CN118470015B - Visual detection method and system for production quality of titanium alloy rod - Google Patents
Visual detection method and system for production quality of titanium alloy rod Download PDFInfo
- Publication number
- CN118470015B CN118470015B CN202410926525.1A CN202410926525A CN118470015B CN 118470015 B CN118470015 B CN 118470015B CN 202410926525 A CN202410926525 A CN 202410926525A CN 118470015 B CN118470015 B CN 118470015B
- Authority
- CN
- China
- Prior art keywords
- gray
- region
- titanium alloy
- alloy rod
- crack
- 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.)
- Active
Links
- 229910001069 Ti alloy Inorganic materials 0.000 title claims abstract description 146
- 238000001514 detection method Methods 0.000 title claims abstract description 84
- 238000004519 manufacturing process Methods 0.000 title claims abstract description 53
- 230000000007 visual effect Effects 0.000 title abstract description 16
- 238000012545 processing Methods 0.000 claims abstract description 41
- 238000009826 distribution Methods 0.000 claims abstract description 21
- 238000000034 method Methods 0.000 claims description 34
- 238000004364 calculation method Methods 0.000 claims description 18
- 230000011218 segmentation Effects 0.000 claims description 17
- 238000011179 visual inspection Methods 0.000 claims description 12
- 238000010606 normalization Methods 0.000 claims description 9
- 238000003709 image segmentation Methods 0.000 claims description 8
- 238000012423 maintenance Methods 0.000 claims description 5
- 238000000605 extraction Methods 0.000 claims description 2
- 238000013527 convolutional neural network Methods 0.000 description 7
- 239000002245 particle Substances 0.000 description 5
- 230000000694 effects Effects 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 3
- 238000000149 argon plasma sintering Methods 0.000 description 3
- 230000007547 defect Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000001914 filtration Methods 0.000 description 3
- 238000007689 inspection Methods 0.000 description 3
- 230000003746 surface roughness Effects 0.000 description 3
- RTAQQCXQSZGOHL-UHFFFAOYSA-N Titanium Chemical compound [Ti] RTAQQCXQSZGOHL-UHFFFAOYSA-N 0.000 description 2
- 238000005242 forging Methods 0.000 description 2
- 239000012634 fragment Substances 0.000 description 2
- 239000010936 titanium Substances 0.000 description 2
- 229910052719 titanium Inorganic materials 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 239000000956 alloy Substances 0.000 description 1
- 238000010924 continuous production Methods 0.000 description 1
- 230000007797 corrosion Effects 0.000 description 1
- 238000005260 corrosion Methods 0.000 description 1
- 239000000428 dust Substances 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000001125 extrusion Methods 0.000 description 1
- 239000010419 fine particle Substances 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 229910052751 metal Inorganic materials 0.000 description 1
- 239000002184 metal Substances 0.000 description 1
- 238000000465 moulding Methods 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000003825 pressing Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 239000002994 raw material Substances 0.000 description 1
Abstract
The invention relates to the technical field of image processing of visual detection, in particular to a visual detection method and a visual detection system for production quality of a titanium alloy rod. According to the technical scheme, a target image of the titanium alloy rod is obtained, wherein the target image is an image of the end face of the titanium alloy rod after turning, and target characteristics of different gray-scale symmetrical diagonal distribution of each gray-scale area in the end face turning center of the titanium alloy rod exist on the target image; the target image can be subjected to region division according to target features on the target image so as to obtain a plurality of groups of symmetrically and diagonally distributed gray scale regions; further, end face crack identification of the titanium alloy rod is implemented according to each group of symmetrically and diagonally distributed gray areas, so that a crack detection result of the titanium alloy rod can be rapidly and accurately obtained; and the production equipment of the titanium alloy rod is correspondingly maintained according to the crack detection result, so that the influence of production equipment factors on cracks is reduced, and the efficiency and the accuracy are considered when the titanium alloy rod is subjected to crack detection.
Description
Technical Field
The invention relates to the technical field of image processing of visual detection, in particular to a visual detection method and a visual detection system for production quality of a titanium alloy rod.
Background
Titanium alloys are widely used in various fields because of their high strength, good corrosion resistance, high heat resistance, and bio-philic properties. Because of the special application of the titanium product, the requirements on the surface quality of the titanium product are very high, and defects such as pit pressing, scratch, crack, needle hole and the like are not allowed to occur, and the quality problems can seriously affect the performance and the application of the titanium alloy rod.
The titanium alloy rod is used as an important component of the titanium alloy material, and in the titanium alloy production quality detection process, it is important to grasp the production quality of the titanium alloy rod. When a crack occurs on the surface of a titanium alloy rod, the crack typically appears as a linear or slightly curved linear defect. In the prior art, cracks can be identified through edge information by a computer vision technology, but the types of the cracks are complex because of the special shape and the specificity of the turning scene of the titanium alloy rod, and the accurate detection is not easy to carry out.
Disclosure of Invention
The visual detection method and the visual detection system for the production quality of the titanium alloy rod can give consideration to efficiency and accuracy when the titanium alloy rod is subjected to crack detection.
The embodiment of the invention provides the following scheme:
in a first aspect, an embodiment of the present invention provides a visual inspection method for production quality of a titanium alloy rod, where the method includes:
acquiring a target image of the titanium alloy rod, wherein the target image is an image of the end face of the titanium alloy rod subjected to turning;
dividing the target image according to target features on the target image to obtain a plurality of groups of symmetrically and diagonally distributed gray areas, wherein the target features are features that the gray areas are symmetrically and diagonally distributed in different gray levels by using the end face turning center of the titanium alloy rod;
Performing end surface crack identification on the titanium alloy rod according to each group of symmetrically and diagonally distributed gray areas to obtain a crack detection result of the titanium alloy rod;
And correspondingly maintaining production equipment of the titanium alloy rod according to the crack detection result.
In an alternative embodiment, the method for dividing the target image into regions according to the target features on the target image to obtain multiple groups of symmetrically diagonally distributed gray scale regions includes:
obtaining a region dividing line of each gray region on the target image according to gray values corresponding to different gray regions characterized by the target features;
and obtaining each group of symmetrically diagonally distributed gray scale areas according to the segmentation result of all the area segmentation lines on the upper end surface area of the target image.
