A sheet metal welding defect intelligent detection method and device based on data feedback

Abstract

The present invention discloses a sheet metal welding defect intelligent detection method and device based on data feedback, which relates to the technical field of sheet metal welding defect detection, including: obtaining external welding defect parameters of the sheet metal to be tested through multi-directional images of the sheet metal to be tested, obtaining internal welding defect parameters of the sheet metal to be tested through X-ray images of the sheet metal to be tested, and obtaining a welding defect index by combining the external welding defect parameters and the internal welding defect parameters of the sheet metal to be tested, and then evaluating the sheet metal to be tested according to the welding defect index to obtain total defect information of the sheet metal to be tested, and then obtaining corresponding repair information of the sheet metal to be tested, so as to realize rapid judgment of the defect type of the sheet metal, formulate corresponding welding defect repair plan, and at the same time realize more accurate welding defect detection of special sheet metal parts whose welding areas are located at the angle or inner side of the sheet metal parts without being affected by light.

Description

Technical Field

[0001] The present invention relates to the technical field of sheet metal welding defect detection, and in particular to a sheet metal welding defect intelligent detection method and device based on data feedback. Background Art

[0002] With the development of industrialization and modernization, the quality requirements for sheet metal parts are getting higher and higher. Welding defects may lead to unqualified product quality, safety hazards and production accidents, so timely and effective defect detection is needed. The development of sheet metal welding defect detection technology is to meet quality requirements, improve the level of automation, and improve detection accuracy and speed, thereby ensuring the quality and reliability of the welding process.

[0003] The existing sheet metal welding defect detection method is to obtain the surface image of the sheet metal through an image acquisition device, and perform image processing on the surface image of the sheet metal to obtain a grayscale image, and then obtain the defect area in the surface image through the pixel points in the grayscale image, so as to judge the defect type of the sheet metal. However, after welding, the existing sheet metal parts have different shapes and uneven distribution of welding areas. Some special sheet metal parts even have welding areas located at the angle or inside of the sheet metal, which makes it difficult for the detection device to obtain the welding image. In addition, the detection device is easily affected by light when obtaining the image. Therefore, it is necessary to provide a sheet metal welding defect intelligent detection method and device based on data feedback to solve the above-mentioned problems. Summary of the invention

[0004] In order to solve the above technical problems, a method and device for intelligent detection of sheet metal welding defects based on data feedback are provided. The technical solution solves the problem proposed in the above background technology that the existing sheet metal parts have different shapes and uneven welding distribution after welding, and even some special sheet metal parts have welding areas located at the angle or inner side of the sheet metal parts, which makes it difficult for the detection device to obtain the welding image and is easily affected by light when obtaining the image.

[0005] In order to achieve the above purpose, the technical solution adopted by the present invention is:

[0006] An intelligent detection method for sheet metal welding defects based on data feedback, comprising:

[0007] Obtain multi-directional images of the sheet metal part to be tested;

[0008] Based on the importance of multi-directional images, a library of images to be tested is established, and the images to be tested in the library are sorted;

[0009] According to the image to be tested in the image library to be tested, the external welding defect parameters of the sheet metal part to be tested are obtained;

[0010] Use X-ray detection to obtain X-ray images of the sheet metal to be tested;

[0011] Based on the X-ray image of the sheet metal part to be tested, internal welding defect parameters of the sheet metal part to be tested are obtained;

[0012] Obtaining a welding defect index according to external welding defect parameters and internal welding defect parameters of the sheet metal to be tested;

[0013] According to the welding defect index, the sheet metal part to be tested is evaluated to obtain the total defect information of the sheet metal part to be tested;

[0014] Based on the total defect information of the sheet metal part to be tested, corresponding repair information of the sheet metal part to be tested is obtained;

[0015] According to the corresponding repair information of the sheet metal part to be tested, the welding defects of the sheet metal part to be tested are repaired.

[0016] In an optional embodiment, based on the importance of the multi-directional images, a library of images to be tested is established, and the images to be tested in the library are sorted, specifically including:

[0017] Based on Canny edge detection, the position of the sheet metal part to be tested in the multi-directional image is determined, and the image of the sheet metal part to be tested is separated from the multi-directional image to obtain an image library of the sheet metal part to be tested;

[0018] Determine the position of the welding area of ​​the sheet metal part to be tested in the image of the sheet metal part to be tested, separate the image of the welding area from the image of the sheet metal part to be tested, and obtain a welding area image library;

[0019] Taking the edge position of the welding area image as the maximum size of the welding area, obtaining the welding image size parameters of each welding area image;

[0020] Based on the images of the sheet metal parts to be tested in the image library of the sheet metal parts to be tested, obtaining the maximum size parameter of each image of the sheet metal parts to be tested;

[0021] A blank image is constructed with the maximum size of the welding area, and each welding area image is placed at the center of a blank image, and the background filling color of the blank image outside each welding area image is black, the image to be tested is obtained, and a library of images to be tested is established;

[0022] The ratio of the welding image size parameter of each welding image region image to the maximum size parameter of the image of the sheet metal part to be tested is used as an evaluation index of the importance of the multi-directional image to obtain the important image index of each image to be tested;

[0023] The images to be tested in the image library to be tested are sorted according to the size of the important image index of each image to be tested.

[0024] In an optional embodiment, determining the position of the welding area of ​​the sheet metal part to be tested in each image in the sheet metal part image library to obtain the welding area image library specifically includes:

[0025] Get historical sheet metal image library;

[0026] Based on the visibility of the welding area of ​​sheet metal parts in the historical sheet metal parts image library, images without blur and overexposure are screened out from the historical sheet metal parts image library to obtain a standard welding area image library;

[0027] According to the standard welding area image library, the resolution information of each standard welding area image is obtained;

[0028] Based on the resolution information of each standard welding area image, a resolution index of each standard welding area image is obtained;

[0029] According to the size of the resolution index of each standard welding area image, the standard welding area images in the standard welding area image library are sorted, and the standard welding area image with the first sorting is extracted to obtain a welding feature image;

[0030] Grayscale processing is performed on the welding feature image, and edge enhancement is performed on the grayscale image to obtain a welding feature template;

[0031] Using the image processing library to load the welding feature template and the image of the sheet metal part to be tested in the image library of the sheet metal part to be tested;

[0032] Based on the normalized correlation matching method, every possible position in the image of the sheet metal part to be tested is traversed with the preset image size of the welding feature template, and its correlation with the welding feature template is calculated, and the normalized correlation matching value of each position and the corresponding position coordinates are recorded;

[0033] The position with the highest normalized correlation matching value is taken as the position with the maximum correlation value, and half of the maximum normalized correlation matching value is taken as the correlation matching threshold;

[0034] Based on the correlation matching threshold, each possible position in the image of the sheet metal part to be tested is screened, and the position above the correlation matching threshold is marked in the image of the sheet metal part to be tested;

[0035] The marked positions in the image of the sheet metal part to be tested are connected, and the part of the image of the sheet metal part to be tested with the largest connected area is taken as the welding area image to obtain the welding area image.

