CN110490864B - Self-adaptive defect detection method for image - Google Patents
Self-adaptive defect detection method for image Download PDFInfo
- Publication number
- CN110490864B CN110490864B CN201910779680.4A CN201910779680A CN110490864B CN 110490864 B CN110490864 B CN 110490864B CN 201910779680 A CN201910779680 A CN 201910779680A CN 110490864 B CN110490864 B CN 110490864B
- Authority
- CN
- China
- Prior art keywords
- image
- mask
- value
- detected
- pixel
- 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
- 238000001514 detection method Methods 0.000 title claims abstract description 27
- 230000007547 defect Effects 0.000 title claims abstract description 26
- 238000000034 method Methods 0.000 claims abstract description 8
- 230000003044 adaptive effect Effects 0.000 claims abstract description 7
- 230000002159 abnormal effect Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000005856 abnormality Effects 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 239000003973 paint Substances 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10004—Still image; Photographic image
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Quality & Reliability (AREA)
- Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
- Preparing Plates And Mask In Photomechanical Process (AREA)
- Image Analysis (AREA)
Abstract
The invention disclosesAn adaptive defect detection method for images comprises the following steps: 1) generating mask images with the same size for the images to be detected, counting the mean value mu of the images to be detected, wherein the initial mask value is 00And standard deviation σ0Determining a coefficient M; 2) judging whether the gray value of each pixel is larger than mu0+M×σ0Or less than mu0‑M×σ0If yes, setting the mask value to be 1; 3) counting the mean value mu of the part with the mask value of 0 in the image to be measurediAnd standard deviation σiSetting the gray value of the pixel corresponding to the mask value 1 on the image to be measured as the mean value muiObtaining an image I; 4) generating a mask image I by the same method and judging; 5) i +1, repeating the steps 3) and 4) until the mask values are all 0; 6) and merging and displaying all the areas with the mask value of 1 to obtain a final defect mask image. The method avoids false detection or missing detection caused by unreasonable threshold setting, and can effectively realize defect detection and identification of each scale.
Description
Technical Field
The invention relates to the field of image detection, in particular to a self-adaptive defect detection method for an image.
Background
The image detection technology is a visual detection method widely applied at present, and for a picture with a single ground color, a method of setting a threshold value and identifying an abnormal value in the picture is generally adopted to detect the abnormal point on the surface of the picture, and the abnormal point is a defect on a product. However, setting a single threshold often fails to accurately detect all outliers in an image. This is because the small-sized abnormality is missed if the threshold value is set too large, and a large number of invalid regions are erroneously detected if the threshold value is set too small.
Disclosure of Invention
In order to solve the technical problems, the invention provides an image self-adaptive defect detection method, which can effectively solve the problem that a threshold is difficult to determine during defect identification, avoid false detection or missing detection caused by unreasonable threshold setting, and effectively realize defect detection and identification of each scale.
Therefore, the technical scheme of the invention is as follows:
an adaptive defect detection method for an image comprises the following steps:
1) generating mask graphs with the same size for an image to be detected, setting the initial mask value of each pixel in the mask graphs to be 0, and counting the mean value mu of the gray value of each pixel of the image to be detected0And standard deviation σ0Determining a coefficient M;
the image to be detected is an image with single and uniform ground color;
2) judging whether the gray value of each pixel in the image to be detected is large or notAt mu0+M×σ0Or less than mu0-M×σ0If yes, setting the mask value of the corresponding position of the mask map as 1;
3) counting the gray value mean value mu of the corresponding pixel of the part with the mask value of 0 in the mask image in the image to be detectediAnd standard deviation σiSetting the gray value of the pixel of the part with the mask value of 1 in the mask image on the image to be measured as the mean value muiObtaining an image I; i is a positive integer; i ═ I;
4) generating a mask I with the same size as the image I, setting the initial mask value of the mask I to be 0, and judging whether the gray value of each pixel in the image I is larger than mu or noti+M×σiOr less than mui-M×σiIf yes, setting the mask value of the position corresponding to the mask image I as 1;
5) i +1, replacing the mean value mu of the gray values calculated in step 3) with the mask image I and the image I in step 4)iAnd standard deviation σiRepeating the steps 3) and 4) on the used mask image and the image to be detected until all mask values in the mask image are 0;
6) and combining and displaying the areas with the mask value of 1 in multiple cycles to obtain a final defect mask image. The mask image can be used for acquiring the defect area on the image to be detected.