In an alternative embodiment, obtaining the region dividing line of each gray scale region on the target image according to the gray scale values corresponding to different gray scale regions characterized by the target feature includes:
image segmentation is carried out on the target image by a face turning center according to the target characteristics so as to segment the target image into a first image and a second image;
Obtaining a region dividing line of each gray scale region on the first image according to different gray scale values presented by adjacent gray scale regions on the first image;
And extending the region dividing line of each gray scale region on the first image to the second image to obtain the region dividing line of each gray scale region on the target image.
In an alternative embodiment, obtaining the region dividing line of each gray scale region on the first image according to the different gray scale values presented by the adjacent gray scale regions on the first image includes:
Extracting gray values of all pixel points on the first image to obtain gray data of each radial line on the first image;
respectively carrying out mean value calculation, difference value calculation and the same gray value statistics according to the gray data of each radial line to obtain an average gray value, a maximum gray difference value and a maximum frequency gray of each radial line;
obtaining gray level similarity of all adjacent radial lines according to the average gray level value and the maximum frequency gray level of the adjacent radial lines;
obtaining the gray level similarity of all adjacent radial lines according to the maximum gray level difference value, the gray level similarity degree and a preset gray level deviation threshold value;
When the gray level similarity of the adjacent radial lines is larger than a preset similarity threshold value, determining that the adjacent radial lines are in the same gray level region;
And when the gray level similarity of the adjacent radial lines is smaller than or equal to a similarity threshold value, determining that the adjacent radial lines are in different gray level areas so as to obtain an area dividing line of each gray level area on the first image.
In an alternative embodiment, the obtaining the gray level similarity of all adjacent radial lines according to the maximum gray level difference value, the gray level similarity degree and the preset gray level deviation threshold value includes:
According to the formula:
Obtaining gray level similarity of radial line k and radial line k+1 Wherein, the method comprises the steps of, wherein,For the gray level deviation threshold of radial line k from radial line k +1,To the extent that radial line k is similar to the gray scale of radial line k +1,For the maximum gray difference value of the radial line k, norm () is a normalization function.
In an alternative embodiment, performing end face crack identification of a titanium alloy rod according to each set of symmetrically diagonally distributed gray scale regions to obtain a crack detection result of the titanium alloy rod, comprising:
Obtaining an average gray level difference value, a gray level variance difference value and region roughness of each gray level region according to the gray level data of each gray level region;
Obtaining the region similarity of each gray region according to the average gray difference value and the gray variance difference value;
When the region similarity of the gray scale region of the current group is smaller than a preset similarity threshold, judging whether the region roughness of the gray scale region of the current group is larger than a preset roughness threshold;
When any area of the current group of gray areas is larger than the roughness threshold, determining that a crack exists in the corresponding gray area;
and outputting the gray scale region with the cracks as a crack detection result of the titanium alloy rod.
In an alternative embodiment, obtaining the region roughness of each gray region from the gray data of each gray region includes:
Performing corner detection on the gray data of each gray area to obtain the number of corners of each gray area in each group of gray areas;
inputting the number of corner points and gray data of each gray area into a preset roughness processing model;
and obtaining the region roughness of each gray region according to the output result of the roughness processing model.
In an alternative embodiment, outputting the gray scale region where the crack exists as a crack detection result of the titanium alloy rod includes:
image segmentation is carried out on the gray scale area with the cracks according to a preset crack segmentation threshold value so as to obtain the distribution distance between the pixel points of each crack;
when the distribution distance between the crack pixel points is in a preset range, determining that the crack detection result is a branch crack;
and when the distribution distance between the crack pixel points is not in a preset range, determining that the crack detection result is a discontinuous crack.
In an alternative embodiment, before the region division of the target image according to the target feature on the target image, the method further includes:
carrying out noise reduction treatment on the target image to obtain a target image after the noise reduction treatment;
Inputting the target image after the noise reduction treatment into a preset semantic segmentation model to segment an end face part and a background part of the target image after the noise reduction treatment;
the target image divided into the end face portions is determined as a target image in which region division is performed based on the target features.
In a second aspect, an embodiment of the present invention further provides a visual inspection system for production quality of a titanium alloy rod, where the system includes:
The image shooting terminal is used for outputting a target image of the titanium alloy rod, wherein the target image is an image of the end face of the titanium alloy rod subjected to turning;
The data processing terminal is connected with the image shooting terminal; the data processing terminal is used for dividing the target image according to target characteristics on the target image to obtain a plurality of groups of gray areas which are symmetrically and diagonally distributed, wherein the target characteristics are characteristics that the gray areas are symmetrically and diagonally distributed in different gray levels by the end face turning center of the titanium alloy rod;
the data processing terminal is also used for carrying out end face crack identification of the titanium alloy rod according to each group of symmetrically and diagonally distributed gray areas so as to obtain a crack detection result of the titanium alloy rod;
the data processing terminal is also used for carrying out corresponding maintenance on production equipment of the titanium alloy rod according to the crack detection result.
Compared with the prior art, the visual detection method and the visual detection system for the production quality of the titanium alloy rod have the following advantages:
According to the technical scheme, the target image of the titanium alloy rod is obtained, wherein the target image is an image of the end face of the titanium alloy rod after turning, and as the turning face has bright reflected light and certain light scattering, the target image has target characteristics that all gray areas are symmetrically distributed diagonally in different gray levels by the end face turning center of the titanium alloy rod; the target image can be subjected to region division according to target features on the target image so as to obtain a plurality of groups of symmetrically and diagonally distributed gray scale regions; further, end face crack identification of the titanium alloy rod is implemented according to each group of symmetrically and diagonally distributed gray areas, so that a crack detection result of the titanium alloy rod can be rapidly and accurately obtained; the production equipment of the titanium alloy rod is correspondingly maintained according to the crack detection result, so that the influence of production equipment factors on cracks is reduced, and efficiency and accuracy can be considered when the titanium alloy rod is subjected to crack detection.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions and advantages of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are only some embodiments of the invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of a visual inspection method for production quality of a titanium alloy rod according to an embodiment of the present invention;
FIG. 2 is a target image of a titanium alloy rod according to one embodiment of the present invention;
FIG. 3 is a schematic diagram of a visual inspection system for production quality of a titanium alloy rod according to an embodiment of the present invention.
