8. The automatic welding method for wire mesh pallets based on visual analysis as claimed in claim 1, characterized in that: The process of obtaining the overlapping corner points of the horizontally and vertically overlapping images is specifically as follows: The number of times all overlapping suspected corner points are marked is segmented using a threshold value to obtain a segmentation threshold value; and the overlapping suspected corner points whose number of times are marked is greater than or equal to the segmentation threshold value are recorded as overlapping corner points.

9. The automatic welding method for wire mesh pallets based on visual analysis as claimed in claim 2, characterized in that: The clustering algorithm is used for all overlapping corner points of the horizontally and vertically overlapping images, including: The coordinates of all overlapping corner points of the horizontally and vertically overlapping images are used as inputs of the DBSCAN clustering algorithm, the neighborhood radius is taken as the upper limit value of the preset interval; the minimum number is taken as the third preset value, and each cluster is obtained.

10. The automatic welding method for wire mesh pallets based on visual analysis as claimed in claim 1, characterized in that: The optimal welding point of the wire mesh pallet is determined by the centroid of the actual coordinates of all elements in each cluster of the horizontal and vertical overlapping images on the wire mesh pallet.

Images