CN114354631A - Valve blank surface defect detection method based on vision - Google Patents
Valve blank surface defect detection method based on vision Download PDFInfo
- Publication number
- CN114354631A CN114354631A CN202210030525.4A CN202210030525A CN114354631A CN 114354631 A CN114354631 A CN 114354631A CN 202210030525 A CN202210030525 A CN 202210030525A CN 114354631 A CN114354631 A CN 114354631A
- Authority
- CN
- China
- Prior art keywords
- valve blank
- image
- valve
- vision
- blank
- 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.)
- Pending
Links
- 230000007547 defect Effects 0.000 title claims abstract description 54
- 238000001514 detection method Methods 0.000 title claims abstract description 35
- 238000012545 processing Methods 0.000 claims abstract description 35
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 22
- 238000000034 method Methods 0.000 claims abstract description 21
- 238000007781 pre-processing Methods 0.000 claims abstract description 11
- 239000013074 reference sample Substances 0.000 claims abstract description 6
- 238000003708 edge detection Methods 0.000 claims description 7
- 230000011218 segmentation Effects 0.000 claims description 7
- 238000000605 extraction Methods 0.000 claims description 6
- 239000004519 grease Substances 0.000 claims description 6
- 239000012535 impurity Substances 0.000 claims description 6
- 230000008569 process Effects 0.000 abstract description 12
- 238000004519 manufacturing process Methods 0.000 abstract description 5
- 239000002699 waste material Substances 0.000 abstract description 5
- 230000000007 visual effect Effects 0.000 description 7
- 239000000463 material Substances 0.000 description 6
- 238000011179 visual inspection Methods 0.000 description 5
- 239000002184 metal Substances 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 3
- 230000001629 suppression Effects 0.000 description 3
- 229910001369 Brass Inorganic materials 0.000 description 2
- 239000002253 acid Substances 0.000 description 2
- 239000010951 brass Substances 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 238000005266 casting Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 238000004140 cleaning Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- 238000007667 floating Methods 0.000 description 2
- 238000009499 grossing Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 239000002245 particle Substances 0.000 description 2
- 238000007517 polishing process Methods 0.000 description 2
- 238000004080 punching Methods 0.000 description 2
- 239000004576 sand Substances 0.000 description 2
- 229910001220 stainless steel Inorganic materials 0.000 description 2
- 239000010935 stainless steel Substances 0.000 description 2
- 238000004381 surface treatment Methods 0.000 description 2
- 238000005406 washing Methods 0.000 description 2
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 2
- 229910000831 Steel Inorganic materials 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 238000005422 blasting Methods 0.000 description 1
- 238000007635 classification algorithm Methods 0.000 description 1
- 230000002950 deficient Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000005242 forging Methods 0.000 description 1
- 238000010191 image analysis Methods 0.000 description 1
- 230000005764 inhibitory process Effects 0.000 description 1
- 230000001050 lubricating effect Effects 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 230000000877 morphologic effect Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000005488 sandblasting Methods 0.000 description 1
- 238000001228 spectrum Methods 0.000 description 1
- 239000010959 steel Substances 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 238000003786 synthesis reaction Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 238000013519 translation Methods 0.000 description 1
- 238000004506 ultrasonic cleaning Methods 0.000 description 1
Images
Classifications
-
- 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
- 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
Abstract
A valve blank surface defect detection method based on vision belongs to the technical field of vision detection. The invention comprises the following steps: step S1, processing the surface of the valve blank and preprocessing the three-dimensional reference sample picture of the valve blank; step S2, acquiring high dynamic graphic information of the surface of the valve blank; step S3, processing the high dynamic graph by using an image processing algorithm, and extracting the characteristic information of the valve blank image; step S4, comparing and identifying the characteristic information of the valve blank image with the preprocessed reference image one by one, and detecting the surface defect of the valve blank; and step S5, sorting the valve blanks with the unqualified products and the surface defects. The invention can identify the problems of the body and the surface defects of the valve blank, can sort out most of the defects of the valve blank, reduces unqualified products from flowing into the next process, reduces the waste of the production process, improves the detection efficiency of the valve blank and reduces the detection cost.
Description
Technical Field
The invention belongs to the technical field of visual detection, and particularly relates to a valve blank surface defect detection method based on vision.