Detailed Description
In order to further describe the technical means and effects adopted by the present invention to achieve the preset purpose, the following detailed description is given below of the specific implementation, structure, features and effects of the technical solution according to the present invention with reference to the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" means that the embodiments are not necessarily the same. Furthermore, the particular features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
The invention provides a visual detection method and a visual detection system for production quality of a titanium alloy rod, which are specifically described below with reference to the accompanying drawings.
Referring to fig. 1, a flowchart of a visual inspection method for production quality of a titanium alloy rod according to an embodiment of the present invention is shown, where the inspection method may be applied to an inspection terminal of a titanium alloy rod or a control terminal of a production device, and the inspection method may be executed, and is not limited herein. The detection method comprises the following steps:
S11, acquiring a target image of the titanium alloy rod, wherein the target image is an image of the end face of the titanium alloy rod after turning.
Specifically, the titanium alloy rod is a metal section bar which is produced and is off-line by production equipment, the production equipment can be extrusion molding production line equipment or forging equipment for forging and molding, the production equipment needs to be provided with various parameters, and continuous production of the titanium alloy rod is implemented by titanium alloy raw materials. The cross section of the titanium alloy rod can be rectangular, circular or elliptical, and after the titanium alloy rod is produced and is taken off line, in order to check the production quality, the end face of the titanium alloy rod is turned by a lathe to remove an oxide layer on the end face so as to facilitate the quality detection. The target image can be obtained by photographing the end face of the turned titanium alloy rod through an image photographing terminal, and the image photographing terminal can be a mobile phone, a quality detection camera and the like.
Because the shooting environment of the titanium alloy rod is located in the production workshop, the shooting environment is influenced by various factors such as dust and light in the production workshop, noise exists on a target image, and the accuracy of a subsequent detection result is influenced. Based on this, in a specific embodiment, preprocessing is further required for the target image, specifically including:
and carrying out noise reduction processing on the target image to obtain the target image after the noise reduction processing. The method can select a proper noise reduction processing mode based on actual demands, such as Gaussian filtering, mean filtering, median filtering and the like, and image noise reduction can be implemented, so that the image quality can be improved, noise can be removed or lightened, the definition and detail performance of the image can be improved, the influence due to abnormal factors in the shooting process can be reduced, and the subsequent image processing and analysis can be facilitated.
After the image noise reduction processing is completed, inputting the noise-reduced target image into a preset semantic segmentation model to perform segmentation processing on an end face part and a background part of the noise-reduced target image. The target image may be semantically segmented using a CNN (Convolutional Neural Networks, convolutional neural network) model of the encoding-decoding structure. The split labels are divided into two categories, namely end face labels and background labels of the titanium alloy rods. The method is pixel-level classification, namely, corresponding labels need to be marked on all pixels in the image. Pixels belonging to the end face of the titanium alloy rod, and the value of the pixels is marked as 1; the pixel belonging to the background has a value marked as 0, the CNN model performs segmentation processing by adopting a cross entropy loss function, and the target image segmented into end face parts is determined as the target image for implementing region division based on the target characteristics.
Since the direct-shooting target image includes the end face portion and the background portion, the noise reduction and the segmentation processing described above are performed, and the processing efficiency and accuracy of the subsequent image can be improved. Referring to fig. 2, fig. 2 is a schematic diagram of a target image, in which a cross section of a titanium alloy rod is circular, a circular region in a coordinate system XOY is an end face portion of the titanium alloy rod, an image other than the end face portion is a background portion, and a subsequent process is performed based on the target image of the end face portion.
S12, dividing the target image according to target features on the target image to obtain a plurality of groups of symmetrically and diagonally distributed gray areas, wherein the target features are features that the gray areas are symmetrically and diagonally distributed in different gray levels by using the end face turning center of the titanium alloy rod.
Specifically, the end face of the titanium alloy rod is turned, so that the surface of the end face has bright reflected light under the irradiation of light, and meanwhile, certain light scattering exists. For the titanium alloy rod, under the irradiation of normal light, the collected target image on the surface of the titanium alloy rod shows certain regularity, specifically, the target image is divided into a plurality of areas, and the brightness of two areas which are diagonal areas has a great degree of similarity, as shown in fig. 2. According to the rule expression, the gray level of the pixel points in the target image is combined, the target image of the titanium alloy rod is analyzed, the area where the crack condition possibly exists is judged, the suspicious area is further accurately detected through the smoothness characteristic of the image surface, and therefore the production quality detection of the titanium alloy rod is completed according to the crack condition in the defect area after the accurate detection.
The target image is divided into areas, namely the end face image of the titanium alloy rod is divided into areas, and grouping is completed according to a diagonal rule. When the region division is performed, the target image may be divided based on the gradation values of the different regions to divide the end face of the titanium alloy rod into a plurality of regions. Taking a titanium alloy rod with a circular cross section as an example, the end face of the titanium alloy rod can be divided into a plurality of sector areas, the central angle vertexes of all the sector areas are the end face turning center of the titanium alloy rod, the symmetrical diagonal distribution is that two sector areas are the same group of gray areas, the gray values presented by the two sector areas are the same, and after the target image is segmented, a plurality of groups of gray areas which are symmetrically and diagonally distributed can be obtained.
Illustratively, step S12 includes substeps S12-1 through S12-2, which specifically include:
S12-1, obtaining an area dividing line of each gray area on the target image according to gray values corresponding to different gray areas represented by the target features. The gray value of each pixel point can be extracted based on the target characteristic represented by the target image, the region formed by the pixel points covered by the near gray value is divided into the same gray region, and the boundary line of the adjacent gray regions is the region dividing line of the gray region.
In practical applications, the number of pixels in the target image is large, and if gray scale calculation is performed one by one, a large amount of processing time is required, resulting in a large amount of redundant calculation in the image processing process. Based on this, in a specific embodiment, substep S12-1 includes S12-1-1 to S12-1-3, specifically including:
S12-1-1, performing image segmentation on the target image by using a face turning center according to the target characteristics so as to segment the target image into a first image and a second image. With continued reference to fig. 2, taking titanium alloy rod detection with a circular cross section as an example, the acquisition target image may be considered to belong to a circle, and the coordinate system XOY may be constructed based on the center of the circle as the origin of the coordinate axes, with the horizontal direction as the X axis and the vertical direction as the Y axis. And (3) image segmentation is carried out on the target image through a coordinate system XOY, and the target image is segmented into a first image and a second image in an X axis or a Y axis, wherein the first image and the second image are mirror images.