Further, the cycle termination condition of the step 5) is the preset maximum cycle number or the same mask image obtained after two adjacent cycles.
Further, M is 1 to 5.
Further, the image to be detected is a difference or convolution graph of phase, gradient and curvature, and a modulation graph.
According to the image self-adaptive defect detection method, the detection threshold is set repeatedly, the problem that the threshold is difficult to determine during defect identification can be effectively solved, false detection or missing detection caused by unreasonable threshold setting is avoided, and defect detection and identification of all scales can be effectively realized.
Drawings
FIG. 1 is a diagram illustrating the detection effect of a fixed threshold value that is larger according to a conventional method;
FIG. 2 is a diagram illustrating the detection effect when the threshold is fixed by the prior art method;
FIG. 3 is a mask obtained in step 2) of the present embodiment;
FIG. 4 is a mask map obtained after cycle 1 in an embodiment;
FIG. 5 is a mask map obtained after the 2 nd cycle in an embodiment;
FIG. 6 is a mask map obtained after the 3 rd cycle in the preferred embodiment;
FIG. 7 is a final defect mask map that results from the method of the embodiment.
Note: the same sheet of artwork is processed in figures 1, 2 and 3-7.
Detailed Description
The technical solution of the present invention is described in detail below with reference to the accompanying drawings and the detailed description.
An adaptive defect detection method for an image comprises the following steps:
1) generating mask images with the same size for the image to be detected, setting the initial mask value of each pixel in the mask images to be 0, and counting the mean value mu of the gray value of each pixel of the image to be detected0And standard deviation σ0Determining a coefficient M;
the image to be detected is an image with single and uniform ground color; mirror surface/mirror-like surface images such as automobile paint surfaces, mobile phone screens and the like;
the image to be detected is a difference or convolution graph of phase, gradient and curvature and a modulation graph, and an original image used in the attached figures 3-7 is a phase difference graph after defect modulation;
2) judging whether the gray value of each pixel in the image to be detected is greater than mu0+M×σ0Or less than mu0-M×σ0If yes, setting the mask value of the corresponding position of the mask map as 1; preferably, M is 1 to 5, where M is 3, to obtain a first mask map, fig. 3;
3) counting the gray value mean value mu of the corresponding pixel of the part with the mask value of 0 in the mask image in the image to be detectediAnd standard deviation σiThen, the image to be measured is placed onThe gray value of the pixel of the part with the mask value of 1 in the mask image is set as the mean value muiObtaining an image I; i is a positive integer; i ═ I;
4) generating a mask I with the same size as the image I, setting the initial mask value of the mask I to be 0, and judging whether the gray value of each pixel in the image I is larger than mui+M×σiOr less than mui-M×σiIf yes, setting the mask value of the position corresponding to the mask image I as 1;
5) i +1, replacing the mean value mu of the gray values calculated in step 3) with the mask image I and the image I in step 4)iAnd standard deviation σiRepeating the steps 3) and 4) on the used mask image and the image to be detected until all mask values in the mask image are 0; when i is 1, the mask diagram 1 is fig. 4; when i is 2, the mask fig. 1 is fig. 5; when i is 3, the mask fig. 1 is fig. 6;
6) and combining and displaying the areas with the mask value of 1 in multiple cycles to obtain a final defect mask map, as shown in FIG. 7. The mask image can be used for acquiring the defect area on the image to be detected. It can be seen that fig. 7 does not have the clutter in the lower right corner of fig. 2, relative to the existing fixed threshold approach employed in fig. 2.
In order to prevent the loop from being unable to be terminated due to program error, the loop termination condition of step 5) is also the same as the preset maximum loop number or the mask obtained after two adjacent loops.