Background
The visual inspection means that a detected product target is converted into image information through machine vision, the image information is transmitted to a set image processing system for analysis and processing, and the image information is converted into digital information according to various information such as the size, brightness, color difference, outline and the like of the detected target image; the image system performs specific operation on the input information, analyzes the target characteristics, and then outputs a signal or equipment action according to a judgment result.
Machine visual inspection compares artifical visual inspection, has efficiently, and the recognition accuracy is high, and it is good to last work endurance, and can use in dangerous work occasion to and the unable occasion that detects the target feature of artifical visual. With the development of camera processing systems and computer algorithm technologies, visual inspection sets different visual inspection methods for different target characteristics, and is widely applied in various industries.
In the valve industry, a valve blank is an early-stage material for manufacturing a valve component, the valve blank is usually formed by red punching or casting, the problems of cold shut, air holes, cracks, flash, material shortage and the like exist, the defects cannot be eliminated in the blank stage, and the waste of a later-stage processing process and the inflow of unqualified products can be caused. The manual detection wastes time and labor, has low efficiency, and is easy to generate detection fatigue, so that defective products flow into the next link.
The valve blank is of a three-dimensional structure, and on an assembly line detection platform, the valve blank provides new challenges for the conventional visual detection mode of a planar product in different forms, angles and positions.
Disclosure of Invention
The invention mainly solves the technical problems in the prior art and provides a valve blank surface defect detection method based on vision.
The technical problem of the invention is mainly solved by the following technical scheme: a valve blank surface defect detection method based on vision comprises the following steps:
step S1, processing the surface of the valve blank and preprocessing the three-dimensional reference sample picture of the valve blank;
step S2, acquiring high dynamic graphic information of the surface of the valve blank;
step S3, processing the high dynamic graph by using an image processing algorithm, and extracting the characteristic information of the valve blank image;
step S4, comparing and identifying the characteristic information of the valve blank image with the preprocessed reference image one by one, and detecting the surface defect of the valve blank;
and step S5, sorting the valve blanks with the unqualified products and the surface defects.
Preferably, step S1 specifically includes the following steps:
step a, eliminating surface color difference, grease and impurities of a valve blank to form uniform surface metallic color;
and b, continuously recording frames of the three-dimensional shape of the valve blank at different angles, and preprocessing the frame records to form a reference sample picture.
Preferably, the three-dimensional shape of the valve blank is recorded in successive frames along the X-axis, Y-axis and Z-axis, respectively.
Preferably, the preprocessing of the frame recording is to perform image binarization processing on the continuous frame recording pictures one by one, and set the pixel gray scale under the threshold value to a uniform specific value.
Preferably, the valve blank is placed on a monochromatic conveyor belt of the production line, and passes through a light source and an industrial CCD or CMOS camera, so that high dynamic graphic information of the surface of the valve blank is obtained.
Preferably, the valve blank image is subjected to image binarization processing, a dynamic threshold segmentation 0tsu algorithm is introduced, the valve blank image is divided into a target area and a background area, and the maximum inter-class variance of the target area and the background area is calculated.
Preferably, after the valve blank image is processed, the edge contour of the valve blank is extracted, and the edge contour and the contour of the reference sample picture are compared and identified one by one.
Preferably, the edge profile of the valve blank is extracted by a Canny edge detection algorithm.
Preferably, after the valve blank image is processed, the valve blank surface defect feature information is subjected to matching extraction and defect classification processing by adopting a fast Fourier transform method and an R-FCN algorithm.
Preferably, the method further comprises extracting defect gray scale features or/and defect texture feature information.
The invention has the following beneficial effects: the method can identify the body problems of overlarge flash, deformation, material shortage and the like of the valve blank and the surface defects of cold shut, cracks and the like, can sort out most of the defects of the valve blank, reduces unqualified products from flowing into the next process link, reduces the waste of the production process, improves the detection efficiency of the valve blank, reduces the detection cost, has obvious benefits when being applied to metal processing industries such as valve pipelines and the like, and has market popularization value.
Drawings
FIG. 1 is a view of a valve blank of the present invention in a transport position on an assembly line;
FIG. 2 is a template diagram for gradient computation according to the present invention.
In the figure: 1. valve blanks; 2. a single color conveyor belt; 3. a light source; 4. a CMOS camera; 5. a valve blank image; 6. and (5) taking reference pictures.
Detailed Description
The technical scheme of the invention is further specifically described by the following embodiments and the accompanying drawings.
Example (b): a valve blank surface defect detection method based on vision comprises the following steps:
step S1, processing the surface of the valve blank 1 and preprocessing the three-dimensional reference picture 6 of the valve blank 1;
step S2, acquiring high dynamic graphic information of the surface of the valve blank 1;
step S3, processing the high dynamic graph by using an image processing algorithm, and extracting the characteristic information of the valve blank image 5;
step S4, comparing and identifying the characteristic information of the valve blank image 5 with the preprocessed reference image 6 one by one, and detecting the surface defects of the valve blank 1;
and step S5, sorting the valve blanks 1 with the unqualified products and the surface defects.
Wherein, step S1 specifically includes the following steps:
step a, eliminating surface color difference, grease and impurities of a valve blank 1 to form uniform surface metallic color;
and step b, continuously recording frames of the three-dimensional form of the valve blank 1 at different angles, and preprocessing the frame records to form a reference picture 6.
After the valve blank 1 is formed by red punching, casting or forging, the surface of the valve blank 1 is more or less provided with lubricating grease, shallow black skin, floating particles and the like, and the judgment of vision on the valve blank 1 can be seriously influenced by the formed surface impurity color or blackening. In order to improve the detection recognition rate, the valve blank 1 is firstly subjected to sand blasting or acid washing or water blasting treatment to eliminate surface color difference, grease and impurities and form uniform surface metal color, such as brass natural color or stainless steel natural color. The sand throwing process is to use a tiny steel ball to impact the surface of the valve blank 1 to remove floating impurities, grease and black skin on the surface. The water polishing process is similar to the sand polishing process, and is a physical surface treatment process, and the acid cleaning process is a chemical surface treatment process.
The sand-throwing, acid-washing and water-throwing treatment can form concave-convex dermatoglyph on the surface of the valve blank 1, but has little influence on visual judgment, thus being a more ideal treatment process. The surface of the valve blank 1 can be well protected by ultrasonic cleaning, but the cleaning effects of flow marks, attached particles and the like are general, and more visual misjudgments exist in the later period.
Different from plane or specification visual angle structure, valve blank 1 is three-dimensional, presents different three-dimensional forms and puts the angle on the assembly line, and valve blank 1 volume is big simultaneously, can't regularly put one by one, and along with assembly line moving vibrations, puts the change of regular valve blank 1 also can produce three-dimensional form and put the angle, has increaseed the degree of difficulty of visual judgement.
Therefore, the invention innovatively adopts a reference picture library, continuous frame recording is carried out on the three-dimensional form of the valve blank 1 at different angles in advance, and the frame recording is preprocessed to form a reference picture 6. The three-dimensional shape of the valve blank 1 is drawn according to the valve blank 1 with standard size by three-dimensional software.
And the three-dimensional form of the valve blank 1 is subjected to continuous frame recording along an X axis, a Y axis and a Z axis respectively to form a three-dimensional form image library of the valve blank. For example, the three-dimensional shape of the valve blank 1 is firstly exported by standard pictures after rotating for 0.5 degree or 1 degree along the X axis, and the standard pictures are rotated for 360 degrees; then, after rotating every 0.5 degree or 1 degree along the Y axis, exporting the standard photo, and rotating 360 degrees; then, after rotating every 0.5 degree or 1 degree along the Z axis, exporting the standard photo, and rotating 360 degrees; the control of the rotational angle accuracy is performed based on the accuracy and the degree of recognition. During the recording of successive frames, the corresponding standard photographs can also be derived in terms of axis or specific angle.
Meanwhile, the output frame recording standard photo is preprocessed to form a reference photo 6. The preprocessing mainly relates to filtering and denoising of the photos, image binarization processing, image morphological analysis processing and the like, the noise points of the photos of the standard photos are generally less, background pictures can be removed, and the filtering and denoising process can be reduced according to the situation. The preprocessing of the frame records is to perform image binarization processing on continuous frame record pictures one by one and set the pixel gray level under a threshold value as a uniform specific value.
The image binarization is to set the pixel gray level under a certain threshold value as a uniform specific value, so as to provide a basis for subsequent edge detection and image analysis processing, and reduce the complexity of subsequent algorithm processing of the image.
In step S2, the valve blank 1 is placed on a monochrome conveyor 2 of an assembly line, as shown in fig. 1, through a selected light source 3 and an industrial CCD or CMOS camera 4, thereby acquiring high dynamic pattern information of the surface of the valve blank 1. The monochromatic conveyor belt 2 can be selected from green, blue, red and the like, particularly green, and compared with the brass natural color or stainless steel color of a metal valve, the color difference is large, the color is single, and the image processing noise is low.
In step S3, during the image binarization processing, the Otsu algorithm of dynamic threshold segmentation is introduced into the valve blank image 5 to divide the valve blank image 5 into two areas, namely, the target area and the background area. The Otsu algorithm, namely the maximum between-class variance algorithm, divides the valve blank image 5 into two areas of a target and a background, calculates the maximum between-class variance of the target and the background area by adopting different binarization threshold values, wherein the larger the variance is, the larger the difference between the target and the background is, the more ideal the segmentation effect is, and the lower the error probability is. And when the traversal is completed, the obtained maximum value of the variance between the pixel classes is used as the optimal value of the image global threshold.
In the aspect of contour extraction, after the valve blank image 5 is subjected to image processing, a target object area in the valve blank image 5 needs to be extracted, the edge of the target object area is extracted by adopting edge detection according to the forming process and the surface defect condition of the valve blank 1, the external graph with the largest target is obtained by utilizing a contour representation algorithm, and then the target object area is intercepted according to the required size. In order to balance edge accurate positioning and noise suppression, a self-adaptive Canny edge detection algorithm is adopted, the suppression of multi-response edges can be enhanced, the missing rate of the edges is reduced, and therefore the edge positioning accuracy is improved.
Assuming that an original image is I (x, y) and a finite difference gradient is S (x, y), performing Gaussian smoothing filter convolution operation on the image by using a Sobel operator in a Canny edge detection algorithm:
S(x，y)＝G(x，y，σ)*I(x，y)；
wherein, σ is a smoothing degree parameter of the gaussian filter, when σ takes a larger value, noise can be suppressed, but edge positioning accuracy is lower, and when σ takes a smaller value, edge positioning accuracy is higher, but noise suppression capability is reduced. Therefore, the value of the noise is required to be set according to the noise condition of the image.
Meanwhile, by using 4 gradient calculation templates, as shown in fig. 2, the total gradient of the gray value of the current pixel point in two directions, i.e. horizontal and vertical, of 4 gradient synthesis is calculated as follows:
therefore, the gradient amplitude S (x, y) and the gradient angle theta (x, y) of the pixel point of the current gray value can be calculated;
the phenomenon of edge break or false edge may occur during the calculation process, and a high threshold value T is also needed to be setHAnd a low threshold TLWhen the gradient value of the detected pixel point is greater than TH, the pixel point is at the edge, and when the gradient value of the pixel point is less than TLThen the pixel is non-edge. Meanwhile, the gray value of the non-edge pixel point is set to be 0 by utilizing non-maximum value inhibition, the point with the maximum amplitude value is reserved as the pixel point with edge refinement, and accurate positioning and extraction of the edge contour are achieved.
After the valve blank image 5 is processed, extracting the edge contour of the valve blank 1, comparing and identifying the edge contour with the contour of the reference picture 6 one by one, and mainly identifying the body abnormity such as overlarge flash, deformation, material shortage and the like of the valve blank 1.
Meanwhile, after the valve blank image 5 is processed, the valve blank 1 surface defect feature information is subjected to matching extraction and defect classification processing by adopting a fast Fourier transform method and an R-FCN algorithm.
After image threshold segmentation and edge contour extraction, the valve blank image 5 is subjected to threshold segmentation into different areas, a certain amount of false targets and noise are contained, and the correspondingly obtained image needs to be subjected to secondary processing so as to extract a real defect characteristic area. Because the valve blank 1 may have various phenomena such as cold shut, cracks, air holes and the like, and the geometric characteristics of the surface defects need to be determined at first, the target image and the template are subjected to image matching by adopting fast Fourier transform, the target geometric airspace image is transformed into a frequency domain character area in the template, and the geometric airspace is inversely transformed after the matching is finished. Assuming that the valve blank image 5f1(x, y) is moved in the coordinate system to obtain an image f1(x, y), the amount of translation is (x, y)0，y0) If the rotation angle is α, the corresponding fourier transform domain relationship and cross power spectra of f1 and f1 are respectively: f2(u，v)＝|F1((ucosα+vsinα)，(-usinα+vcosα))|；
Then, inverse transformation is performed on the obtained matching image, and the obtained result is compared with the valve blank image 5 to extract the geometric feature information of the defect.
Meanwhile, in order to extract the comprehensive information of the surface defect characteristic, the gray characteristic or/and the texture characteristic information of the defect can be extracted. The defect gray level features can be extracted through the features of the gray level mean value, the gray level variance and the gray level entropy in the image gray level histogram; the defect culture characteristic can be extracted through the characteristics of contrast, correlation and entropy in the image gray level co-occurrence matrix.
In addition, a proper defect classification algorithm is selected for processing according to the characteristics of the defects and the expected sorting target. The R-FCN is used as a full convolution network structure and has the characteristics of high target identification speed, high accuracy and more precise target classification.
According to the invention, the characteristic information of the reference picture 6 and the valve blank picture 5 is subjected to overlapping analysis and comparison, so that the body abnormity such as overlarge flash, deformation and material shortage of the valve blank 1 can be better identified; the valve blank 1 is placed on a monochromatic conveyor belt 2 of an assembly line, the noise point of an image can be effectively reduced, the contrast is obvious, and a dynamic threshold segmentation 0tsu algorithm is introduced, so that the target and the background of the valve blank 1 can be effectively segmented; after the Canny edge detection algorithm is adopted to operate the image, the algorithm can well detect the edge and has robust performance. The field detection result of a manufacturer shows that the defect detection is carried out on 3000 valve blank 1 samples, the defect detection rate of the detection method is higher than that of a conventional detection method, the defect detection rate reaches 85%, and the detection efficiency is improved by more than one time.
In conclusion, the valve blank detection device can identify the body problems of overlarge flash, deformation, material shortage and the like of the valve blank and the surface defects of cold shut, cracks and the like, can sort out most of the defects of the valve blank, reduces unqualified products from flowing into the next process, reduces the waste of the production process, improves the detection efficiency of the valve blank, reduces the detection cost, has obvious benefits when being applied to metal processing industries such as valve pipelines and the like, and has market popularization value.
Finally, it should be noted that the above embodiments are merely representative examples of the present invention. It is obvious that the invention is not limited to the above-described embodiments, but that many variations are possible. Any simple modification, equivalent change and modification made to the above embodiments in accordance with the technical spirit of the present invention should be considered to be within the scope of the present invention.
Claims (10)
1. A valve blank surface defect detection method based on vision is characterized by comprising the following steps:
step S1, processing the surface of the valve blank and preprocessing the three-dimensional reference sample picture of the valve blank;
step S2, acquiring high dynamic graphic information of the surface of the valve blank;
step S3, processing the high dynamic graph by using an image processing algorithm, and extracting the characteristic information of the valve blank image;
step S4, comparing and identifying the characteristic information of the valve blank image with the preprocessed reference image one by one, and detecting the surface defect of the valve blank;
and step S5, sorting the valve blanks with the unqualified products and the surface defects.
2. The vision-based valve blank surface defect detecting method of claim 1, wherein the step S1 specifically comprises the following steps:
step a, eliminating surface color difference, grease and impurities of a valve blank to form uniform surface metallic color;
and b, continuously recording frames of the three-dimensional shape of the valve blank at different angles, and preprocessing the frame records to form a reference sample picture.
3. The vision-based method for detecting surface defects of a valve blank according to claim 2, wherein the three-dimensional shape of the valve blank is recorded in successive frames along the X-axis, the Y-axis and the Z-axis.
4. The vision-based valve blank surface defect detection method as claimed in claim 2, wherein the frame recording preprocessing is to perform image binarization processing on successive frame recording pictures one by one, and to set the pixel gray level under the threshold value to a uniform specific value.
5. The vision-based valve blank surface defect detection method of claim 1, wherein the valve blank is placed on a monochrome conveyor belt of an assembly line, and passes through a light source and an industrial CCD or CMOS camera, thereby obtaining high dynamic pattern information of the valve blank surface.
6. The vision-based valve blank surface defect detection method as claimed in claim 1, wherein the valve blank image is subjected to image binarization processing, and an Otsu algorithm of dynamic threshold segmentation is introduced, so that the valve blank image is divided into two areas of a target area and a background area, and the maximum inter-class variance of the target area and the background area is calculated.
7. The vision-based valve blank surface defect detection method of claim 1, wherein the valve blank image is processed, and the edge contour of the valve blank is extracted and compared with the contour of the reference picture one by one for identification.
8. The vision-based valve blank surface defect detection method of claim 7, wherein the edge profile of the valve blank is extracted by a Canny edge detection algorithm.
9. The vision-based valve blank surface defect detection method of claim 1, wherein after the valve blank image is processed, the valve blank surface defect feature information is subjected to matching extraction and defect classification processing by using a fast fourier transform method and an R-FCN algorithm.
10. The vision-based method for detecting surface defects of a valve blank according to claim 9, further comprising extracting information of gray scale features and/or texture features of the defects.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210030525.4A CN114354631A (en) | 2022-01-12 | 2022-01-12 | Valve blank surface defect detection method based on vision |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210030525.4A CN114354631A (en) | 2022-01-12 | 2022-01-12 | Valve blank surface defect detection method based on vision |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114354631A true CN114354631A (en) | 2022-04-15 |
Family
ID=81109128
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210030525.4A Pending CN114354631A (en) | 2022-01-12 | 2022-01-12 | Valve blank surface defect detection method based on vision |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114354631A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114612471A (en) * | 2022-05-10 | 2022-06-10 | 佛山市阿瑞斯数字设备有限公司 | Method for detecting surface defects of ceramic tiles with complex textures |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070280501A1 (en) * | 2006-05-31 | 2007-12-06 | The Boeing Company | Method and System for Two-Dimensional and Three-Dimensional Inspection of a Workpiece |
| CN103229045A (en) * | 2010-10-08 | 2013-07-31 | 比利时胶囊公司 | Apparatus and method for acquiring a two-dimensional image of the surface of a three-dimensional object |
| JP2016090401A (en) * | 2014-11-05 | 2016-05-23 | 古河電気工業株式会社 | Appearance inspection device, appearance inspection system and appearance inspection method |
| CN108184108A (en) * | 2018-01-12 | 2018-06-19 | 盎锐（上海）信息科技有限公司 | Image generating method and device based on 3D imagings |
| US20180350057A1 (en) * | 2017-06-06 | 2018-12-06 | Nuflare Technology, Inc. | Inspection method and inspection system |
| CN109523541A (en) * | 2018-11-23 | 2019-03-26 | 五邑大学 | A kind of metal surface fine defects detection method of view-based access control model |
| CN109737877A (en) * | 2019-03-15 | 2019-05-10 | 定襄县格美莱智能科技有限公司 | Part blank detection system and method |
| CN110108712A (en) * | 2019-04-20 | 2019-08-09 | 东莞中科蓝海智能视觉科技有限公司 | Multifunctional visual sense defect detecting system |
| KR102177726B1 (en) * | 2019-05-15 | 2020-11-12 | 한국기계연구원 | Method and apparatus for inspecting workpiece |
| WO2020248439A1 (en) * | 2019-06-11 | 2020-12-17 | 江苏农林职业技术学院 | Crown cap surface defect online inspection method employing image processing |
| CN112304958A (en) * | 2020-11-25 | 2021-02-02 | 四川晶剑电子材料有限公司 | Copper strip surface defect detection method and system based on machine vision |
-
2022
- 2022-01-12 CN CN202210030525.4A patent/CN114354631A/en active Pending
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070280501A1 (en) * | 2006-05-31 | 2007-12-06 | The Boeing Company | Method and System for Two-Dimensional and Three-Dimensional Inspection of a Workpiece |
| CN103229045A (en) * | 2010-10-08 | 2013-07-31 | 比利时胶囊公司 | Apparatus and method for acquiring a two-dimensional image of the surface of a three-dimensional object |
| JP2016090401A (en) * | 2014-11-05 | 2016-05-23 | 古河電気工業株式会社 | Appearance inspection device, appearance inspection system and appearance inspection method |
| US20180350057A1 (en) * | 2017-06-06 | 2018-12-06 | Nuflare Technology, Inc. | Inspection method and inspection system |
| CN108184108A (en) * | 2018-01-12 | 2018-06-19 | 盎锐（上海）信息科技有限公司 | Image generating method and device based on 3D imagings |
| CN109523541A (en) * | 2018-11-23 | 2019-03-26 | 五邑大学 | A kind of metal surface fine defects detection method of view-based access control model |
| CN109737877A (en) * | 2019-03-15 | 2019-05-10 | 定襄县格美莱智能科技有限公司 | Part blank detection system and method |
| CN110108712A (en) * | 2019-04-20 | 2019-08-09 | 东莞中科蓝海智能视觉科技有限公司 | Multifunctional visual sense defect detecting system |
| KR102177726B1 (en) * | 2019-05-15 | 2020-11-12 | 한국기계연구원 | Method and apparatus for inspecting workpiece |
| WO2020248439A1 (en) * | 2019-06-11 | 2020-12-17 | 江苏农林职业技术学院 | Crown cap surface defect online inspection method employing image processing |
| CN112304958A (en) * | 2020-11-25 | 2021-02-02 | 四川晶剑电子材料有限公司 | Copper strip surface defect detection method and system based on machine vision |
Non-Patent Citations (2)
| Title |
|---|
| 徐伟峰等: "基于机器视觉的接头组件表面缺陷检测系统研究", 《机床与液压》, vol. 48, no. 16, 31 August 2020 (2020-08-31), pages 72 - 77 * |
| 段春梅等: "基于机器视觉的瓷砖素坯表面缺陷无损检测算法研究", 智能计算机与应用, vol. 07, no. 03, 30 June 2017 (2017-06-30), pages 37 - 40 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114612471A (en) * | 2022-05-10 | 2022-06-10 | 佛山市阿瑞斯数字设备有限公司 | Method for detecting surface defects of ceramic tiles with complex textures |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109816652B (en) | Complex casting defect identification method based on gray level significance | |
| CN109872300B (en) | Visual saliency detection method for appearance defects of friction plate | |
| CN106404793B (en) | Bearing sealing element defect detection method based on vision | |
| CN107490582B (en) | Assembly line workpiece detection system | |
| CN106780526A (en) | A kind of ferrite wafer alligatoring recognition methods | |
| WO2021109697A1 (en) | Character segmentation method and apparatus, and computer-readable storage medium | |
| CN109658391B (en) | Circle radius measuring method based on contour merging and convex hull fitting | |
| CN109584215A (en) | A kind of online vision detection system of circuit board | |
| CN113077437B (en) | Workpiece quality detection method and system | |
| CN113298776B (en) | Method for detecting appearance defects of metal closed water pump impeller | |
| CN106780437B (en) | A kind of quick QFN chip plastic packaging image obtains and amplification method | |
| CN115100191A (en) | Metal casting defect identification method based on industrial detection | |
| CN111402238A (en) | Defect identification system realized through machine vision | |
| CN108020554A (en) | A kind of steel strip surface defect recognition detection method | |
| CN115684176A (en) | Online visual inspection system for film surface defects | |
| CN111667475A (en) | Machine vision-based Chinese date grading detection method | |
| CN113177924A (en) | Industrial production line product flaw detection method | |
| CN115311289A (en) | Method for detecting oil stain defects of plain-color cloth | |
| CN112669265B (en) | Method for realizing surface defect detection based on Fourier transform and image gradient characteristics | |
| CN114354631A (en) | Valve blank surface defect detection method based on vision | |
| Tang et al. | Surface inspection system of steel strip based on machine vision | |
| CN112381140B (en) | Abrasive particle image machine learning identification method based on new characteristic parameters | |
| CN111815575B (en) | Bearing steel ball part detection method based on machine vision | |
| CN108269264B (en) | Denoising and fractal method of bean kernel image | |
| CN113516619A (en) | Product surface flaw identification method based on image processing technology |
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 |