S12-1-2, obtaining the region dividing line of each gray scale region on the first image according to different gray scale values presented by the adjacent gray scale regions on the first image. And dividing the region based on the gray values of the adjacent gray areas on the first image to obtain a dividing line between the adjacent gray areas, namely a region dividing line. The target pixel point can be taken in any quadrant of the coordinate system XOY far away from the coordinate origin point, and the gray level set of all the pixel points on the connecting line between the target pixel point and the coordinate origin point is obtained:
And carrying out gray scale region division based on gray scale data characterized by the gray scale set to obtain a region division line of each gray scale region on the first image.
Illustratively, step S12-1-2 includes:
And firstly, extracting gray values of all pixel points on the first image to obtain gray data of each radial line on the first image. The gray level set is the gray level data of a radial line, and M is the number of pixel points on one radial line. The gray value extraction is performed on all pixel points on the first image, so that gray data of each radial line can be obtained, and the gray data of each radial line is characterized as a corresponding gray set.
And secondly, respectively carrying out mean value calculation, difference value calculation and the same gray value statistics according to the gray data of each radial line to obtain the average gray value, the maximum gray difference value and the maximum frequency gray of each radial line. The gray values of the pixel points in the same area are approximately the same, and the gray data of each radial line can be analyzedGray scale set of radial linesObtaining the gray value distribution range of each radial lineAnd the corresponding number of pixelsAnd the gray value corresponding to the most pixel points. Carrying out average value calculation on the gray data of each radial line to obtain an average gray value; maximum gray valueAnd minimum gray valueThe difference value of (2) is the maximum gray level difference value; the maximum frequency gray scale is the gray scale value with the maximum number of corresponding pixel points in the gray scale data of each radial line。
And thirdly, obtaining the gray level similarity degree of all the adjacent radial lines according to the average gray level value and the maximum frequency gray level of the adjacent radial lines. The target image has a gray level similar relation of diagonal areas, and for adjacent radial lines, the target image may exist in the same area, and if the gray level distribution ranges on the connecting lines are similar and the distribution conditions are similar, the target image belongs to the same gray level area. Therefore, it is possible to determine whether or not adjacent radial lines are the same gray scale region based on the gray scale proximity degree of the adjacent radial lines. The formula can be based on:
obtaining the gray level similarity degree of the radial line k and the radial line k+1 In which, in the process,Representing radial linesIs used for the color filter,Represents the average gray value of radial line k +1,The gray value of the k-th radial line having the largest number of corresponding pixels in the gray data,The gray value corresponding to the most pixel point in the gray data of the (k+1) th radial line,Representing absolute value symbols.
It can be appreciated that the above formula utilizes the maximum gray value difference between the radial line k and the corresponding pixel point on the radial line k+1And the average gray value of bothAnd (5) performing calculation. For the same gray scale region on the titanium alloy rod, the corresponding average gray scale values will be approximately the same, and the gray scale value distribution will be similar. The smaller the maximum gray value difference of the corresponding pixel points on the radial line is, the smaller the average gray value is, the larger the corresponding calculation result is, and the gray similarity degree of all adjacent radial lines is accurately calculated.
Fourth, according to the maximum gray difference value, the gray similarity degree and the preset gray deviation threshold value, gray similarity of all adjacent radial lines is obtained. The similarity calculation model can be configured based on actual requirements, the maximum gray difference value, the gray similarity degree and the gray deviation threshold value are input into the similarity calculation model, the gray similarity of the adjacent radial lines is obtained based on the output result of the similarity calculation model, and the gray similarity represents the gray similarity degree between the adjacent radial lines.
As an example, the formula may be according to:
Obtaining gray level similarity of radial line k and radial line k+1 Wherein, the method comprises the steps of, wherein,For the gray level deviation threshold of radial line k from radial line k +1,To the extent that radial line k is similar to the gray scale of radial line k +1,For the maximum gray difference value of the radial line k, norm () is a normalization function. It will be appreciated that for the target image of the titanium alloy rod, the overall colour will be biased towards white, and the normally larger grey values, so that the calculation is performed taking into account the maximum grey difference. For the minimum gray value, there are fewer pixel points and smaller gray values because there may be crack features, but they have no substantial effect on the division of gray areas.
It should be noted that, because the gray values of the pixels in the same area are similar, however, due to the influence of other factors in the shooting process, such as light factors, there may be a certain deviation in the gray values of the pixels on the connecting line in the picture. For this purpose, a gray level deviation threshold is set, which allows gray level value differences of the gray level deviation threshold to exist in the same gray level region. The gray level deviation threshold can be set based on experience of technicians or calibration experiments without affecting accuracy of gray level region division, such as setting gray level deviation threshold5.
And fifthly, when the gray level similarity of the adjacent radial lines is larger than a preset similarity threshold value, the gray level characteristics represented by the adjacent radial lines are very similar, and the adjacent radial lines are determined to be in the same gray level region. When the gray level similarity of the adjacent radial lines is smaller than or equal to a similarity threshold value, the difference of gray level characteristics represented by the adjacent radial lines is indicated, the adjacent radial lines are determined to be in different gray level areas, and after similarity comparison is carried out on all the adjacent radial lines, each gray level area on the first image can be obtained. The similarity threshold may be set based on actual requirements, for example, 0.7, or may be another value, which is not particularly limited herein.
Thus, division of each gradation region on the first image is completed, and a region dividing line of each gradation region on the first image is obtained based on the boundary line of the adjacent gradation regions.
S12-1-3, extending the region dividing line of each gray scale region on the first image to the second image so as to obtain the region dividing line of each gray scale region on the target image. And (3) extending the region dividing line on the first image based on the diagonal rule of each gray scale region on the target image, so that the region dividing line can be extended to the outer contour of the second image, and the region dividing line of each gray scale region on the target image can be obtained.