The foregoing descriptions of specific exemplary embodiments of the present invention have been presented for purposes of illustration and description. The foregoing description is not intended to be exhaustive or to limit the invention to the precise form disclosed, and obviously many modifications and variations are possible in light of the above teaching. The exemplary embodiments were chosen and described in order to explain certain principles of the invention and its practical application to enable others skilled in the art to make and use various exemplary embodiments of the invention and various alternatives and modifications thereof. It is intended that the scope of the invention be defined by the following claims and their equivalents.
Claims (4)
1. An adaptive defect detection method for an image is characterized by comprising the following steps:
1) generating mask graphs with the same size for an image to be detected, setting the initial mask value of each pixel in the mask graphs to be 0, and counting the mean value mu of the gray value of each pixel of the image to be detected0And standard deviation σ0Determining a coefficient M;
the image to be detected is an image with single and uniform ground color;
2) judging whether the gray value of each pixel in the image to be detected is greater than mu0+M×σ0Or less than mu0-M×σ0If yes, setting the mask value of the corresponding position of the mask map as 1;
3) counting the gray value mean value mu of the corresponding pixel of the part with the mask value of 0 in the mask image in the image to be detectediAnd standard deviation σiSetting the gray value of the pixel of the part with the mask value of 1 in the mask image on the image to be measured as the mean value muiObtaining an image I; i is a positive integer; i ═ I;
4) generating a mask I with the same size as the image I, setting the initial mask value of the mask I to be 0, and judging whether the gray value of each pixel in the image I is larger than mu or noti+M×σiOr less than mui-M×σiIf yes, setting the mask value of the position corresponding to the mask image I as 1;
5) i +1, replacing the mean value mu of the gray values calculated in step 3) with the mask image I and the image I in step 4)iAnd standard deviation σiRepeating the steps 3) and 4) on the used mask image and the image to be detected until all mask values in the mask image are 0;
6) and combining and displaying the areas with the mask value of 1 in multiple cycles to obtain a final defect mask image.
2. The method for adaptive defect detection of an image according to claim 1, wherein: the cycle termination condition of the step 5) is also the preset maximum cycle number or the same mask image obtained after two adjacent cycles.
3. The method for adaptive defect detection of an image according to claim 1, wherein: m is 1-5.
4. The method for adaptive defect detection of an image according to claim 1, wherein: the image to be measured is a difference or convolution graph of phase, gradient and curvature and a modulation graph.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910779680.4A CN110490864B (en) | 2019-08-22 | 2019-08-22 | Self-adaptive defect detection method for image |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910779680.4A CN110490864B (en) | 2019-08-22 | 2019-08-22 | Self-adaptive defect detection method for image |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110490864A CN110490864A (en) | 2019-11-22 |
CN110490864B true CN110490864B (en) | 2021-09-21 |
Family
ID=68552984
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910779680.4A Active CN110490864B (en) | 2019-08-22 | 2019-08-22 | Self-adaptive defect detection method for image |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110490864B (en) |
Families Citing this family (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111127415B (en) * | 2019-12-19 | 2023-07-25 | 信利(仁寿)高端显示科技有限公司 | Mura quantification method based on excimer laser annealing |
CN112598647A (en) * | 2020-12-24 | 2021-04-02 | 凌云光技术股份有限公司 | Method for detecting weak line defect under arc-shaped surface gradual change background |
CN113284148B (en) * | 2021-07-23 | 2021-10-15 | 苏州高视半导体技术有限公司 | Screen dust filtering method |
CN113888537B (en) * | 2021-12-03 | 2022-04-12 | 深圳市网旭科技有限公司 | Mask extraction method, device, equipment and storage medium |
