CN115018827A - Automatic detection method for quality of building material weld joint - Google Patents
Automatic detection method for quality of building material weld joint Download PDFInfo
- Publication number
- CN115018827A CN115018827A CN202210923913.5A CN202210923913A CN115018827A CN 115018827 A CN115018827 A CN 115018827A CN 202210923913 A CN202210923913 A CN 202210923913A CN 115018827 A CN115018827 A CN 115018827A
- Authority
- CN
- China
- Prior art keywords
- defect
- undercut
- row
- undercut defect
- rectangle
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 239000004566 building material Substances 0.000 title claims abstract description 65
- 238000001514 detection method Methods 0.000 title claims abstract description 36
- 230000007547 defect Effects 0.000 claims abstract description 272
- 238000003466 welding Methods 0.000 claims abstract description 157
- 238000000034 method Methods 0.000 claims abstract description 25
- 239000000463 material Substances 0.000 claims abstract description 20
- 239000010953 base metal Substances 0.000 claims description 11
- 238000012545 processing Methods 0.000 abstract description 5
- 238000012549 training Methods 0.000 abstract description 4
- 230000002950 deficient Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 238000004519 manufacturing process Methods 0.000 description 3
- 239000002893 slag Substances 0.000 description 3
- 230000000694 effects Effects 0.000 description 2
- 239000002184 metal Substances 0.000 description 2
- 230000011218 segmentation Effects 0.000 description 2
- 230000009466 transformation Effects 0.000 description 2
- 230000001133 acceleration Effects 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 238000013441 quality evaluation Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
- G06V10/267—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion by performing operations on regions, e.g. growing, shrinking or watersheds
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
- G01N2021/8887—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges based on image processing techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Theoretical Computer Science (AREA)
- Biochemistry (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Quality & Reliability (AREA)
- Signal Processing (AREA)
- Multimedia (AREA)
- Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
Abstract
The invention relates to the technical field of image data processing, in particular to an automatic detection method for building material weld joint quality. Converting a welding line image on the surface of a building material into an HSV image and determining a welding line area, fitting the average value of V channel values on each row of the welding line area to obtain a standard smooth curve, and determining the defect degree of the welding line area according to the deviation of each V channel value on each row of the welding line area and the standard smooth curve; calculating the variance of V channel values on each line in a welding line and base material connecting area to determine whether the line is an undercut defect line, determining the defect degree of the undercut defect line according to the undercut defect depth, length and kurtosis, determining the weight of the defect degree of the undercut defect line according to the undercut size of the undercut defect line to obtain the undercut defect degree of the area, and completing welding line quality detection according to the obtained defect degrees of the two areas. The method has the advantages that the judgment threshold is set in a self-adaptive mode, the detection is comprehensive, a large amount of training data or point cloud data does not need to be acquired, and the efficient and accurate detection of the quality of the building material welding seam is realized.
Description
Technical Field
The invention relates to the technical field of image data processing, in particular to an automatic detection method for building material weld joint quality.
Background
In recent years, with the continuous acceleration of the industrialization process of China, the market competition is increasingly fierce, and the upgrading of the processing and manufacturing technology of the product becomes an important means for improving the product competitiveness of each enterprise. Welding is used as a mature material forming process, is widely applied in the building manufacturing industry, and the number of products produced by adopting welding as a building material connecting means every year is huge. Detecting the quality of a welded seam after welding is an essential link in welding production, and the quality of the welded seam not only influences the service performance and service life of a welded product, but also influences personal safety and property safety.
The most common welding quality detection methods in the prior art comprise methods such as high-precision eddy current and ultrasonic flaw detection, which can well judge whether a welding seam has defects such as bubbles, but have complex implementation process, cannot quantitatively acquire information such as width and area of the welding defect, and has high detection cost; however, in the existing means for detecting the weld defects by using the image processing method, if the means for detecting the weld defects by using the neural network is used, the problem that the required training sample is large and the processing speed is slow exists, and if the means for judging the weld defects by using three-dimensional point cloud data of the weld obtained by three-dimensional scanning is used, the problem that the value is single and the applicable scene is narrow exists when the judgment threshold value used for judging by using the point cloud data obtained in real time, so that the weld quality detection effect is poor.
In conclusion, the existing weld joint quality detection method has the problems of low detection efficiency and poor detection effect.