S12-2, according to the segmentation result of all the region segmentation lines on the upper end surface region of the target image, each group of symmetrically and diagonally distributed gray scale regions are obtained. The target image can be divided into a plurality of gray scale areas through the area dividing line, the gray scale characteristics represented by the adjacent gray scale areas are different, the gray scale areas which are symmetrically and diagonally distributed are the same group of gray scale areas, and the gray scale characteristics represented by the same group of gray scale areas are the same. By implementing the gray scale region division and the region division line extension of the first image, the calculation redundancy of data in the gray scale region division process can be effectively reduced, and the processing efficiency of the titanium alloy rod for implementing crack detection is improved.
Thus, the grouping of the regions of the target image of the titanium alloy rod is completed.
S13, carrying out end face crack identification on the titanium alloy rods according to each group of symmetrically and diagonally distributed gray scale areas so as to obtain crack detection results of the titanium alloy rods.
Specifically, since the gray scale of the same group of gray scale regions has the same gray scale, the existence of the crack can be primarily judged according to the similarity degree of two different regions in the same group of gray scale regions. The end face crack identification is to analyze the gray level expression of the same group of areas and accurately judge the gray level areas with crack conditions by combining the characteristic smoothness characteristics. The training of end face crack identification can be implemented based on the CNN model, and when the training has higher accuracy, the end face crack identification is implemented based on the CNN model, and the output result of the CNN model is determined as the crack detection result of the titanium alloy rod.
Illustratively, step S13 includes substeps S13-1 through S13-5, which specifically include:
S13-1, obtaining an average gray level difference value, a gray level variance difference value and region roughness of each gray level region according to the gray level data of each gray level region. The average gray level difference value can be obtained based on the average gray level difference of each gray level region; the gray variance difference is derived based on the gray variance of each gray region. The region roughness characterizes the smoothness of the gray scale region, and the presence of cracks can lead to an increase in roughness of the titanium alloy rod surface, as crack edges are often irregular and may be accompanied by tiny fragments or broken grains, thus allowing further precision of the crack region depending on the roughness exhibited by the image within the gray scale region.
The method for acquiring the region roughness comprises the substeps S13-1-1 to S13-1-3, specifically:
S13-1-1, carrying out corner detection on the gray level data of each gray level region to obtain the number of corners of each gray level region in each group of gray level regions. When cracks exist on the surface of the titanium alloy rod, tiny fragments and other small particles exist on two sides of the cracks, so that the roughness of the surface of the titanium alloy rod is improved, and the larger the number of the small particles on the crack area is, the larger the roughness is reflected. Because the color difference between the small particles on the surface of the titanium alloy rod and the edge area is larger, the corner detection operator of SIFT (Scale-INVARIANT FEATURE TRANSFORM, scale invariant feature transform) corner detection operator can be used for detecting the corner of each gray Scale area, and the quantity of the corner in each gray Scale area is obtained.
S13-1-2, inputting the number of corner points of each gray scale area and gray scale data into a preset roughness processing model. Similarly, the roughness processing model can be configured based on actual requirements, and the region roughness can be accurately calculated based on the number of corner points and gray data.
Since the pixels on the small protruding particles behave differently from the surrounding neighboring pixels under light irradiation, analysis can be performed by selecting a plurality of neighboring pixels as a whole, for example, 9 (3×3) neighboring pixels as a whole. As an example, the formula may be according to:
Obtaining the region roughness of each gray region In which, in the process,Represent the firstGroup gray level region inThe first of the areasGray values of individual pixels.Represent the firstGroup gray level region inThe first of the areasAverage gray values of 8 neighborhood pixels around each pixel.Represent the firstGroup IIIAverage gradient value of individual regions.Is shown in the firstGroup IIIThe gradient value of the pixels of each region is smaller than the number of the pixels of 10.Represent the firstGroup IIICorner number of corner detection is performed for each region. norm ()' is a normalization function.
It can be understood that in the above calculation, when the difference between the gray values of the pixel points and the surrounding 8 neighboring pixel points is larger, the gradient value of the pixel points in the whole area is larger, and the smaller the number of the pixel point gradient values is smaller than the set threshold value, the more fine particles exist on the titanium alloy rod area, and the roughness of the area is larger. The greater the number of corner points detected in the region, the greater the likelihood that there is a greater number of micro-chips in the region, i.e., the greater the roughness of the region.
S13-1-3, obtaining the regional roughness of each gray region according to the output result of the roughness processing model. The marking can be performed based on each gray scale region, the marking result of each gray scale region and the corresponding region roughness are correspondingly stored, and the step S13-2 is performed after the region roughness of each gray scale region is obtained.
S13-2, obtaining the region similarity of each group of gray regions according to the average gray difference value and the gray variance difference value. Also, the region similarity between two gray regions in each group of gray regions can be calculated based on the similarity calculation model. For two regions belonging to the same group on the surface of the titanium alloy rod, the titanium alloy rod should generally be similar in overall gray scale. But when a certain area has a crack, it may exhibit a certain degree of color difference with the corresponding area. Thereby obtaining the firstThe degree of regional similarity of two regions within a group:
In the formula, Representing the degree of regional similarity of two regions in the same set of gray scale regions,Representing the average gray level difference of two regions in the same set of gray level regions,The difference of gray variance of pixel points in the two areas is represented, and norm () is a normalization function.
It should be further noted that, when the average gray differences of the pixels in the two regions in the same gray region are equalThe smaller the gray variance differenceThe smaller the gray scale features representing the same set of two gray scale regions on the surface of the titanium alloy rod, the closer the gray scale features are to each other, further illustrating that both regions are either normal-performing regions or regions where cracks exist. Further analysis is required.
S13-3, when the region similarity of the gray scale region of the current group is smaller than a preset similarity threshold, the fact that the similarity of the two gray scale regions of the gray scale region of the current group is very high is indicated, and whether the region roughness of the gray scale region of the current group is larger than the preset roughness threshold is continuously judged.
S13-4, when any region of the gray scale regions in the current group is larger than the roughness threshold value, the fact that the surface roughness of the gray scale regions larger than the roughness threshold value is larger is indicated, and the fact that cracks exist in the corresponding gray scale regions is determined.