[0036] In an optional embodiment, obtaining the external welding defect parameters of the sheet metal part to be tested according to the image to be tested in the image library to be tested specifically includes:

[0037] Performing binarization processing on the image to be tested in the image library to be tested, and obtaining the binary image library to be tested;

[0038] Based on the binary images to be tested in the binary image library to be tested, respectively record the external welding defect type information corresponding to each binary image to be tested, wherein the external welding defect type information includes a first pore area, a first burn-through area, a first slag inclusion area, and a crack area;

[0039] According to the external welding defect type information, the external welding defect parameters of the sheet metal part to be tested are obtained;

[0040] The calculation steps of the first pore area, the first burn-through area, the first slag inclusion area and the crack area are as follows:

[0041] S401: Acquire one of the binary images to be tested;

[0042] S402: Use the cv.findContours function in the image processing library to perform contour detection on the binary image to be tested, and obtain the boundary of each defect area, wherein the white area in the shape of a circle or an ellipse corresponds to the first pore area, the white area in the shape of a line or a dot along the weld contour corresponds to the first burn-through area, the black area in the shape of an irregular shape corresponds to the first slag inclusion area, and the white area in the shape of a thin line or a fracture corresponds to the crack area;

[0043] S403: Calculating the sum of the number of pixels of the defect areas corresponding to the first pore area, the first burn-through area, the first slag inclusion area and the crack area respectively, namely, the first pore area, the first burn-through area, the first slag inclusion area and the crack area;

[0044] The calculation formula of the external welding defect parameter is:

[0045]

[0046] Where W represents the external welding defect parameter of the sheet metal to be tested, ω i represents the importance weight of the i-th binary image to be tested, represents the weight of the first pore area in the i-th binary image to be tested, represents the weight of the first burn-through area in the i-th binary image to be tested, represents the weight of the first slag inclusion area in the i-th binary image to be tested, represents the weight of the crack area in the i-th image to be tested, represents the first pore area in the i-th binary image to be tested, represents the first burn-through area in the i-th binary image to be tested, represents the first slag inclusion area in the i-th binary image to be tested, represents the crack area in the i-th binary image to be tested, A total represents the total area of ​​the sheet metal to be tested, and j represents the total number of binary images to be tested.

[0047] In an optional embodiment, based on the X-ray image of the sheet metal part to be tested, obtaining the internal welding defect parameters of the sheet metal part to be tested specifically includes:

[0048] Obtain a background image without the sheet metal part to be tested as a background image template;

[0049] Based on grayscale standardization, the X-ray image of the sheet metal to be tested and the background image template are preprocessed so that the X-ray image of the sheet metal to be tested and the background image template have the same size and grayscale range;

[0050] The X-ray image of the sheet metal to be tested and the background image template are grayscale standardized, and then each pixel value in the X-ray image is subtracted from the corresponding pixel value in the background image to obtain the image to be segmented;

[0051] Obtaining the segmentation threshold based on the average gray value of the X-ray image and the background image template;

[0052] Based on the segmentation threshold, threshold processing is performed on the image to be segmented to obtain a binary image to be segmented;

[0053] Based on region growth, fill and connect the partial images of the sheet metal to be tested in the binary image to be segmented to obtain a full-range image of the sheet metal to be tested;

[0054] Based on the omnidirectional image of the sheet metal part to be tested, the corresponding internal welding defect type information in the omnidirectional image is recorded respectively, wherein the internal welding defect type information includes a second pore area, a second slag inclusion area, a second crack area, an unmelted region area, an impurity area, and an uneven structure area;

[0055] According to the internal welding defect type information, the internal defect parameters of the sheet metal part to be tested are obtained;

[0056] The calculation steps of the second pore area, the second slag inclusion area, the second crack area, the unmelted region area, the impurity area and the non-uniform structure area are as follows:

[0057] S501: Acquire a full-range image of the sheet metal part to be tested;

[0058] S502: Use the cv.findContours function in the image processing library to perform contour detection on the omnidirectional image to obtain the boundary of each defect area, wherein the dark spots and dark spots in the form of circles and ellipses correspond to the second pore area, the irregular bright spots and bright spots correspond to the second slag inclusion area, the thinner dark lines and dark shadows correspond to the second crack area, the irregular and lighter areas around the welding area correspond to the unmelted area, the dark spots in the form of obvious irregular shapes correspond to the impurity area, and the areas in the form of uneven density and grayscale changes correspond to the uneven structure area;

[0059] S503: respectively calculating the sum of the number of pixels of the defective areas corresponding to the second pore area, the second slag inclusion area, the second crack area, the unmelted region area, the impurity area and the non-uniform structure area, that is, the second pore area, the second slag inclusion area, the second crack area, the unmelted region area, the impurity area and the non-uniform structure area;

[0060] The calculation formula of the internal welding defect parameter is:

[0061]

[0062] Where N represents the internal defect parameter of the sheet metal to be tested, B n Represents the area of ​​various defects, γ n represents the weight factor of each defect, and m represents the total number of defect types.

[0063] In an optional embodiment, according to the welding defect index, the sheet metal part to be tested is evaluated to obtain the total defect information of the sheet metal part to be tested, which specifically includes:

[0064] Based on the external welding defect type information, external defect position information and external defect size information of each defect are obtained;

[0065] Based on the external defect position information and the external defect size information, the external defect influencing factor is obtained;

[0066] Based on the internal welding defect type information, internal defect position information and internal defect size information of each defect are obtained;

[0067] Based on the internal defect position information and the internal defect size information, the internal defect influencing factor is obtained;

[0068] 1% of the average value of the external defect influencing factor and the internal defect influencing factor is used as the welding defect index threshold of the sheet metal to be tested;

[0069] Obtaining a welding defect index according to external welding defect parameters and internal welding defect parameters;

[0070] Determine whether the welding defect index exceeds the welding defect index threshold. If so, perform welding defect repair on the sheet metal to be tested. If not, continue to obtain the multi-directional image and X-ray image of the next sheet metal to be tested.