Disclosure of Invention
The invention provides an automatic detection method for the quality of a building material weld joint, which is used for solving the problems of low efficiency and inaccuracy of the detection of the quality of the building material weld joint in the prior art, and adopts the following technical scheme:
the invention relates to an automatic detection method for the weld quality of building materials, which comprises the following steps:
acquiring a building material surface weld image, converting the building material surface weld image into an HSV (hue, saturation and value) image and then determining a weld area;
taking the minimum external rectangle of the welding seam region as a first rectangle, taking the longitudinal direction of the first rectangle as the welding seam trend, determining the mean value of the V channel values of the pixel points on each column in the first rectangle, then fitting according to the determined mean value of the V channel values of the pixel points on each column in the first rectangle to obtain a standard smooth curve of the welding seam region in the transverse direction, calculating the deviation degree of the V channel values of the pixel points on each row in the first rectangle and the standard smooth curve, and summing the deviation degrees of the V channel values of the pixel points on all rows and the standard smooth curve to obtain the defect degree of the welding seam region;
respectively extending the first rectangle leftwards and rightwards in the transverse direction by a set distance to obtain a second rectangle, taking the area between the left edge of the first rectangle and the left edge of the second rectangle as a left extending area, taking the area between the right edge of the first rectangle and the right edge of the second rectangle as a right extending area, calculating the variance of the V channel values of the pixel points on each line in the left extending area and the variance of the V channel values of the pixel points on each line in the right extending area, and if the variances are larger than a variance threshold value, taking the corresponding line as a undercut defect line;
respectively counting the longitudinal length value of the undercut defect line when the undercut defect line continuously appears once in the left extending area and the right extending area, calculating the ratio of the longitudinal length value of the undercut defect line when the undercut defect line continuously appears once to the longitudinal total length of the second rectangle, taking the ratio as the defect weight of each undercut defect line continuously appearing this time, and repeating the acquiring process of the defect weight of the undercut defect line, thereby obtaining the defect weight of all undercut defect lines;
determining the undercut defect depth of the undercut defect row according to the maximum value of the V channel value and the minimum value of the V channel value at each pixel point in the undercut defect row; taking the number of pixel points with V channel values smaller than the undercut defect threshold value in the undercut defect row as the undercut defect length of the undercut defect row; calculating the undercut defect kurtosis of the undercut defect row according to the undercut defect length of the undercut defect row and the V channel value of each pixel point in the undercut defect length;
calculating the defect degree of the undercut defect row according to the undercut defect depth, the undercut defect length and the undercut defect kurtosis of the undercut defect row, and calculating the undercut defect degree in the connecting area between the welding line and the base metal according to the defect degree and the defect weight of each undercut defect row;
and calculating the damage degree of the welding line according to the defect degree of the welding line area and the undercut defect degree in the connecting area between the welding line and the base metal, and judging the welding line quality according to the damage degree to finish the detection of the welding line quality of the building materials.
The invention has the beneficial effects that:
according to the automatic detection method for the welding quality of the building material, for a welding seam area, a standard smooth curve of the welding seam area in the transverse direction is obtained by fitting the mean value of the V channel value of each row of pixel points on the minimum external rectangle of the welding seam area, the standard smooth curve is a defect judgment threshold value which is obtained by self-adapting the condition of the overall V channel value of the welding seam area in the welding seam image on the surface of the building material, and then the defect degree of the welding seam area is determined by calculating the sum of the deviation degree of each pixel point on each row of the minimum external rectangle of the welding seam area and the standard smooth curve, the defect judgment threshold value is set in a self-adapting mode, the welding quality representation of the welding seam area is more comprehensive, and the more accurate evaluation of the welding quality of the welding seam area is realized; for a connecting area of a welding line and a base material, firstly calculating the variance of V channel values of pixel points on each row in the area, comparing with a variance threshold value to determine whether the corresponding row contains pixel points of undercut defects, taking the row containing the pixel points of the undercut defects as the undercut defect row, then performing defect weight assignment on each undercut defect row according to the proportion of the longitudinal length of each undercut defect in the connecting area of the welding line and the base material in the whole area, then accurately representing the undercut defect degree of the undercut defect row by calculating the undercut defect depth, the undercut defect length and the undercut defect kurtosis of the undercut defect row, and finally obtaining the undercut defect degree of the connecting area of the welding line and the base material. Finally, according to the defect degree of the obtained welding line area and the undercut defect degree of the connecting area of the welding line and the base metal, the invention can realize the rapid, efficient and accurate detection of the quality of the welding line of the building material.
Further, the undercut defect kurtosis of the undercut defect row is:
wherein,
is the undercut defect kurtosis of the z-th undercut defect row in the left and right side extension regions,
is the V channel value of the D-th pixel point in the undercut defect length D of the z-th undercut defect row in the extension areas at the left and right sides,
the average value of the V channels of all pixel points in the undercut defect length D of the z-th undercut defect row in the extension areas on the left side and the right side.
Further, the process of obtaining the defect degree of the weld joint region comprises:
calculating the defect degree of the k-th row on the first rectangle:
wherein,
is the defect level of the k-th row on the first rectangle,
is the V channel value of the j pixel point in the k line on the first rectangle,
represents the ordinate value corresponding to the abscissa of the standard smooth curve being j,
and
respectively representing the abscissa minimum value and the abscissa maximum value of the first rectangle in a coordinate system;
then calculating the defect degree of the welding seam region:
wherein P is the defect degree of the welding seam area, and M is the longitudinal length of the first rectangle.
Further, the undercut defect depth of the undercut defect row is as follows:
wherein,
the undercut defect depth of the z-th undercut defect row in the left and right side extension regions,
the maximum V channel value in the z-th undercut defect line in the left and right extended regions,
the minimum V channel value in the z th undercut defect line in the left and right extended regions.
Further, the defect degree of the undercut defect row is as follows:
wherein,
the defect degree of the z-th undercut defect row in the left and right extended regions,
the undercut defect depth of the z-th undercut defect row in the left and right side extension regions,
the crest factor of the undercut defect of the z-th undercut defect row in the left and right extended regions.