S13-5, outputting a gray scale region with cracks as a crack detection result of the titanium alloy rod.
The crack determination process will be specifically described below. Setting a similarity thresholdRoughness thresholdWhen the regions are similarAnd (2) andAnd (3) withThe similarity of two gray areas of the same group is high, the surface roughness is low, and the two gray areas are normal areas; when (when)If (if)，Indicating that the two have low similarity and different surface roughness degree, the firstThe 1 st region in the group is a crack region, and the 2 nd region is a normal region. Other cases are also analyzed and will not be further described herein.
So far, all crack areas on the titanium alloy rod are obtained through the data processing mode.
Through the treatment, all crack areas on the titanium alloy rod can be obtained, but the production process of the titanium alloy rod is complex, different crack types need to be correspondingly adjusted parameters to be different, and if only the crack areas are determined, accurate process parameter adjustment cannot be implemented. Based on this, in a specific embodiment, substep S13-5 includes S13-5-1 to S13-5-2, comprising in particular:
s13-5-1, performing image segmentation on the gray scale region with the cracks according to a preset crack segmentation threshold value so as to obtain the distribution distance between the pixel points of each crack. Because crack features and some small particle features exist in the crack region, the distribution of crack pixels in the image after threshold segmentation can reflect the type of the crack, and the type of the crack is judged according to the continuity of pixel coordinates on the crack. Reconstructing an XY coordinate system on a gray scale region with cracks to obtain coordinates of pixel points on the cracks . And selecting a first pixel point from left to right as a starting pixel point from top to bottom, and calculating the distance between the first pixel point and the pixel point with the nearest surrounding distance. Iterative acquisition of the firstDistance between each pixel point and the next pixel pointThe distribution distance between the pixel points of each crack can be obtained.
S13-5-2, when the distribution distance between the crack pixel points is in a preset range, indicating that the crack presents a branch structure, and determining that the crack detection result is a branch crack. When the distribution distance between the crack pixel points is not in the preset range, the crack forms are dispersed, and the crack detection result is determined to be discontinuous cracks. The preset range can be set based on actual requirements, for example. When measuring the distribution distance between the crack pixels, a distance threshold K3 may be set, allowing for a measurement error of k3=0.5.
Thus, a crack detection result of the titanium alloy rod is obtained, and the process proceeds to step S14.
And S14, correspondingly maintaining production equipment of the titanium alloy rod according to the crack detection result.
Specifically, corresponding maintenance of the production equipment can be implemented based on actual crack detection results, for example, cracks exist in all continuous 5 gray areas, the parameter setting of the production equipment is possibly unreasonable, or the production equipment has faults, so that the problem of cracks exists in the production of the titanium alloy rod, an alarm can be sent to on-site staff to remind the staff to analyze the corresponding problems according to different crack conditions, the maintenance of the production equipment is facilitated, and the product quality of the titanium alloy rod is ensured.
Of course, the corresponding crack detection result may be output based on the specifically detected branch crack or discontinuous crack, and the worker may implement the corresponding production equipment maintenance strategy based on the specific crack type.
Based on the same technical conception as the detection method, the embodiment of the invention also provides a visual detection system for the production quality of the titanium alloy rod, referring to fig. 3, and fig. 3 is a schematic structural diagram of the detection system. The system includes an image capturing terminal 301 and a data processing terminal 302.
The image capturing terminal 301 is configured to output a target image of the titanium alloy rod, where the target image is an image of the end surface of the titanium alloy rod after turning.
The data processing terminal 302 is connected to the image capturing terminal 301; the data processing terminal 302 is configured to divide a target image into regions according to target features on the target image, so as to obtain multiple groups of symmetrically and diagonally distributed gray areas, where the target features are features that each gray area is symmetrically and diagonally distributed in different gray levels by using a turning center of an end face of the titanium alloy rod.
The data processing terminal 302 is further configured to perform end face crack identification of the titanium alloy rod according to each set of symmetrically diagonally distributed gray scale regions, so as to obtain a crack detection result of the titanium alloy rod.
The data processing terminal 302 is further used for correspondingly maintaining production equipment of the titanium alloy rod according to the crack detection result.
The data processing terminal 302 may also be used to process related processing steps in the above detection method, which are not described in detail herein.
The technical scheme provided by the embodiment of the invention has at least the following technical effects or advantages:
1. The method comprises the steps of obtaining a target image of a titanium alloy rod, wherein the target image is an image of the end face of the titanium alloy rod after turning, and because the turning face has bright reflected light and certain light scattering, the target image has target characteristics that all gray areas are symmetrically distributed diagonally in different gray levels by the end face turning center of the titanium alloy rod; the target image can be subjected to region division according to target features on the target image so as to obtain a plurality of groups of symmetrically and diagonally distributed gray scale regions; further, end face crack identification of the titanium alloy rod is implemented according to each group of symmetrically and diagonally distributed gray areas, so that a crack detection result of the titanium alloy rod can be rapidly and accurately obtained; and the production equipment of the titanium alloy rod is correspondingly maintained according to the crack detection result, so that the influence of production equipment factors on cracks is reduced, and the efficiency and the accuracy are considered when the titanium alloy rod is subjected to crack detection.
2. The embodiment of the invention implements visual detection of the production quality of the titanium alloy rod based on the target characteristics of the titanium alloy rod after turning, gives consideration to detection efficiency and accuracy, can be suitable for field fault treatment of production equipment, helps staff master the quality detection result of products, and ensures the production quality of the titanium alloy rod.
It should be noted that: the sequence of the embodiments of the present invention is only for description, and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
In this specification, each embodiment is described in a progressive manner, and identical and similar parts of each embodiment are all referred to each other, and each embodiment mainly describes differences from other embodiments.