[0071] Wherein, the calculation formula of the external defect influencing factor is:

[0072]

[0073] In the formula, I out represents the external defect influencing factor, θα represents the weight of the αth external welding defect position parameter, C α represents the location parameter of the αth external welding defect, represents the weight of the αth external defect size parameter, D α represents the αth external defect size parameter, and β represents the total number of external defect types;

[0074] The calculation formula of the internal defect influence factor is:

[0075]

[0076] In the formula, I in represents the internal defect influence factor, ρ μ represents the weight of the μth internal welding defect location parameter, E μ represents the μth internal welding defect location parameter, σ μ represents the weight of the μth internal defect size parameter, F μ represents the μth internal defect size parameter, and π represents the total number of internal defect types;

[0077] The calculation formula of the welding defect index is:

[0078] G z =δW+εN;

[0079] In the formula, G z represents the welding defect index of the zth sheet metal part to be tested, W represents the external welding defect parameter, δ represents the weight of the external welding defect parameter, N represents the internal welding defect parameter, and ε represents the weight of the internal welding defect parameter.

[0080] Furthermore, a sheet metal welding defect intelligent detection device based on data feedback is proposed to implement the above detection method, including:

[0081] A data acquisition module, the data acquisition module is used to acquire multi-directional images of the sheet metal part to be tested, X-ray images of the sheet metal part to be tested, historical sheet metal part images and background image templates;

[0082] A data processing module, the data processing module is used to determine the position of the sheet metal to be tested in the multi-directional image based on Canny edge detection, and separate the image of the sheet metal to be tested from the multi-directional image, obtain the image library of the sheet metal to be tested, and pre-process the X-ray image and background image template of the sheet metal to be tested based on grayscale standardization, so that the X-ray image and the background image template of the sheet metal to be tested have the same size and grayscale range, and grayscale standardize the X-ray image and the background image template of the sheet metal to be tested, and then subtract the corresponding pixel value in the background image from each pixel value in the X-ray image to obtain the image to be segmented, which is used to calculate external welding defect parameters, internal welding defect parameters, external defect influencing factors, internal defect influencing factors, welding defect index and welding defect index threshold, and is used to determine whether the welding defect index exceeds the welding defect index threshold, and determine whether the sheet metal to be tested needs welding defect repair;

[0083] A data transmission module, which is used to transmit the images and data in the data acquisition module to the data processing module and the display module;

[0084] The display module is used to display the multi-directional images of the sheet metal part to be tested and the image processing process and to display the X-ray image of the sheet metal part to be tested and the image processing process.

[0085] In an optional embodiment, the data acquisition module specifically includes:

[0086] A first acquisition unit, the first acquisition unit is used to acquire a multi-directional image of the sheet metal part to be tested and an X-ray image of the sheet metal part to be tested;

[0087] The second acquisition unit is used to acquire a historical sheet metal image library and a background image template.

[0088] In an optional embodiment, the data processing module specifically includes:

[0089] An image processing unit, the image processing unit is used to determine the position of the sheet metal to be tested in the multi-directional image based on Canny edge detection, separate the image of the sheet metal to be tested from the multi-directional image, obtain an image library of the sheet metal to be tested, pre-process the X-ray image of the sheet metal to be tested and the background image template based on grayscale standardization, so that the X-ray image of the sheet metal to be tested and the background image template have the same size and grayscale range, grayscale standardize the X-ray image of the sheet metal to be tested and the background image template, and then subtract the corresponding pixel value in the background image from each pixel value in the X-ray image to obtain an image to be segmented;

[0090] A calculation unit, the calculation unit is used to calculate external welding defect parameters, internal welding defect parameters, external defect influencing factors, internal defect influencing factors, welding defect index and welding defect index threshold;

[0091] The judging unit is used to judge whether the welding defect index exceeds the welding defect index threshold value, and judge whether the sheet metal part to be tested needs welding defect repair.

[0092] In an optional embodiment, the display module specifically includes:

[0093] A first display unit, the first display unit is used to display the multi-directional images of the sheet metal part to be tested and the image processing process;

[0094] A second display unit, wherein the second display unit is used to display the X-ray image of the sheet metal part to be tested and the image processing process.

[0095] Compared with the prior art, the present invention has the following beneficial effects:

[0096] This scheme proposes an intelligent detection method for sheet metal welding defects based on data feedback. Based on the importance of multi-directional images, a library of images to be tested is established, and the images to be tested in the library are binarized to obtain the binary image library to be tested, and then the external welding defect parameters of the sheet metal to be tested are obtained. By adopting X-ray detection, the X-ray image of the sheet metal to be tested is obtained, and the X-ray image of the sheet metal to be tested and the background image template are gray-scale standardized. Then, each pixel value in the X-ray image is subtracted from the corresponding pixel value in the background image to obtain the image to be segmented, and then the internal welding defect parameters of the sheet metal to be tested are obtained. The welding defect index is obtained through the external welding defect parameters and internal welding defect parameters of the sheet metal to be tested. Then, according to the welding defect index, the sheet metal to be tested is evaluated, and the total defect information of the sheet metal to be tested is obtained, so as to realize rapid judgment of the defect type of the sheet metal, formulate corresponding welding defect repair plan, and at the same time, more accurate welding defect detection is realized for special sheet metal parts whose welding areas are located at the angle or inside of the sheet metal parts, which will not be affected by light. BRIEF DESCRIPTION OF THE DRAWINGS

[0097] Figure 1 A flow chart of a sheet metal welding defect intelligent detection method based on data feedback proposed by the present invention;

[0098] Figure 2 is a flow chart of obtaining the image library to be tested in the present invention;

[0099] Figure 3 is an evaluation flow chart of the sheet metal part to be tested in the present invention;

[0100] Figure 4This is a system framework diagram of a sheet metal welding defect intelligent detection device based on data feedback proposed in the present invention. DETAILED DESCRIPTION

[0101] The following description is used to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are only examples, and those skilled in the art may think of other obvious variations.

[0102] Reference Figure 1 - Figure 4 As shown, a sheet metal welding defect intelligent detection method based on data feedback includes:

[0103] Obtain multi-directional images of the sheet metal part to be tested;

[0104] Based on the importance of multi-directional images, a library of images to be tested is established, and the images to be tested in the library are sorted;

[0105] According to the image to be tested in the image library to be tested, the external welding defect parameters of the sheet metal part to be tested are obtained;

[0106] Use X-ray detection to obtain X-ray images of the sheet metal to be tested;

[0107] Based on the X-ray image of the sheet metal part to be tested, internal welding defect parameters of the sheet metal part to be tested are obtained;

[0108] Obtaining a welding defect index according to external welding defect parameters and internal welding defect parameters of the sheet metal to be tested;

[0109] According to the welding defect index, the sheet metal part to be tested is evaluated to obtain the total defect information of the sheet metal part to be tested;

[0110] Based on the total defect information of the sheet metal part to be tested, corresponding repair information of the sheet metal part to be tested is obtained;

[0111] According to the corresponding repair information of the sheet metal part to be tested, the welding defects of the sheet metal part to be tested are repaired.

[0112] Specifically, the multi-directional images of the sheet metal to be tested are photos or images of the sheet metal to be tested at different angles or directions, aiming to fully display the various parts and welding features of the sheet metal to be tested. These images may include shots from the front, side, back, top view and other angles. When collecting multi-directional images of the sheet metal to be tested, high-definition acquisition equipment can be used, and it is necessary to ensure that the distance between the high-definition acquisition equipment and the sheet metal to be tested is the same when collecting images. For some special sheet metal parts, the welding area may exist inside and at the angle. X-ray detection equipment can be used to perform non-destructive testing on the sheet metal to be tested and collect X-ray images of the sheet metal to be tested.

[0113] It can be understood that the external welding defect parameters of the sheet metal to be tested are used to describe the comprehensive indicators of the outer surface of the sheet metal to be tested and some welding defects that can be observed from the outside, which are calculated using multi-directional images after multi-directional image acquisition of welding defects that are easy to observe from the outside of the sheet metal to be tested, while the internal welding defect parameters of the sheet metal to be tested are comprehensive indicators of welding defects that are difficult to observe from the outside of the sheet metal to be tested. For example, a box-shaped sheet metal part may be composed of several panels welded together, and these panels form a closed shell on the outside. In this case, welding is usually completed inside the box to ensure that the external surface remains flat and seamless, so as to meet the requirements of appearance and function. At this time, it is difficult for ordinary testing equipment to detect the internal welding area of ​​the box-shaped sheet metal part.

[0114] Furthermore, based on the importance of the multi-directional images, a library of images to be tested is established, and the images to be tested in the library are sorted, specifically including:

[0115] Based on Canny edge detection, the position of the sheet metal part to be tested in the multi-directional image is determined, and the image of the sheet metal part to be tested is separated from the multi-directional image to obtain an image library of the sheet metal part to be tested;

[0116] Determine the position of the welding area of ​​the sheet metal part to be tested in the image of the sheet metal part to be tested, separate the image of the welding area from the image of the sheet metal part to be tested, and obtain a welding area image library;

[0117] Taking the edge position of the welding area image as the maximum size of the welding area, obtaining the welding image size parameters of each welding area image;

[0118] Based on the images of the sheet metal parts to be tested in the image library of the sheet metal parts to be tested, obtaining the maximum size parameter of each image of the sheet metal parts to be tested;

[0119] A blank image is constructed with the maximum size of the welding area, and each welding area image is placed at the center of a blank image, and the background filling color of the blank image outside each welding area image is black, the image to be tested is obtained, and a library of images to be tested is established;

[0120] The ratio of the welding image size parameter of each welding image region image to the maximum size parameter of the image of the sheet metal part to be tested is used as an evaluation index of the importance of the multi-directional image to obtain the important image index of each image to be tested;

[0121] The images to be tested in the image library to be tested are sorted according to the size of the important image index of each image to be tested.

[0122] Further, determining the position of the welding area of ​​the sheet metal part to be tested in each image in the sheet metal part image library to obtain the welding area image library specifically includes:

[0123] Get historical sheet metal image library;

[0124] Based on the visibility of the welding area of ​​sheet metal parts in the historical sheet metal parts image library, images without blur and overexposure are screened out from the historical sheet metal parts image library to obtain a standard welding area image library;

[0125] According to the standard welding area image library, the resolution information of each standard welding area image is obtained;

[0126] Based on the resolution information of each standard welding area image, a resolution index of each standard welding area image is obtained;

[0127] According to the size of the resolution index of each standard welding area image, the standard welding area images in the standard welding area image library are sorted, and the standard welding area image with the first sorting is extracted to obtain a welding feature image;

[0128] Grayscale processing is performed on the welding feature image, and edge enhancement is performed on the grayscale image to obtain a welding feature template;

[0129] Using the image processing library to load the welding feature template and the image of the sheet metal part to be tested in the image library of the sheet metal part to be tested;

[0130] Based on the normalized correlation matching method, every possible position in the image of the sheet metal part to be tested is traversed with the preset image size of the welding feature template, and its correlation with the welding feature template is calculated, and the normalized correlation matching value of each position and the corresponding position coordinates are recorded;

[0131] The position with the highest normalized correlation matching value is taken as the position with the maximum correlation value, and half of the maximum normalized correlation matching value is taken as the correlation matching threshold;

[0132] Based on the correlation matching threshold, each possible position in the image of the sheet metal part to be tested is screened, and the position above the correlation matching threshold is marked in the image of the sheet metal part to be tested;

[0133] The marked positions in the image of the sheet metal part to be tested are connected, and the part of the image of the sheet metal part to be tested with the largest connected area is taken as the welding area image to obtain the welding area image.

[0134] Specifically, based on Canny edge detection, the position of the sheet metal to be tested in the multi-directional image is determined, and the image of the sheet metal to be tested is separated from the multi-directional image, and the image library of the sheet metal to be tested is obtained. The edge detection algorithm is used in the multi-directional image to extract the contour or edge of the sheet metal. Then, the template matching technology can be used to compare a predefined sheet metal template with the corresponding area in the image to find the best matching position, so as to determine the position of the sheet metal to be tested in the multi-directional image. The edge position of the welding area image is used as the maximum size of the welding area, and the welding image size parameters of each welding area image are obtained. Each welding area image has different size parameters, and each sheet metal image to be tested also has different size parameters. Therefore, each image in the multi-directional image has a different degree of importance, and the degree of importance is numerically equal to the important image index of the corresponding image to be tested. For example, if the welding area displayed in one of the multi-directional images accounts for a large proportion, while the image displayed by the sheet metal to be tested collected from this angle or orientation accounts for a small proportion, its importance will be greater and its ranking in the image library to be tested will be higher. If the welding area displayed in one of the multi-directional images accounts for a small proportion or is zero, its importance will be smaller or even zero. Based on the normalized correlation matching method, every possible position in the image of the sheet metal to be tested is traversed with the preset image size of the welding feature template. The welding feature template can be used as a sliding window, one side of which is aligned with one side of the image of the sheet metal to be tested, and all positions in the image of the sheet metal to be tested are traversed in a sliding manner, and the distance of each sliding is a unit pixel.

[0135] Furthermore, according to the image to be tested in the image library to be tested, the external welding defect parameters of the sheet metal part to be tested are obtained, which specifically includes:

[0136] Performing binarization processing on the image to be tested in the image library to be tested, and obtaining the binary image library to be tested;

[0137] Based on the binary images to be tested in the binary image library to be tested, the external welding defect type information corresponding to each binary image to be tested is recorded respectively, and the external welding defect type information includes a first pore area, a first burn-through area, a first slag inclusion area and a crack area;

[0138] According to the external welding defect type information, the external welding defect parameters of the sheet metal part to be tested are obtained;

[0139] The calculation steps of the first pore area, the first burn-through area, the first slag inclusion area and the crack area are as follows:

[0140] S401: Acquire one of the binary images to be tested;

[0141] S402: Use the cv.findContours function in the image processing library to perform contour detection on the binary image to be tested, and obtain the boundary of each defect area, wherein the white area in the shape of a circle or an ellipse corresponds to the first pore area, the white area in the shape of a line or a dot along the weld contour corresponds to the first burn-through area, the black area in the shape of an irregular shape corresponds to the first slag inclusion area, and the white area in the shape of a thin line or a fracture corresponds to the crack area;

[0142] S403: Calculating the sum of the number of pixels of the defect areas corresponding to the first pore area, the first burn-through area, the first slag inclusion area and the crack area respectively, namely, the first pore area, the first burn-through area, the first slag inclusion area and the crack area;

[0143] The calculation formula for external welding defect parameters is:

[0144]

[0145] Where W represents the external welding defect parameter of the sheet metal to be tested, ω i represents the importance weight of the i-th binary image to be tested, represents the weight of the first pore area in the i-th binary image to be tested, represents the weight of the first burn-through area in the i-th binary image to be tested, represents the weight of the first slag inclusion area in the i-th binary image to be tested, represents the weight of the crack area in the i-th image to be tested, represents the first pore area in the i-th binary image to be tested, A Bi represents the first burn-through area in the i-th binary image to be tested, A Si represents the first slag area in the i-th binary image to be tested, A Ci represents the crack area in the i-th binary image to be tested, A total represents the total area of ​​the sheet metal to be tested, and j represents the total number of binary images to be tested.

[0146] Furthermore, based on the X-ray image of the sheet metal part to be tested, the internal welding defect parameters of the sheet metal part to be tested are obtained, specifically including:

[0147] Obtain a background image without the sheet metal part to be tested as a background image template;

[0148] Based on grayscale standardization, the X-ray image of the sheet metal to be tested and the background image template are preprocessed so that the X-ray image of the sheet metal to be tested and the background image template have the same size and grayscale range;

[0149] The X-ray image of the sheet metal to be tested and the background image template are grayscale standardized, and then each pixel value in the X-ray image is subtracted from the corresponding pixel value in the background image to obtain the image to be segmented;

[0150] Obtaining the segmentation threshold based on the average gray value of the X-ray image and the background image template;

[0151] Based on the segmentation threshold, threshold processing is performed on the image to be segmented to obtain a binary image to be segmented;

[0152] Based on region growth, fill and connect the partial images of the sheet metal to be tested in the binary image to be segmented to obtain a full-range image of the sheet metal to be tested;

[0153] Based on the omnidirectional image of the sheet metal part to be tested, the corresponding internal welding defect type information in the omnidirectional image is recorded respectively, and the internal welding defect type information includes the second pore area, the second slag inclusion area, the second crack area, the unmelted region area, the impurity area and the uneven structure area;

[0154] According to the internal welding defect type information, the internal defect parameters of the sheet metal part to be tested are obtained;

[0155] The calculation steps of the second pore area, the second slag inclusion area, the second crack area, the unmelted region area, the impurity area and the non-uniform structure area are as follows:

[0156] S501: Acquire a full-range image of the sheet metal part to be tested;

[0157] S502: Use the cv.findContours function in the image processing library to perform contour detection on the omnidirectional image to obtain the boundary of each defect area, wherein the dark spots and dark spots in the form of circles and ellipses correspond to the second pore area, the irregular bright spots and bright spots correspond to the second slag inclusion area, the thinner dark lines and dark shadows correspond to the second crack area, the irregular and lighter areas around the welding area correspond to the unmelted area, the dark spots in the form of obvious irregular shapes correspond to the impurity area, and the areas in the form of uneven density and grayscale changes correspond to the uneven structure area;

[0158] S503: respectively calculating the sum of the number of pixels of the defective areas corresponding to the second pore area, the second slag inclusion area, the second crack area, the unmelted region area, the impurity area and the non-uniform structure area, that is, the second pore area, the second slag inclusion area, the second crack area, the unmelted region area, the impurity area and the non-uniform structure area;

[0159] The calculation formula for internal welding defect parameters is:

[0160]

[0161] Where N represents the internal defect parameter of the sheet metal to be tested, B n Represents the area of ​​various defects, γ n represents the weight factor of each defect, and m represents the total number of defect types.

[0162] Furthermore, according to the welding defect index, the sheet metal part to be tested is evaluated to obtain the total defect information of the sheet metal part to be tested, including:

[0163] Based on the external welding defect type information, external defect position information and external defect size information of each defect are obtained;

[0164] Based on the external defect position information and the external defect size information, the external defect influencing factor is obtained;

[0165] Based on the internal welding defect type information, internal defect position information and internal defect size information of each defect are obtained;

[0166] Based on the internal defect position information and the internal defect size information, the internal defect influencing factor is obtained;

[0167] 1% of the average value of the external defect influencing factor and the internal defect influencing factor is used as the welding defect index threshold of the sheet metal to be tested;

[0168] Obtaining a welding defect index according to external welding defect parameters and internal welding defect parameters;

[0169] Determine whether the welding defect index exceeds the welding defect index threshold. If so, perform welding defect repair on the sheet metal to be tested. If not, continue to obtain the multi-directional image and X-ray image of the next sheet metal to be tested.

[0170] Among them, the calculation formula of the external defect influence factor is:

[0171]

[0172] In the formula, I out represents the external defect influence factor, θ α represents the weight of the αth external welding defect location parameter, C α represents the location parameter of the αth external welding defect, represents the weight of the αth external defect size parameter, D α represents the αth external defect size parameter, and β represents the total number of external defect types;

[0173] The calculation formula of internal defect influence factor is:

[0174]

[0175] In the formula, I in represents the internal defect influence factor, ρ μ represents the weight of the μth internal welding defect location parameter, E μ represents the μth internal welding defect location parameter, σ μ represents the weight of the μth internal defect size parameter, F μ represents the μth internal defect size parameter, and π represents the total number of internal defect types;

[0176] The calculation formula of welding defect index is:

[0177] G z =δW+εN;

[0178] In the formula, G z represents the welding defect index of the zth sheet metal part to be tested, W represents the external welding defect parameter, δ represents the weight of the external welding defect parameter, N represents the internal welding defect parameter, and ε represents the weight of the internal welding defect parameter.

[0179] Furthermore, a sheet metal welding defect intelligent detection device based on data feedback is proposed to implement the above detection method, including:

[0180] A data acquisition module, the data acquisition module is used to acquire multi-directional images of the sheet metal part to be tested, X-ray images of the sheet metal part to be tested, historical sheet metal part images and background image templates;

[0181] A data processing module, which is used to determine the position of the sheet metal to be tested in the multi-directional image based on Canny edge detection, separate the image of the sheet metal to be tested from the multi-directional image, obtain the image library of the sheet metal to be tested, pre-process the X-ray image and the background image template of the sheet metal to be tested based on grayscale standardization, so that the X-ray image and the background image template of the sheet metal to be tested have the same size and grayscale range, grayscale standardize the X-ray image and the background image template of the sheet metal to be tested, and then subtract the corresponding pixel value in the background image from each pixel value in the X-ray image to obtain the image to be segmented, which is used to calculate external welding defect parameters, internal welding defect parameters, external defect influencing factors, internal defect influencing factors, welding defect index and welding defect index threshold, and is used to determine whether the welding defect index exceeds the welding defect index threshold, and determine whether the sheet metal to be tested needs welding defect repair;

[0182] A data transmission module, the data transmission module is used to transmit the image and data in the data acquisition module to the data processing module and the display module;

[0183] The display module is used to display the multi-directional images of the sheet metal part to be tested and the image processing process and to display the X-ray images of the sheet metal part to be tested and the image processing process.

[0184] Furthermore, the data acquisition module specifically includes:

[0185] A first acquisition unit, the first acquisition unit is used to acquire a multi-directional image of the sheet metal part to be tested and an X-ray image of the sheet metal part to be tested;

[0186] The second acquisition unit is used to acquire a historical sheet metal image library and a background image template.

[0187] Furthermore, the data processing module specifically includes:

[0188] An image processing unit, the image processing unit is used to determine the position of the sheet metal to be tested in the multi-directional image based on Canny edge detection, separate the image of the sheet metal to be tested from the multi-directional image, obtain the image library of the sheet metal to be tested, pre-process the X-ray image and the background image template of the sheet metal to be tested based on grayscale standardization, so that the X-ray image and the background image template of the sheet metal to be tested have the same size and grayscale range, grayscale standardize the X-ray image and the background image template of the sheet metal to be tested, and then subtract the corresponding pixel value in the background image from each pixel value in the X-ray image to obtain the image to be segmented;

[0189] A calculation unit, the calculation unit is used to calculate external welding defect parameters, internal welding defect parameters, external defect influencing factors, internal defect influencing factors, welding defect index and welding defect index threshold;

[0190] The judgment unit is used to judge whether the welding defect index exceeds the welding defect index threshold value and whether the sheet metal part to be tested needs welding defect repair.

[0191] Furthermore, the display module specifically includes:

[0192] A first display unit, the first display unit is used to display the multi-directional image of the sheet metal part to be tested and the image processing process;

[0193] The second display unit is used to display the X-ray image of the sheet metal part to be tested and the image processing process.

[0194] In summary, the advantages of the present invention are: by collecting multi-directional images of the sheet metal to be tested, based on the importance of the multi-directional images, a library of images to be tested is established, and the images to be tested in the library of images to be tested are binarized to obtain the library of binary images to be tested, and then the external welding defect parameters of the sheet metal to be tested are obtained, and by adopting X-ray detection, the X-ray image of the sheet metal to be tested is obtained, the X-ray image of the sheet metal to be tested and the background image template are gray-scale standardized, and then each pixel value in the X-ray image is subtracted from the corresponding pixel value in the background image to obtain the image to be segmented, and then the internal welding defect parameters of the sheet metal to be tested are obtained, and the welding defect index is obtained through the external welding defect parameters and the internal welding defect parameters of the sheet metal to be tested. Then, according to the welding defect index, the sheet metal to be tested is evaluated, and the total defect information of the sheet metal to be tested is obtained, so as to realize rapid judgment of the defect type of the sheet metal, formulate a corresponding welding defect repair plan, and at the same time, realize more accurate welding defect detection of special sheet metal parts whose welding areas are located at the angle or inner side of the sheet metal parts, which will not be affected by light.

[0195] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions only describe the principles of the present invention. The present invention may be subject to various changes and improvements without departing from the spirit and scope of the present invention. These changes and improvements fall within the scope of the present invention. The scope of protection claimed by the present invention is defined by the attached claims and their equivalents.

Claims

1. A sheet metal welding defect intelligent detection method based on data feedback, characterized in that: include: Obtain multi-directional images of the sheet metal part to be tested; Based on the importance of multi-directional images, a library of images to be tested is established, and the images to be tested in the library are sorted; According to the image to be tested in the image library to be tested, the external welding defect parameters of the sheet metal part to be tested are obtained; Use X-ray detection to obtain X-ray images of the sheet metal to be tested; Based on the X-ray image of the sheet metal part to be tested, internal welding defect parameters of the sheet metal part to be tested are obtained; Obtaining a welding defect index according to external welding defect parameters and internal welding defect parameters of the sheet metal to be tested; According to the welding defect index, the sheet metal parts to be tested are evaluated to obtain the total defect information of the sheet metal parts to be tested; Based on the total defect information of the sheet metal part to be tested, corresponding repair information of the sheet metal part to be tested is obtained; Repair welding defects of the sheet metal parts to be tested according to the corresponding repair information of the sheet metal parts to be tested; Based on the importance of multi-directional images, a library of images to be tested is established, and the images to be tested in the library are sorted, including: Based on Canny edge detection, the position of the sheet metal part to be tested in the multi-directional image is determined, and the image of the sheet metal part to be tested is separated from the multi-directional image to obtain an image library of the sheet metal part to be tested; Determine the position of the welding area of ​​the sheet metal part to be tested in the image of the sheet metal part to be tested, separate the image of the welding area from the image of the sheet metal part to be tested, and obtain a welding area image library; Taking the edge position of the welding area image as the maximum size of the welding area, obtaining the welding image size parameters of each welding area image; Based on the images of the sheet metal parts to be tested in the image library of the sheet metal parts to be tested, obtaining the maximum size parameter of each image of the sheet metal parts to be tested; A blank image is constructed with the maximum size of the welding area, and each welding area image is placed at the center of a blank image, and the background filling color of the blank image outside each welding area image is black, the image to be tested is obtained, and a library of images to be tested is established; The ratio of the welding image size parameter of each welding image region image to the maximum size parameter of the image of the sheet metal part to be tested is used as an evaluation index of the importance of the multi-directional image to obtain the important image index of each image to be tested; The images to be tested in the image library to be tested are sorted according to the size of the important image index of each image to be tested; According to the image to be tested in the image library to be tested, the external welding defect parameters of the sheet metal part to be tested are obtained, including: Performing binarization processing on the image to be tested in the image library to be tested, and obtaining the binary image library to be tested; Based on the binary images to be tested in the binary image library to be tested, respectively record the external welding defect type information corresponding to each binary image to be tested, wherein the external welding defect type information includes a first pore area, a first burn-through area, a first slag inclusion area, and a crack area; According to the external welding defect type information, the external welding defect parameters of the sheet metal part to be tested are obtained; The calculation steps of the first pore area, the first burn-through area, the first slag inclusion area and the crack area are as follows: S401: Acquire one of the binary images to be tested; S402: Use the cv.findContours function in the image processing library to perform contour detection on the binary image to be tested, and obtain the boundary of each defect area, wherein the white area in the shape of a circle or an ellipse corresponds to the first pore area, the white area in the shape of a line or a dot along the weld contour corresponds to the first burn-through area, the black area in the shape of an irregular shape corresponds to the first slag inclusion area, and the white area in the shape of a thin line or a fracture corresponds to the crack area; S403: Calculating the sum of the number of pixels of the defect areas corresponding to the first pore area, the first burn-through area, the first slag inclusion area and the crack area respectively, namely, the first pore area, the first burn-through area, the first slag inclusion area and the crack area; The calculation formula of the external welding defect parameter is: Where W represents the external welding defect parameter of the sheet metal to be tested, ω i represents the importance weight of the i-th binary image to be tested, represents the weight of the first pore area in the i-th binary image to be tested, represents the weight of the first burn-through area in the i-th binary image to be tested, represents the weight of the first slag inclusion area in the i-th binary image to be tested, represents the weight of the crack area in the i-th image to be tested, represents the first pore area in the i-th binary image to be tested, represents the first burn-through area in the i-th binary image to be tested, represents the first slag inclusion area in the i-th binary image to be tested, represents the crack area in the i-th binary image to be tested, A total represents the total area of ​​the sheet metal to be tested, and j represents the total number of binary images to be tested; based on the X-ray image of the sheet metal to be tested, the internal welding defect parameters of the sheet metal to be tested are obtained, specifically including: Obtain a background image without the sheet metal part to be tested as a background image template; Based on grayscale standardization, the X-ray image of the sheet metal to be tested and the background image template are preprocessed so that the X-ray image of the sheet metal to be tested and the background image template have the same size and grayscale range; The X-ray image of the sheet metal to be tested and the background image template are grayscale standardized, and then each pixel value in the X-ray image is subtracted from the corresponding pixel value in the background image to obtain the image to be segmented; Obtaining the segmentation threshold based on the average gray value of the X-ray image and the background image template; Based on the segmentation threshold, threshold processing is performed on the image to be segmented to obtain a binary image to be segmented; Based on region growth, fill and connect the partial images of the sheet metal to be tested in the binary image to be segmented to obtain a full-range image of the sheet metal to be tested; Based on the omnidirectional image of the sheet metal part to be tested, the corresponding internal welding defect type information in the omnidirectional image is recorded respectively, wherein the internal welding defect type information includes a second pore area, a second slag inclusion area, a second crack area, an unmelted region area, an impurity area, and an uneven structure area; According to the internal welding defect type information, the internal defect parameters of the sheet metal part to be tested are obtained; The calculation steps of the second pore area, the second slag inclusion area, the second crack area, the unmelted region area, the impurity area and the non-uniform structure area are as follows: S501: Acquire a full-range image of the sheet metal part to be tested; S502: Use the cv.findContours function in the image processing library to perform contour detection on the omnidirectional image to obtain the boundary of each defect area, wherein the dark spots and dark spots in the form of circles and ellipses correspond to the second pore area, the irregular bright spots and bright spots correspond to the second slag inclusion area, the thinner dark lines and dark shadows correspond to the second crack area, the irregular and lighter areas around the welding area correspond to the unmelted area, the dark spots in the form of obvious irregular shapes correspond to the impurity area, and the areas in the form of uneven density and grayscale changes correspond to the uneven structure area; S503: respectively calculating the sum of the number of pixels of the defective areas corresponding to the second pore area, the second slag inclusion area, the second crack area, the unmelted region area, the impurity area and the non-uniform structure area, that is, the second pore area, the second slag inclusion area, the second crack area, the unmelted region area, the impurity area and the non-uniform structure area; The calculation formula of the internal welding defect parameter is: Where N represents the internal defect parameter of the sheet metal to be tested, B n Represents the area of ​​various defects, γ n represents the weight factor of each defect, and m represents the total number of defect types; According to the welding defect index, the sheet metal parts to be tested are evaluated to obtain the total defect information of the sheet metal parts to be tested, including: Based on the external welding defect type information, external defect position information and external defect size information of each defect are obtained; Based on the external defect position information and the external defect size information, the external defect influencing factor is obtained; Based on the internal welding defect type information, internal defect position information and internal defect size information of each defect are obtained; Based on the internal defect position information and the internal defect size information, the internal defect influencing factor is obtained; 1% of the average value of the external defect influencing factor and the internal defect influencing factor is used as the welding defect index threshold of the sheet metal to be tested; Obtaining a welding defect index according to external welding defect parameters and internal welding defect parameters; Determine whether the welding defect index exceeds the welding defect index threshold. If so, perform welding defect repair on the sheet metal to be tested. If not, continue to obtain the multi-directional image and X-ray image of the next sheet metal to be tested. Wherein, the calculation formula of the external defect influencing factor is: In the formula, I out represents the external defect influence factor, θ α represents the weight of the αth external welding defect location parameter, C α represents the location parameter of the αth external welding defect, represents the weight of the αth external defect size parameter, D α represents the αth external defect size parameter, and β represents the total number of external defect types; The calculation formula of the internal defect influence factor is: In the formula, I in represents the internal defect influence factor, ρ μ represents the weight of the μth internal welding defect location parameter, E μ represents the μth internal welding defect location parameter, σ μ represents the weight of the μth internal defect size parameter, F μ represents the μth internal defect size parameter, and π represents the total number of internal defect types; The calculation formula of the welding defect index is: G z =δW+εN; In the formula, G z represents the welding defect index of the zth sheet metal part to be tested, W represents the external welding defect parameter, δ represents the weight of the external welding defect parameter, N represents the internal welding defect parameter, and ε represents the weight of the internal welding defect parameter.

2. The intelligent detection method for sheet metal welding defects based on data feedback according to claim 1 is characterized in that: Determine the position of the welding area of ​​the sheet metal part to be tested in each image in the sheet metal part image library to obtain the welding area image library, specifically including: Get historical sheet metal image library; Based on the visibility of the welding area of ​​sheet metal parts in the historical sheet metal parts image library, images without blur and overexposure are screened out from the historical sheet metal parts image library to obtain a standard welding area image library; According to the standard welding area image library, the resolution information of each standard welding area image is obtained; Based on the resolution information of each standard welding area image, a resolution index of each standard welding area image is obtained; According to the size of the resolution index of each standard welding area image, the standard welding area images in the standard welding area image library are sorted, and the standard welding area image with the first sorting is extracted to obtain a welding feature image; Grayscale processing is performed on the welding feature image, and edge enhancement is performed on the grayscale image to obtain a welding feature template; Using the image processing library to load the welding feature template and the image of the sheet metal part to be tested in the image library of the sheet metal part to be tested; Based on the normalized correlation matching method, every possible position in the image of the sheet metal part to be tested is traversed with the preset image size of the welding feature template, and its correlation with the welding feature template is calculated, and the normalized correlation matching value of each position and the corresponding position coordinates are recorded; The position with the highest normalized correlation matching value is taken as the position with the maximum correlation value, and half of the maximum normalized correlation matching value is taken as the correlation matching threshold; Based on the correlation matching threshold, each possible position in the image of the sheet metal part to be tested is screened, and the position above the correlation matching threshold is marked in the image of the sheet metal part to be tested; The marked positions in the image of the sheet metal part to be tested are connected, and the part of the image of the sheet metal part to be tested with the largest connected area is taken as the welding area image to obtain the welding area image.

3. A sheet metal welding defect intelligent detection device based on data feedback, used to implement the detection method according to any one of claims 1-2, characterized in that: include: A data acquisition module, the data acquisition module is used to acquire multi-directional images of the sheet metal part to be tested, X-ray images of the sheet metal part to be tested, historical sheet metal part images and background image templates; A data processing module, the data processing module is used to determine the position of the sheet metal to be tested in the multi-directional image based on Canny edge detection, and separate the image of the sheet metal to be tested from the multi-directional image, obtain the image library of the sheet metal to be tested, and pre-process the X-ray image and background image template of the sheet metal to be tested based on grayscale standardization, so that the X-ray image and the background image template of the sheet metal to be tested have the same size and grayscale range, and grayscale standardize the X-ray image and the background image template of the sheet metal to be tested, and then subtract the corresponding pixel value in the background image from each pixel value in the X-ray image to obtain the image to be segmented, which is used to calculate external welding defect parameters, internal welding defect parameters, external defect influencing factors, internal defect influencing factors, welding defect index and welding defect index threshold, and is used to determine whether the welding defect index exceeds the welding defect index threshold, and determine whether the sheet metal to be tested needs welding defect repair; A data transmission module, which is used to transmit the images and data in the data acquisition module to the data processing module and the display module; The display module is used to display the multi-directional images of the sheet metal part to be tested and the image processing process and to display the X-ray image of the sheet metal part to be tested and the image processing process.

4. The intelligent detection device for sheet metal welding defects based on data feedback according to claim 3 is characterized in that: The data acquisition module specifically includes: A first acquisition unit, the first acquisition unit is used to acquire a multi-directional image of the sheet metal part to be tested and an X-ray image of the sheet metal part to be tested; The second acquisition unit is used to acquire a historical sheet metal image library and a background image template.

5. The intelligent detection device for sheet metal welding defects based on data feedback according to claim 3 is characterized in that: The data processing module specifically includes: An image processing unit, the image processing unit is used to determine the position of the sheet metal to be tested in the multi-directional image based on Canny edge detection, separate the image of the sheet metal to be tested from the multi-directional image, obtain an image library of the sheet metal to be tested, pre-process the X-ray image of the sheet metal to be tested and the background image template based on grayscale standardization, so that the X-ray image of the sheet metal to be tested and the background image template have the same size and grayscale range, grayscale standardize the X-ray image of the sheet metal to be tested and the background image template, and then subtract the corresponding pixel value in the background image from each pixel value in the X-ray image to obtain an image to be segmented; A calculation unit, the calculation unit is used to calculate external welding defect parameters, internal welding defect parameters, external defect influencing factors, internal defect influencing factors, welding defect index and welding defect index threshold; The judging unit is used to judge whether the welding defect index exceeds the welding defect index threshold value, and judge whether the sheet metal part to be tested needs welding defect repair.

6. The intelligent detection device for sheet metal welding defects based on data feedback according to claim 3 is characterized in that: The display module specifically includes: A first display unit, the first display unit is used to display the multi-directional images of the sheet metal part to be tested and the image processing process; A second display unit, wherein the second display unit is used to display the X-ray image of the sheet metal part to be tested and the image processing process.

Images