Further, the undercut defect degree in the connecting region between the weld joint and the base material is as follows:
wherein,
the degree of undercut defect in the connecting region between the weld and the base material,
the defect degree of the z-th undercut defect row in the left and right extended regions,
the defect weight of the z-th undercut defect row in the left and right extended regions.
Further, the damage degree of the welding seam is as follows:
wherein,
as the degree of damage to the weld,
is a weight of the degree of defect in the weld area,
is a weight of the degree of defects in the connecting region between the weld and the parent material.
Further, the obtaining process of the standard smooth curve is as follows:
calculating the average value of the V channel values of the pixel points on each row of the first rectangle from left to right in sequence to obtain a set of the average values of the V channels on each row of the first rectangle
Where N is the lateral length of the first rectangle, then set
And performing Gaussian fitting to obtain a standard smooth curve of the welding seam area in the transverse direction.
Further, the process of determining the variance threshold is as follows:
determining a welding line region in a welding line image on the surface of the building material without undercut defects to obtain a minimum circumscribed rectangle of the welding line region, respectively extending the minimum circumscribed rectangle leftwards and rightwards in the transverse direction by the set distance, determining a leftwards extending region and a rightwards extending region in the extending process, calculating the variance of V channel values of all pixel points in the leftwards extending region and the rightwards extending region as a standard variance, and taking the standard variance of a set multiple as the variance threshold.
Drawings
FIG. 1 is a flow chart of the automatic detection method for the welding seam quality of the building material;
FIG. 2 is a schematic view of the normal weld shape of the present invention;
FIG. 3 is a schematic view of the weld shape when the weld region of the present invention is defective;
FIG. 4 is a schematic view of a minimum circumscribed rectangle corresponding to a weld region of the present invention;
FIG. 5 is a schematic diagram of a standard smooth curve of the weld region in the transverse direction obtained by the V-channel mean fitting of each row of the minimum circumscribed rectangle corresponding to the weld region;
FIG. 6 is a schematic view of an undercut defect of the present invention;
FIG. 7 is a schematic view showing the shape of a weld in the case where an undercut defect is present in the joint region between the weld and the base material in the present invention.
Detailed Description
The main purposes of the invention are: and (3) by utilizing a computer vision technology, processing the collected building material surface weld image, calculating the defect degree of the weld and base metal connecting area according to the weld characteristics on the weld area and the characteristics of the weld and base metal connecting area, further obtaining the damage degree of the building material weld, determining the welding quality grade of the building material weld according to the damage degree grading basis of the building material weld, and completing the automatic detection of the building material weld quality.
The following describes a method for automatically detecting the weld quality of a building material according to the present invention in detail with reference to the accompanying drawings and embodiments.
The method comprises the following steps:
the embodiment of the automatic detection method for the quality of the welding line of the building material, disclosed by the invention, has the overall flow as shown in figure 1, and comprises the following specific processes:
the method comprises the steps of firstly, obtaining a building material surface weld image, converting the building material surface weld image into an HSV image, and then segmenting the converted building material surface weld image to determine a weld area.
In order to determine the welding quality of the welding seam on the surface of the building material by using the computer vision technology, it is necessary to first obtain an image of the welding seam on the surface of the building material.
The weld formed by welding the building material is required to be smooth and free from welding defects such as air holes, slag inclusion, welding beading, burn-through and the like, so that the weld is in a shape slightly higher than the surface of the base material under normal conditions, and when the building material is horizontally viewed, the weld takes an arc shape as shown in fig. 2. However, when the weld has the above-mentioned weld defects such as blowholes, slag inclusions, flash, and burn-through, these defects all cause deformation of the weld, changing the weld from a circular arc shape in plan view to a shape having a central recess as shown in fig. 3.
The change of the shape of the welding seam caused by the welding defect is corresponding to the change of the image characteristics on the acquired welding seam image, and in the embodiment, the acquired welding seam image on the surface of the building material is preferably denoised and then is converted into an HSV (hue, saturation and value) image by an RBG (radial basis function) image to determine the image characteristics of the welding seam image on the surface of the building material, so that whether the welding defect is generated in the welding seam area and the area between the welding seam and the base metal is determined. Of course, in other embodiments, other forms of transformation may be performed on the acquired building material surface weld image to determine image characteristics, such as grayscale transformation, etc. The denoising method, which is the preferred median filtering in this embodiment, may also be implemented by other feasible methods in the prior art in other embodiments.
After the welding seam image on the building material surface is converted into the HSV image, the DNN semantic segmentation method is adopted to identify the welding seam area in the image, and the related content of the used DNN network is as follows:
a. the used training data set is a building material surface welding seam image data set collected in an overlook mode;
b. the pixels to be segmented are divided into 2 types, namely the labeling process of the corresponding labels of the training data set is as follows: in the semantic label of the single channel, the pixel at the corresponding position belongs to the background class, namely the label of the non-welding seam is 0, and the label of the pixel belonging to the welding seam is 1;
c. the task of the network is classification, so the loss function used is a cross entropy loss function.
Therefore, the weld joint area can be successfully divided and determined from the weld joint image on the surface of the building material by a DNN semantic division method.
And step two, determining the defect degree in the welding seam area and the defect degree in the welding seam and base metal connecting area according to the brightness change characteristics, thereby determining the damage degree of the welding seam.