Claims (7)
1.A visual inspection method for production quality of a titanium alloy rod, which is characterized by comprising the following steps:
Acquiring a target image of a titanium alloy rod, wherein the target image is an image of the end face of the titanium alloy rod after turning;
Dividing the target image according to target features on the target image to obtain a plurality of groups of gray areas which are symmetrically and diagonally distributed, wherein the target features are features in which each gray area is symmetrically and diagonally distributed in different gray with the end face turning center of the titanium alloy rod;
Performing end surface crack identification of the titanium alloy rod according to each group of symmetrically and diagonally distributed gray scale areas to obtain a crack detection result of the titanium alloy rod;
performing corresponding maintenance on production equipment of the titanium alloy rod according to the crack detection result;
the method for identifying the end face cracks of the titanium alloy rod according to the gray scale areas distributed symmetrically and diagonally in each group to obtain the crack detection result of the titanium alloy rod comprises the following steps:
Obtaining an average gray level difference value, a gray level variance difference value and region roughness of each gray level region according to the gray level data of each gray level region;
Obtaining the region similarity of each gray region according to the average gray difference value and the gray variance difference value, wherein the specific formula comprises:
In the formula, Representing the degree of regional similarity of two regions in the same set of gray scale regions,Representing the average gray level difference of two regions in the same set of gray level regions,Representing the gray variance difference value of pixel points in the two areas, wherein norm () is a normalization function;
when the region similarity of the current group gray scale region is smaller than a preset similarity threshold, judging whether the region roughness of the current group gray scale region is larger than a preset roughness threshold;
when any one of the current gray scale areas is larger than the roughness threshold, determining that a crack exists in the corresponding gray scale area;
outputting a gray area with cracks as a crack detection result of the titanium alloy rod;
Obtaining the region roughness of each gray region according to the gray data of each gray region, including:
performing corner detection on the gray data of each gray area to obtain the number of corner points of each gray area in each group of gray areas;
inputting the number of corner points and gray data of each gray area into a preset roughness processing model;
obtaining the region roughness of each gray region according to the output result of the roughness processing model;
the outputting of the gray scale region where the crack exists as the crack detection result of the titanium alloy rod includes:
image segmentation is carried out on the gray scale area with the cracks according to a preset crack segmentation threshold value so as to obtain the distribution distance between the pixel points of each crack;
When the distribution distance between the crack pixel points is in a preset range, determining that the crack detection result is a branch crack;
when the distribution distance between the crack pixel points is not in a preset range, determining that the crack detection result is a discontinuous crack;
the roughness treatment model includes:
Is the first Group gray level region inThe area roughness of the individual areas is determined,Represent the firstGroup gray level region inThe first of the areasGray values of the individual pixels; Represent the first Group gray level region inThe first of the areasAverage gray value of 8 neighborhood pixels around each pixelRepresent the firstGroup IIIAverage gradient value of individual regionsIs shown in the firstGroup IIIThe number of pixels in each region having a gradient value of less than 10Represent the firstGroup IIICorner number of the corner detection is implemented in each region; norm ()' is a normalization function.
2. The visual inspection method for the production quality of the titanium alloy rod according to claim 1, wherein the dividing the target image into areas according to the target features on the target image to obtain a plurality of groups of symmetrically diagonally distributed gray areas comprises:
Obtaining an area dividing line of each gray area on the target image according to gray values corresponding to different gray areas represented by the target features;
And obtaining each group of symmetrically diagonally distributed gray scale areas according to the segmentation result of all the area segmentation lines on the upper end surface area of the target image.
3. The visual inspection method for the production quality of the titanium alloy rod according to claim 2, wherein the obtaining the region dividing line of each gray scale region on the target image according to the gray scale values corresponding to the different gray scale regions characterized by the target features comprises:
image segmentation is carried out on the target image by the end face turning center according to the target characteristics so as to segment the target image into a first image and a second image;
obtaining an area dividing line of each gray level area on the first image according to different gray level values presented by adjacent gray level areas on the first image;
And extending the region dividing line of each gray scale region on the first image to the second image so as to obtain the region dividing line of each gray scale region on the target image.
4. The visual inspection method for the production quality of the titanium alloy rod according to claim 3, wherein the step of obtaining the region dividing line of each gray scale region on the first image according to the different gray scale values represented by the adjacent gray scale regions on the first image comprises the following steps:
gray value extraction is carried out on all pixel points on the first image so as to obtain gray data of each radial line on the first image;
respectively carrying out mean value calculation, difference value calculation and the same gray value statistics according to the gray data of each radial line to obtain an average gray value, a maximum gray difference value and a maximum frequency gray of each radial line;
according to the average gray value and the maximum frequency gray of the adjacent radial lines, the gray approximation degree of all the adjacent radial lines is obtained, and the specific formula comprises:
obtaining the gray level similarity degree of the radial line k and the radial line k+1 In which, in the process,Representing radial linesIs used for the color filter,Represents the average gray value of radial line k +1,The gray value of the k-th radial line having the largest number of corresponding pixels in the gray data,The gray value corresponding to the most pixel point in the gray data of the (k+1) th radial line,Representing absolute value symbols;
Obtaining gray level similarity of all adjacent radial lines according to the maximum gray level difference value, the gray level similarity degree and a preset gray level deviation threshold value;
When the gray level similarity of the adjacent radial lines is larger than a preset similarity threshold value, determining that the adjacent radial lines are in the same gray level region;
And when the gray level similarity of the adjacent radial lines is smaller than or equal to the similarity threshold value, determining that the adjacent radial lines are in different gray level areas so as to obtain an area dividing line of each gray level area on the first image.
5. The visual inspection method for the production quality of the titanium alloy rod according to claim 4, wherein the obtaining the gray level similarity of all adjacent radial lines according to the maximum gray level difference value, the gray level similarity degree and a preset gray level deviation threshold value comprises:
According to the formula:
Obtaining gray level similarity of radial line k and radial line k+1 Wherein, the method comprises the steps of, wherein,For the gray level deviation threshold of radial line k from radial line k +1,To the extent that radial line k is similar to the gray scale of radial line k +1,For the maximum gray difference value of the radial line k, norm () is a normalization function,Representing absolute value symbols.
6. The visual inspection method for the production quality of the titanium alloy rod according to claim 1, wherein before the region division of the target image according to the target feature on the target image, the method further comprises:
carrying out noise reduction processing on the target image to obtain a noise-reduced target image;
Inputting the target image after the noise reduction treatment into a preset semantic segmentation model to segment an end face part and a background part of the target image after the noise reduction treatment;
and determining the target image divided into end face parts as a target image for performing region division based on the target features.