CN116452582A (en) * | 2023-06-13 | 2023-07-18 | 季华实验室 | Mirror-like surface data detection method and device, electronic equipment and storage medium |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6381374B1 (en) * | 1998-10-30 | 2002-04-30 | General Electric Company | Histogram analysis method for defective pixel identification |
CN107703715A (en) * | 2016-08-08 | 2018-02-16 | 中芯国际集成电路制造(上海)有限公司 | A kind of restorative procedure of mask pattern defect |
CN108460757A (en) * | 2018-02-11 | 2018-08-28 | 深圳市鑫信腾科技有限公司 | A kind of mobile phone TFT-LCD screens Mura defects online automatic detection method |
CN109916922A (en) * | 2019-04-02 | 2019-06-21 | 易思维(杭州)科技有限公司 | Mirror surface/class mirror article defect inspection method |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20190238796A1 (en) * | 2017-05-11 | 2019-08-01 | Jacob Nathaniel Allen | Object Inspection System And Method For Inspecting An Object |
-
2019
- 2019-08-22 CN CN201910779680.4A patent/CN110490864B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6381374B1 (en) * | 1998-10-30 | 2002-04-30 | General Electric Company | Histogram analysis method for defective pixel identification |
CN107703715A (en) * | 2016-08-08 | 2018-02-16 | 中芯国际集成电路制造(上海)有限公司 | A kind of restorative procedure of mask pattern defect |
CN108460757A (en) * | 2018-02-11 | 2018-08-28 | 深圳市鑫信腾科技有限公司 | A kind of mobile phone TFT-LCD screens Mura defects online automatic detection method |
CN109916922A (en) * | 2019-04-02 | 2019-06-21 | 易思维(杭州)科技有限公司 | Mirror surface/class mirror article defect inspection method |
Also Published As
Publication number | Publication date |
---|---|
CN110490864A (en) | 2019-11-22 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110490864B (en) | Self-adaptive defect detection method for image | |
CN111758024B (en) | Defect detection method and device | |
CN108896278B (en) | Optical filter silk-screen defect detection method and device and terminal equipment | |
CN108152278B (en) | Urine detection method and device | |
JP6325520B2 (en) | Unevenness inspection system, unevenness inspection method, and unevenness inspection program | |
US20170154234A1 (en) | Information processing device, information processing method, computer-readable recording medium, and inspection system | |
JP6664557B2 (en) | Deformation detection device | |
CN111047615B (en) | Image-based straight line detection method and device and electronic equipment | |
CN110736610B (en) | Method and device for measuring optical center deviation, storage medium and depth camera | |
CN111598801B (en) | Identification method for weak Mura defect | |
CN117058154B (en) | Defect identification method, system and medium for 3DP metal printing powder spreading process | |
CN116109635B (en) | Method, device, equipment and medium for detecting surface quality of composite suspension insulator | |
CN116152261B (en) | Visual inspection system for quality of printed product | |
EP2226763A2 (en) | Image processing method and apparatus for detecting an edge of an object within an image | |
CN107230212B (en) | Vision-based mobile phone size measuring method and system | |
CN116883336A (en) | Image processing method, device, computer equipment and medium | |
KR101842535B1 (en) | Method for the optical detection of symbols | |
CN117557565B (en) | Detection method and device for lithium battery pole piece | |
CN108399617B (en) | Method and device for detecting animal health condition | |
CN113570578B (en) | Method and device for detecting lens ghost phenomenon | |
KR20160001868A (en) | Method for calibrating distortion of image in camera | |
CN109313811B (en) | Automatic correction method, device and system based on vibration displacement of vision system | |
CN115760653A (en) | Image correction method, device, equipment and readable storage medium | |
CN111369513A (en) | Abnormity detection method, abnormity detection device, terminal equipment and storage medium | |
US20200304739A1 (en) | Image processing device, image processing method and storage medium |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
CP01 | Change in the name or title of a patent holder | ||
CP01 | Change in the name or title of a patent holder |
Address after: Room 495, building 3, 1197 Bin'an Road, Binjiang District, Hangzhou City, Zhejiang Province 310051 Patentee after: Yi Si Si (Hangzhou) Technology Co.,Ltd. Address before: Room 495, building 3, 1197 Bin'an Road, Binjiang District, Hangzhou City, Zhejiang Province 310051 Patentee before: ISVISION (HANGZHOU) TECHNOLOGY Co.,Ltd. |