The welding quality of the building material is mainly determined by the defect degree of the welding seam and whether the connecting area between the welding seam and the base metal is smooth, natural and free of defects, so that the welding damage degree in the two areas is determined by analyzing the image characteristics in the two areas, and the quality of the welding seam of the building material is further calculated.
1. And determining the defect degree in the welding seam area according to the brightness change characteristics.
In the embodiment, the acquired welding seam image on the surface of the building material is converted into the HSV image from the RGB image, so that the welding seam area obtained by semantic segmentation in the welding seam image on the surface of the building material is also the HSV image.
And (3) taking a brightness channel image of the welding seam region, namely a V-channel image, and combining the shape characteristics of the welding seam to know that the V-channel value of a pixel point from the edges of two sides of the welding seam to the central line of the welding seam is gradually increased in the welding seam region without welding defects.
And respectively taking the initial and final positions of the rows and the initial and final positions of the columns of the welding seam region to obtain the minimum circumscribed rectangle of the welding seam region, wherein the longitudinal direction of the minimum circumscribed rectangle is the trend of the welding seam as shown in fig. 4. And establishing a plane coordinate system by taking the lower left corner of the minimum circumscribed rectangle as an origin, wherein the rectangle is positioned in the first quadrant. Calculating the average value of the V channel values of the pixel points on each row of the minimum external rectangle from left to right in sequence to obtain the set of the average values of the V channels on each row of the minimum external rectangle
And N is the transverse length of the minimum circumscribed rectangle corresponding to the welding seam area.
Then pair the sets
Gaussian fitting was performed to obtain a standard smooth curve of the weld region in the transverse direction as shown in FIG. 5
Wherein i is the horizontal axis,
the unit is a single pixel point,
and the vertical axis represents the mean value of the V channels of each row of pixel points in the fitted welding seam area.
Determining the abscissa minimum value of the minimum circumscribed rectangle corresponding to the welding seam region in the coordinate system
And maximum value of abscissa
And then calculating the defect degree of the k line on the minimum circumscribed rectangle corresponding to the welding seam region:
wherein,
the defect degree of the k-th row on the minimum circumscribed rectangle corresponding to the welding seam area,
is the V channel value of the jth pixel point in the kth line on the minimum external rectangle corresponding to the welding seam area,
represents the vertical coordinate value corresponding to the horizontal coordinate j of the standard smooth curve, and the value range of j is
,
。
Then calculating the defect degree of the welding seam region:
wherein, P is the defect degree of the welding seam area, and M is the longitudinal length of the minimum circumscribed rectangle corresponding to the welding seam area.
The defect degree P is obtained by comparing the V channel value of each pixel point on the welding line region with the average value of the V channels of the columns of the pixel points, so that the defects of air holes, slag inclusion, welding beading and the like in the welding region can be clearly represented, the smoothness degree of the whole welding line can be represented, and the smoothness degree of the welding line also influences the quality of the welding line, so that the representation of the welding line quality is more comprehensive in the embodiment, the average value of the V channels of each column is used as a comparison threshold value of the V channel value of each pixel point on the column, the self-adaptive setting of the threshold value is realized, and finally, the defect degree P is obtained by the embodiment, and the welding quality evaluation can be more accurately completed.
2. And determining the defect degree of the connecting area between the welding seam and the base material according to the brightness change characteristics.
In the quality inspection of the appearance of the welding seam, the connecting area between the welding seam and the base material is smooth, natural and has no depression, namely, the base material cannot be excessively damaged in the welding process. Therefore, it is necessary to analyze the undercut defect, which is a groove-like defect formed by burning through the welding arc at the edge where the weld base material contacts the weld, as shown in fig. 6. When the building material is viewed in plan with undercut defects, the weld takes the form of a depression outside the arc as shown in fig. 7.
In order to analyze the joint area between the welding seam and the parent metal, the minimum circumscribed rectangle corresponding to the welding seam area is selected to extend leftwards and rightwards in the transverse direction
And obtaining a new rectangular area with the longitudinal length of M and the transverse length of 2N by each pixel point, taking the minimum external rectangle corresponding to the original welding area as a first rectangle, and taking the new rectangular area obtained by transversely extending and expanding the first rectangle as a second rectangle. The present embodiment preferably extends leftwards and rightwards on the basis of the first rectangle
The pixel points, in other embodiments, can be set to other values, such as according to the welding power,The extension length is determined by the type of the base material, and the left and right extension lengths may be the same or different.
The region from the left side boundary of the first rectangle to the left side boundary of the second rectangle is taken as a left extension region, and the region from the right side boundary of the first rectangle to the right side boundary of the second rectangle is taken as a right extension region. Since the undercut region is a pit and the undercut region corresponds to a darker region on the V-channel image, the V-channel value of the region is small, and then if the undercut defect is included in the left-side extended region or the right-side extended region, a row having a large variance of the V-channel value exists in the left-side extended region or the right-side extended region.
Statistics of the first of the left extension region
V channel values of all pixel points in the row and calculating the variance of all the V channel values
Similarly, the first of the right extension region is counted
V channel values of all pixel points in the row and calculating the variance of all the V channel values
Taking a building material surface weld image without undercut defect, determining a left extending area and a right extending area in the building material surface weld image without undercut defect according to the method, and calculating the variance of the V channel values of all pixel points in the left extending area and the right extending area
If, if
Then in the left side extension area
The line contains undercut defect pixel points, which are undercut defect lines, if
In the right extension area
The line contains undercut defect pixel points, which are undercut defect lines.
In the embodiment, after the left extending area and the right extending area are determined by the welding seam image on the building material surface without undercut defect, the variance of the V channel values of all the pixel points in the left extending area and the right extending area is used as the uniform standard variance
In other embodiments, the variance of the V channel values of all the pixel points in the left extension region and the variance of the V channel values of all the pixel points in the right extension region may also be calculated respectively, and then the variances are respectively used as the standard variance on the left side and the standard variance on the right side. In this embodiment, a standard deviation of 1.1 times is preferably used as a determination threshold for determining whether each row contains undercut defective pixel points, and in other embodiments, standard deviations of other times may be set as determination thresholds for determining whether each row contains undercut defective pixel points, of course, according to specific requirements.
Counting the longitudinal length value of the undercut defect line when the u-th continuous occurrence in the left extension area
And the value of the longitudinal length of the line of undercut defects in the right extended area at the time of the v-th continuous occurrence
. Longitudinal length value
And
in practice, the longitudinal length of the u-th undercut defect in the left extended region and the longitudinal length of the v-th undercut defect in the right extended region are shown respectively.
Calculating the longitudinal length value of the undercut defect line when the u-th continuous occurrence in the left extension area
Ratio to the total longitudinal length M of the left-hand extension
The ratio is the undercut defect weight of each undercut defect row continuously appearing at the u-th time in the left extension area, and the longitudinal length value of each undercut defect row continuously appearing at the v-th time in the right extension area is calculated
Ratio to the total longitudinal length M of the right extension
The ratio is the undercut defect weight of each undercut defect row continuously appearing at the v-th time in the right extended area. Thus, the undercut defect weight of each undercut defect row in the left and right extended regions can be obtained, and the undercut defect weight of the z-th undercut defect row in the left and right extended regions can be expressed as
。
Then, in this embodiment, statistics is performed on all undercut defect rows in the left and right extended regions, and the undercut defect depth, the undercut defect length, and the undercut defect kurtosis of the z-th undercut defect row in the left and right extended regions are calculated.
The undercut defect depth of the z-th undercut defect row in the left and right side extension regions is:
wherein,
the undercut defect depth of the z-th undercut defect row in the left and right side extension regions,
the maximum V channel value in the z-th undercut defect line in the left and right extended regions,
the minimum V channel value in the z th undercut defect line in the left and right extended regions.
Regarding the undercut defect length of the z-th undercut defect row in the left and right extended regions, firstly, an undercut defect threshold is determined in the left and right extended regions according to the Otsu threshold method
Then, on the z th undercut defect line, it is judged that the V channel value is less than the undercut defect threshold value
And taking the number of the pixel points as the undercut defect length D of the z-th undercut defect row.
The crest factor of the undercut defect of the z-th undercut defect row in the left and right side extension areas is as follows:
wherein,
is the undercut defect kurtosis of the z-th undercut defect row in the left and right side extension regions,
is the z-th undercut in the left and right side extension areasThe V channel value of the D-th pixel point in the undercut defect length D of the defect row,
is the V channel mean value of all pixel points in the undercut defect length D of the z-th undercut defect row in the extension areas at the left side and the right side,
the fourth order central moment of the V channel values of all pixel points in the undercut defect length D of the z-th undercut defect row is represented,
and the square of the second-order central moment of the V channel values of all pixel points in the undercut defect length D of the z-th undercut defect row is represented.
According to the calculated undercut defect depth of the z-th undercut defect row in the left and right extending areas
Undercut defect length D and undercut defect kurtosis
And calculating the defect degree of the z-th undercut defect row as follows:
wherein,
the defect degree of the z-th undercut defect row in the left and right extended regions.
Finally, the undercut defect level in the connection region between the weld and the parent metal can be calculated:
wherein,
the degree of undercut defect in the connecting region between the weld and the base material,
the defect degree of the z-th undercut defect row in the left and right extended regions,
the defect weight of the z-th undercut defect row in the left and right extended regions.
3. And determining the damage degree of the welding seam according to the defect degree in the welding seam area and the defect degree in the connecting area between the welding seam and the base metal.
The damage degree of the welding seam is as follows:
wherein,
as the degree of damage to the weld,
is a weight of the degree of defect in the weld area,
is a weight of the degree of defects in the connecting region between the weld and the parent material. This embodiment is preferred
In other embodiments, it can be set as required
And
the value of (a).
And step three, determining the quality grade of the welding line according to the damage degree of the welding line, and completing automatic detection of the quality of the welding line of the building material.
According to the second step, the damage degree F of the welding seam can be known, and the quality of the welding seam is classified into first-grade, second-grade and third-grade products and unqualified products from high to low according to the known existing welding seam quality detection standard. According to the specific quality requirement, the first-level welding quality parameters can be respectively and correspondingly set
Second grade welding quality parameter
And three-stage welding quality parameters
。
Therefore, the quality grade of the welding seam can be determined according to the damage degree F of the obtained welding seam and welding quality parameters at all levels, and the detection of the quality of the welding seam of the building materials is completed:
when in use
Judging the quality of the welding seam of the building material to be first grade;
when in use
Judging the quality of the welding seam of the building material to be in a second level;
when in use
Judging the quality of the building material weld joint to be three-level;
when the temperature is higher than the set temperature
And judging that the quality of the building material welding line is unqualified.
The above-mentioned embodiments are only used for illustrating the technical solutions of the present application, and not for limiting the same; although the present application has been described in detail with reference to the foregoing embodiments, it should be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; the modifications or substitutions do not make the essence of the corresponding technical solutions deviate from the technical solutions of the embodiments of the present application, and are included in the protection scope of the present application.
Claims (9)
1. The automatic detection method for the quality of the welding line of the building material is characterized by comprising the following steps of:
acquiring a building material surface weld image, converting the building material surface weld image into an HSV (hue, saturation and value) image and then determining a weld area;
taking the minimum external rectangle of the welding seam region as a first rectangle, taking the longitudinal direction of the first rectangle as the welding seam trend, determining the mean value of the V channel values of the pixel points on each column in the first rectangle, then fitting according to the determined mean value of the V channel values of the pixel points on each column in the first rectangle to obtain a standard smooth curve of the welding seam region in the transverse direction, calculating the deviation degree of the V channel values of the pixel points on each row in the first rectangle and the standard smooth curve, and summing the deviation degrees of the V channel values of the pixel points on all rows and the standard smooth curve to obtain the defect degree of the welding seam region;
respectively extending the first rectangle leftwards and rightwards in the transverse direction by a set distance to obtain a second rectangle, taking the area between the left edge of the first rectangle and the left edge of the second rectangle as a left extending area, taking the area between the right edge of the first rectangle and the right edge of the second rectangle as a right extending area, calculating the variance of the V channel values of the pixel points on each line in the left extending area and the variance of the V channel values of the pixel points on each line in the right extending area, and if the variances are larger than a variance threshold value, taking the corresponding line as a undercut defect line;
respectively counting the longitudinal length value of the undercut defect line when the undercut defect line continuously appears once in the left extending area and the right extending area, calculating the ratio of the longitudinal length value of the undercut defect line when the undercut defect line continuously appears once to the longitudinal total length of the second rectangle, taking the ratio as the defect weight of each undercut defect line continuously appearing this time, and repeating the acquiring process of the defect weight of the undercut defect line, thereby obtaining the defect weight of all undercut defect lines;
determining the undercut defect depth of the undercut defect row according to the maximum value of the V channel value and the minimum value of the V channel value at each pixel point in the undercut defect row; taking the number of pixel points with V channel values smaller than the undercut defect threshold value in the undercut defect row as the undercut defect length of the undercut defect row; calculating the undercut defect kurtosis of the undercut defect row according to the undercut defect length of the undercut defect row and the V channel value of each pixel point in the undercut defect length;
calculating the defect degree of the undercut defect row according to the undercut defect depth, the undercut defect length and the undercut defect kurtosis of the undercut defect row, and calculating the undercut defect degree in the connecting area between the welding line and the base metal according to the defect degree and the defect weight of each undercut defect row;
and calculating the damage degree of the welding line according to the defect degree of the welding line area and the undercut defect degree in the connecting area between the welding line and the base metal, and judging the welding line quality according to the damage degree to finish the detection of the welding line quality of the building materials.
2. A building material weld quality automatic detection method as claimed in claim 1, wherein the undercut defect rows have undercut defect kurtosis of:
wherein,
is the undercut defect kurtosis of the z-th undercut defect row in the left and right side extension regions,
is the z-th region in the left and right extending regionsThe V channel value of the D-th pixel point in the undercut defect length D of each undercut defect row,
the average value of the V channels of all pixel points in the undercut defect length D of the z-th undercut defect row in the extension areas on the left side and the right side.
3. The automatic detection method for the weld quality of the building material according to claim 1 or 2, wherein the process of obtaining the defect degree of the weld area comprises the following steps:
calculating the defect degree of the k-th row on the first rectangle:
wherein,
is the defect level of the k-th row on the first rectangle,
is the V channel value of the j pixel point in the k line on the first rectangle,
represents the ordinate value corresponding to the abscissa of the standard smooth curve being j,
and
respectively representing the abscissa minimum value and the abscissa maximum value of the first rectangle in a coordinate system;
then calculating the defect degree of the welding seam region:
wherein P is the defect degree of the welding seam area, and M is the longitudinal length of the first rectangle.
4. A building material weld quality automatic detection method according to claim 1, characterized in that the undercut defect row has undercut defect depths of:
wherein,
the undercut defect depth of the z-th undercut defect row in the left and right side extension regions,
the maximum V channel value in the z-th undercut defect line in the left and right extended regions,
the minimum V channel value in the z th undercut defect line in the left and right extended regions.
5. The automatic detection method for the weld quality of the building material according to claim 1, wherein the defect degrees of the undercut defect rows are as follows:
wherein,
the defect degree of the z-th undercut defect row in the left and right extended regions,
the undercut defect depth of the z-th undercut defect row in the left and right side extension regions,
the crest factor of the undercut defect of the z-th undercut defect row in the left and right extended regions.
6. A building material weld quality automatic detection method according to claim 1, characterized in that the degree of undercut defects in the connection region between the weld and the base material is:
wherein,
the degree of undercut defect in the connecting region between the weld and the base material,
the defect degree of the z-th undercut defect row in the left and right extended regions,
the defect weight of the z-th undercut defect row in the left and right extended regions.
7. The automatic detection method for the weld quality of the building material according to claim 1, wherein the damage degree of the weld is as follows:
wherein,
as the degree of damage to the weld,
is a weight of the degree of defect in the weld area,
is a weighting of the degree of defects in the connecting region between the weld and the parent material.
8. The automatic detection method for the weld quality of the building material according to claim 1, wherein the standard smooth curve is obtained by the following steps:
calculating the average value of the V channel values of the pixel points on each row of the first rectangle from left to right in sequence to obtain a set of the average values of the V channels on each row of the first rectangle
Where N is the lateral length of the first rectangle, then set
And performing Gaussian fitting to obtain a standard smooth curve of the welding seam area in the transverse direction.
9. The building material weld quality automatic detection method according to claim 1, wherein the variance threshold is determined by the following steps:
determining a welding line region in a welding line image on the surface of the building material without undercut defects to obtain a minimum circumscribed rectangle of the welding line region, respectively extending the minimum circumscribed rectangle leftwards and rightwards in the transverse direction by the set distance, determining a leftwards extending region and a rightwards extending region in the extending process, calculating the variance of V channel values of all pixel points in the leftwards extending region and the rightwards extending region as a standard variance, and taking the standard variance of a set multiple as the variance threshold.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210923913.5A CN115018827B (en) | 2022-08-03 | 2022-08-03 | Automatic detection method for quality of building material weld joint |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210923913.5A CN115018827B (en) | 2022-08-03 | 2022-08-03 | Automatic detection method for quality of building material weld joint |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115018827A true CN115018827A (en) | 2022-09-06 |
| CN115018827B CN115018827B (en) | 2022-11-04 |
Family
ID=83066052
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210923913.5A Active CN115018827B (en) | 2022-08-03 | 2022-08-03 | Automatic detection method for quality of building material weld joint |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115018827B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115213563A (en) * | 2022-09-20 | 2022-10-21 | 扬州耐施工程机械有限公司 | Laser intelligent welding method and system |
| CN115330780A (en) * | 2022-10-13 | 2022-11-11 | 山东盛世恒机械制造有限公司 | Rapid detection method for slag inclusion defect of metal welding |
| CN116342611A (en) * | 2023-05-31 | 2023-06-27 | 无锡日联科技股份有限公司 | Method and device for detecting bonding wires of micro semiconductor |
| CN117708615A (en) * | 2024-02-04 | 2024-03-15 | 芯联集成电路制造股份有限公司 | Product detection method and processing equipment monitoring method |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2008164482A (en) * | 2006-12-28 | 2008-07-17 | Daido Steel Co Ltd | Method for evaluating welding quality |
| WO2014139273A1 (en) * | 2013-03-14 | 2014-09-18 | 苏州华源包装股份有限公司 | Weld seam defect detection method |
| CN104865277A (en) * | 2015-05-11 | 2015-08-26 | 四川大学 | Method for automatically identifying pipe weld defect based on double-wall double-projection transillumination image |
| CN112862794A (en) * | 2021-02-20 | 2021-05-28 | 河北工业大学 | Fillet weld appearance detection method based on structured light |
| CN112950633A (en) * | 2021-04-19 | 2021-06-11 | 上海电机学院 | Aluminum alloy weld surface defect detection method based on line structured light |
| CN113409313A (en) * | 2021-08-18 | 2021-09-17 | 济宁联威车轮制造有限公司 | Wheel weld surface defect detection method based on computer vision |
| WO2021213223A1 (en) * | 2020-04-20 | 2021-10-28 | 广东利元亨智能装备股份有限公司 | Weld quality inspection method, apparatus and system, and electronic device |
| CN113732558A (en) * | 2021-09-06 | 2021-12-03 | 中车工业研究院有限公司 | Online detection method and device for typical quality problems of robot welding |
| CN113989280A (en) * | 2021-12-28 | 2022-01-28 | 武汉市鑫景诚路桥钢模有限公司 | Steel structure welding crack defect detection method based on image processing technology |
-
2022
- 2022-08-03 CN CN202210923913.5A patent/CN115018827B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2008164482A (en) * | 2006-12-28 | 2008-07-17 | Daido Steel Co Ltd | Method for evaluating welding quality |
| WO2014139273A1 (en) * | 2013-03-14 | 2014-09-18 | 苏州华源包装股份有限公司 | Weld seam defect detection method |
| CN104865277A (en) * | 2015-05-11 | 2015-08-26 | 四川大学 | Method for automatically identifying pipe weld defect based on double-wall double-projection transillumination image |
| WO2021213223A1 (en) * | 2020-04-20 | 2021-10-28 | 广东利元亨智能装备股份有限公司 | Weld quality inspection method, apparatus and system, and electronic device |
| CN112862794A (en) * | 2021-02-20 | 2021-05-28 | 河北工业大学 | Fillet weld appearance detection method based on structured light |
| CN112950633A (en) * | 2021-04-19 | 2021-06-11 | 上海电机学院 | Aluminum alloy weld surface defect detection method based on line structured light |
| CN113409313A (en) * | 2021-08-18 | 2021-09-17 | 济宁联威车轮制造有限公司 | Wheel weld surface defect detection method based on computer vision |
| CN113732558A (en) * | 2021-09-06 | 2021-12-03 | 中车工业研究院有限公司 | Online detection method and device for typical quality problems of robot welding |
| CN113989280A (en) * | 2021-12-28 | 2022-01-28 | 武汉市鑫景诚路桥钢模有限公司 | Steel structure welding crack defect detection method based on image processing technology |
Non-Patent Citations (3)
| Title |
|---|
| HUI-HUI CHU 等，: "A vision-based system for post-welding quality measurement", 《THE INTERNATIONAL JOURNAL OF ADVANCED MANUFACTURING TECHNOLOGY》 * |
| JUN SUN 等，: "An Effective Method of Weld Defect Detection and Classification Based on Machine Vision", 《IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS》 * |
| 王维一，: "基于机器视觉的电气焊接在线检测系统研究与设计", 《中国优秀硕士学位论文全文数据库工程科技Ⅰ辑》 * |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115213563A (en) * | 2022-09-20 | 2022-10-21 | 扬州耐施工程机械有限公司 | Laser intelligent welding method and system |
| CN115213563B (en) * | 2022-09-20 | 2022-12-13 | 扬州耐施工程机械有限公司 | Laser intelligent welding method and system |
| CN115330780A (en) * | 2022-10-13 | 2022-11-11 | 山东盛世恒机械制造有限公司 | Rapid detection method for slag inclusion defect of metal welding |
| CN115330780B (en) * | 2022-10-13 | 2023-03-07 | 山东盛世恒机械制造有限公司 | Rapid detection method for slag inclusion defect of metal welding |
| CN116342611A (en) * | 2023-05-31 | 2023-06-27 | 无锡日联科技股份有限公司 | Method and device for detecting bonding wires of micro semiconductor |
| CN116342611B (en) * | 2023-05-31 | 2023-09-08 | 无锡日联科技股份有限公司 | Method and device for detecting bonding wires of micro semiconductor |
| CN117708615A (en) * | 2024-02-04 | 2024-03-15 | 芯联集成电路制造股份有限公司 | Product detection method and processing equipment monitoring method |
| CN117708615B (en) * | 2024-02-04 | 2024-05-17 | 芯联集成电路制造股份有限公司 | Product detection method and processing equipment monitoring method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115018827B (en) | 2022-11-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115018827B (en) | Automatic detection method for quality of building material weld joint | |
| CN111145175B (en) | SMT welding spot defect detection method based on iForest model verification | |
| CN115272316B (en) | Intelligent detection method for welding quality of battery cover based on computer vision | |
| CN115239704B (en) | Accurate detection and repair method for wood surface defects | |
| CN109191459B (en) | Automatic identification and rating method for continuous casting billet macrostructure center segregation defect | |
| CN102175700B (en) | Method for detecting welding seam segmentation and defects of digital X-ray images | |
| CN115082418B (en) | Precise identification method for automobile parts | |
| CN109215009B (en) | Continuous casting billet surface image defect detection method based on deep convolution neural network | |
| CN107607554A (en) | A kind of Defect Detection and sorting technique of the zinc-plated stamping parts based on full convolutional neural networks | |
| CN117152161B (en) | Shaving board quality detection method and system based on image recognition | |
| WO2021168733A1 (en) | Defect detection method and apparatus for defect image, and computer-readable storage medium | |
| CN111815555A (en) | Metal additive manufacturing image detection method and device combining anti-neural network with local binary | |
| CN109886960A (en) | The method of glass edge defects detection based on machine vision | |
| CN109682839A (en) | A kind of metal arc Surface Flaw online test method | |
| CN114219794B (en) | Method and system for evaluating surface quality of shaving board based on machine vision | |
| CN115719332A (en) | Welding quality detection method | |
| CN114897908B (en) | Machine vision-based method and system for analyzing defects of selective laser powder spreading sintering surface | |
| CN113240623A (en) | Pavement disease detection method and device | |
| CN110473174B (en) | Method for calculating accurate number of pencils based on image | |
| TW201512649A (en) | Method of chip detects inspecting, system therefor, and computer program product thereof | |
| CN114820625A (en) | Automobile top block defect detection method | |
| CN114897855A (en) | Method for judging defect type based on X-ray picture gray value distribution | |
| CN110781913A (en) | Zipper cloth belt defect detection method | |
| CN114429445A (en) | PCB defect detection and identification method based on MAIRNet | |
| CN115597494B (en) | Precision detection method and system for prefabricated part preformed hole based on point cloud |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| PE01 | Entry into force of the registration of the contract for pledge of patent right |
Denomination of invention: A Method for Automatic Inspection of Weld Quality in Building Materials Effective date of registration: 20230626 Granted publication date: 20221104 Pledgee: China Postal Savings Bank Limited by Share Ltd. Wenshang County sub branch Pledgor: Shandong Yichang Prefabricated Building Technology Co.,Ltd. Registration number: Y2023980045342 |
|
| PE01 | Entry into force of the registration of the contract for pledge of patent right |