7. A visual inspection system for production quality of a titanium alloy rod, the system comprising:
The image shooting terminal is used for outputting a target image of the titanium alloy rod, wherein the target image is an image of the end face of the titanium alloy rod subjected to turning;
The data processing terminal is connected with the image shooting terminal; the data processing terminal is used for dividing the target image according to target features on the target image to obtain a plurality of groups of symmetrically and diagonally distributed gray areas, wherein the target features are features that the gray areas are symmetrically and diagonally distributed in different gray levels by the end face turning center of the titanium alloy rod;
the data processing terminal is also used for implementing end face crack identification of the titanium alloy rod according to each group of symmetrically and diagonally distributed gray scale areas so as to obtain a crack detection result of the titanium alloy rod;
The data processing terminal is also used for correspondingly maintaining production equipment of the titanium alloy rod according to the crack detection result;
the method for identifying the end face cracks of the titanium alloy rod according to the gray scale areas distributed symmetrically and diagonally in each group to obtain the crack detection result of the titanium alloy rod comprises the following steps:
Obtaining an average gray level difference value, a gray level variance difference value and region roughness of each gray level region according to the gray level data of each gray level region;
Obtaining the region similarity of each gray region according to the average gray difference value and the gray variance difference value, wherein the specific formula comprises:
In the formula, Representing the degree of regional similarity of two regions in the same set of gray scale regions,Representing the average gray level difference of two regions in the same set of gray level regions,Representing the gray variance difference value of pixel points in the two areas, wherein norm () is a normalization function;
when the region similarity of the current group gray scale region is smaller than a preset similarity threshold, judging whether the region roughness of the current group gray scale region is larger than a preset roughness threshold;
when any one of the current gray scale areas is larger than the roughness threshold, determining that a crack exists in the corresponding gray scale area;
outputting a gray area with cracks as a crack detection result of the titanium alloy rod;
Obtaining the region roughness of each gray region according to the gray data of each gray region, including:
performing corner detection on the gray data of each gray area to obtain the number of corner points of each gray area in each group of gray areas;
inputting the number of corner points and gray data of each gray area into a preset roughness processing model;
obtaining the region roughness of each gray region according to the output result of the roughness processing model;
the outputting of the gray scale region where the crack exists as the crack detection result of the titanium alloy rod includes:
image segmentation is carried out on the gray scale area with the cracks according to a preset crack segmentation threshold value so as to obtain the distribution distance between the pixel points of each crack;
When the distribution distance between the crack pixel points is in a preset range, determining that the crack detection result is a branch crack;
when the distribution distance between the crack pixel points is not in a preset range, determining that the crack detection result is a discontinuous crack;
the roughness treatment model includes:
Is the first Group gray level region inThe area roughness of the individual areas is determined,Represent the firstGroup gray level region inThe first of the areasGray values of the individual pixels; Represent the first Group gray level region inThe first of the areasAverage gray value of 8 neighborhood pixels around each pixelRepresent the firstGroup IIIAverage gradient value of individual regionsIs shown in the firstGroup IIIThe number of pixels in each region having a gradient value of less than 10Represent the firstGroup IIICorner number of the corner detection is implemented in each region; norm ()' is a normalization function.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410926525.1A CN118470015B (en) | 2024-07-11 | Visual detection method and system for production quality of titanium alloy rod |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410926525.1A CN118470015B (en) | 2024-07-11 | Visual detection method and system for production quality of titanium alloy rod |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118470015A CN118470015A (en) | 2024-08-09 |
| CN118470015B true CN118470015B (en) | 2024-11-15 |
Family
ID=
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112116550A (en) * | 2019-06-20 | 2020-12-22 | 上海电机学院 | Insulator defect detection method based on multi-corner registration |
| CN115984271A (en) * | 2023-03-20 | 2023-04-18 | 山东鑫科来信息技术有限公司 | Metal burr identification method based on angular point detection |
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112116550A (en) * | 2019-06-20 | 2020-12-22 | 上海电机学院 | Insulator defect detection method based on multi-corner registration |
| CN115984271A (en) * | 2023-03-20 | 2023-04-18 | 山东鑫科来信息技术有限公司 | Metal burr identification method based on angular point detection |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114937055B (en) | Image self-adaptive segmentation method and system based on artificial intelligence | |
| CN111383209B (en) | Unsupervised flaw detection method based on full convolution self-encoder network | |
| CN115082418B (en) | Precise identification method for automobile parts | |
| CN109682839B (en) | Online detection method for surface defects of metal arc-shaped workpiece | |
| CN106780486B (en) | Steel plate surface defect image extraction method | |
| CN115345885A (en) | Method for detecting appearance quality of metal fitness equipment | |
| CN111815555A (en) | Metal additive manufacturing image detection method and device combining anti-neural network with local binary | |
| CN113177924A (en) | Industrial production line product flaw detection method | |
| CN114820625B (en) | Automobile top block defect detection method | |
| CN111539927B (en) | Detection method of automobile plastic assembly fastening buckle missing detection device | |
| CN116152242B (en) | Visual detection system of natural leather defect for basketball | |
| CN115100191A (en) | Metal casting defect identification method based on industrial detection | |
| CN118096739B (en) | Metal forging surface defect detection method based on image processing | |
| CN116883408B (en) | Integrating instrument shell defect detection method based on artificial intelligence | |
| CN114972575A (en) | Linear fitting algorithm based on contour edge | |
| CN112669295A (en) | Lithium battery pole piece defect detection method based on secondary threshold segmentation theory | |
| CN115620061A (en) | Hardware part defect detection method and system based on image recognition technology | |
| CN111724375B (en) | Screen detection method and system | |
| CN115797314A (en) | Part surface defect detection method, system, equipment and storage medium | |
| Li et al. | Surface Defect Detection of Seals Based on K‐Means Clustering Algorithm and Particle Swarm Optimization | |
| CN118470015B (en) | Visual detection method and system for production quality of titanium alloy rod | |
| CN117237747B (en) | Hardware defect classification and identification method based on artificial intelligence | |
| CN118470015A (en) | Visual detection method and system for production quality of titanium alloy rod | |
| CN118212179A (en) | Angle detection method for complex background | |
| CN114332069B (en) | Connector detection method and device based on machine vision |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant |