CN110189297B - Magnetic material appearance defect detection method based on gray level co-occurrence matrix - Google Patents

Magnetic material appearance defect detection method based on gray level co-occurrence matrix Download PDF

Info

Publication number
CN110189297B
CN110189297B CN201910313134.1A CN201910313134A CN110189297B CN 110189297 B CN110189297 B CN 110189297B CN 201910313134 A CN201910313134 A CN 201910313134A CN 110189297 B CN110189297 B CN 110189297B
Authority
CN
China
Prior art keywords
magnetic material
gray level
row
sub
value
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
Application number
CN201910313134.1A
Other languages
Chinese (zh)
Other versions
CN110189297A (en
Inventor
何志伟
江灿军
高明煜
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Hangzhou Dianzi University
Original Assignee
Hangzhou Dianzi University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Hangzhou Dianzi University filed Critical Hangzhou Dianzi University
Priority to CN201910313134.1A priority Critical patent/CN110189297B/en
Publication of CN110189297A publication Critical patent/CN110189297A/en
Application granted granted Critical
Publication of CN110189297B publication Critical patent/CN110189297B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

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
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G06T7/0004—Industrial image inspection
    • G—PHYSICS
    • 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
    • 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
    • G01N2021/8854—Grading and classifying of flaws
    • G01N2021/8861—Determining coordinates of flaws
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/84—Systems specially adapted for particular applications
    • G01N21/88—Investigating the presence of flaws or contamination
    • G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
    • G01N2021/8887—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges based on image processing techniques
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10004—Still image; Photographic image
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20021—Dividing image into blocks, subimages or windows
    • 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
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30108—Industrial image inspection

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Biochemistry (AREA)
  • Chemical & Material Sciences (AREA)
  • Analytical Chemistry (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Immunology (AREA)
  • Pathology (AREA)
  • Quality & Reliability (AREA)
  • Signal Processing (AREA)
  • Investigating Or Analyzing Materials By The Use Of Magnetic Means (AREA)
  • Image Analysis (AREA)

Abstract

The invention discloses a magnetic material appearance defect detection method based on a gray level co-occurrence matrix. The traditional manual quality inspection is not only low in efficiency, but also high in cost. The method comprises the steps of firstly carrying out binarization processing on a magnetic material horizontal projection gray level image, calculating to obtain the positioning of a magnetic material center point, evenly dividing a magnetic material region in a source image into a plurality of sub-regions, calculating gray level co-occurrence matrixes of the plurality of sub-regions, and screening out a defect region by combining the characteristic entropy value of the gray level co-occurrence matrixes, thereby realizing the detection of the appearance defect of the magnetic material. The method disclosed by the invention can be suitable for detecting the appearance defects of various magnetic materials, and not only can solve the problems encountered in the actual production process of an enterprise, but also can make up for the technical vacancy in the field. The method is simple to apply and has high detection efficiency.

Description

Magnetic material appearance defect detection method based on gray level co-occurrence matrix
Technical Field
The invention belongs to the technical field of image processing, and particularly relates to a magnetic material appearance defect detection method based on a gray level co-occurrence matrix.
Background
The magnetic material has wide application field, can be applied to the traditional manufacturing industries of sound equipment, earphones and the like, and can also be applied to a plurality of high and new technology industries of aviation, communication, computers and the like. With the continuous development of high-tech industries, the application of magnetic materials will be more and more extensive. In order to improve the quality of the magnetic material, it is necessary to perform quality inspection on all the magnetic materials before the magnetic materials are shipped. At present, because no mature technology is available for detecting the appearance defects of the magnetic material basically, most of small and medium-sized manufacturers adopt an artificial quality inspection mode. Because the magnetic material has high yield, a large amount of labor is required to be consumed, the enterprise cost is directly increased, and the economic benefit of the enterprise is further reduced. Meanwhile, manual detection has many defects, the health of workers can be damaged when the magnetic material works in a severe factory environment for a long time, and the high quality of the magnetic material cannot be guaranteed due to factors such as manual omission and false detection.
Disclosure of Invention
The invention aims to provide a magnetic material appearance defect detection method based on a gray level co-occurrence matrix.
The method comprises the following specific steps:
step (1), f (x, y) represents a collected magnetic material horizontal projection gray level image with the size of H multiplied by V, and the magnetic material gray level image is subjected to binarization processing:
Figure BDA0002032186460000011
where x, y represent the abscissa and ordinate of the image, η represents the binarization threshold, and b (x, y) represents the binarized image.
Step (2), positioning the center point of the magnetic material:
(2-1) accessing each pixel point of the binary image b (x, y) in a scanning mode from top to bottom, and accumulating the black pixel points in a row until the row of the black pixel points are accumulated and Countup≥η1Stopping, recording the Row number as Rowup；
Accessing each pixel point of the binary image b (x, y) in a scanning mode from bottom to top, and accumulating the black pixel points according to a line until the black pixel points of the line are accumulated and Countdown≥η1Stopping, recording the Row number as Rowdown；
RowupIndicating the ordinate, Row, of the upper edge of the magnetic materialdownRepresenting the ordinate, eta, of the lower edge of the magnetic material1A threshold value representing the upper and lower edges of the magnetic material to be screened;
(2-2) calculating the ordinate Row of the center point of the magnetic materialmiddle：
Figure BDA0002032186460000021
(2-3) scanning from left to rightInquiring each pixel point of binary image b (x, y), accumulating black pixel points according to row until the row of black pixel points are accumulated and Countleft≥η2Stopping, recording the column number as Colleft；
Accessing each pixel point of the binary image b (x, y) in a scanning mode from right to left, and accumulating the black pixel points in a row until the row of black pixel points are accumulated and Countright≥η2Stopping, recording the column number as Colright；
ColleftThe abscissa representing the left edge of the magnetic material, ColrightAbscissa, η, representing right edge of magnetic material2A threshold value representing the left and right edges of the screened magnetic material;
(2-4) calculating the abscissa Col of the center point of the magnetic materialmiddle：
Figure BDA0002032186460000022
(2-5) obtaining the coordinate of the center point of the magnetic material as (Col)middle,Rowmiddle)。
Step (3), equally dividing a magnetic material region in a source image f (x, y) into N sub-regions with the size of M multiplied by N:
(3-1) calculating the Width and Height of the magnetic material:
Figure BDA0002032186460000023
(3-2) selecting a size of the sub-region, mxn, that satisfies:
Figure BDA0002032186460000024
(3-3) traversing all the sub-regions in a scanning mode from left to right and from top to bottom, wherein the coordinate (col) of the center point of the ith sub-regioni,rowi) Comprises the following steps:
Figure BDA0002032186460000025
p denotes the column number traversing the sub-regions in a left-to-right scanning manner, q denotes the scanning manner from top to bottomThe row number of the sub-region is traversed.
Step (4), calculating the gray level co-occurrence matrix of the n sub-regions:
(4-1) selecting a proper gray level k, and carrying out gray level conversion on a source image f (x, y):
Figure BDA0002032186460000026
wherein, [ f (x, y)/T]An integer part is obtained by dividing f (x, y) by T, T represents the maximum difference value of different gray values of the same gray level, and the range of the gray level k is 1-256 levels;
(4-2) calculating a gray level co-occurrence matrix G of the ith sub-regioni：
Figure BDA0002032186460000031
Gray level co-occurrence matrix GiItem g in (1)rsIndicates that a pixel point pair s satisfying the following condition appears in the sub-area1(x1,y1) And s2(x2,y2) The number of times of (2):
Figure BDA0002032186460000032
direct＝0°；
Figure BDA0002032186460000033
direct＝45°；
Figure BDA0002032186460000034
direct＝90°；
Figure BDA0002032186460000035
direct＝135°；
wherein, the pixel point s1(x1,y1) Has a gray value of (r-1) and a pixel point s2(x2,y2) Has a gray value of (s-1), d represents s1And s2Pixel pitch between, direct denotes s1And s2The relative orientation therebetween;
(4-3) computing a gray level co-occurrence matrix GiEntropy value E ofi：
Figure BDA0002032186460000036
arsIs a gray level co-occurrence matrix GiItem g in (1)rsA normalized value;
and (4-4) repeatedly iterating the step (4-2) and the step (4-3) until all the n sub-regions are calculated.
And (5) manufacturing entropy values of gray level co-occurrence matrixes of all defect regions and normal regions into data samples, and carrying out K-Means clustering:
(5-1) randomly selecting w entropy values in a data sample as initial objects, wherein each object represents a clustering center;
(5-2) for all entropy values, according to Euclidean distances between the entropy values and the clustering centers, allocating the entropy values to the class corresponding to the clustering center closest to the entropy values according to the closest criterion;
(5-3) updating the clustering center: taking the mean value corresponding to all the objects in each class as the clustering center of the class, and calculating the value of an objective function, wherein the objective function is as follows:
Figure BDA0002032186460000037
wherein u isjRepresenting the value of the jth one of the w cluster centers, eiRepresenting the ith object value in the m objects of the class;
and (5-4) judging whether the values of the clustering center and the objective function are changed, if not, successfully clustering, otherwise, starting iteration from the step (5-2) again.
Step (6) after the clustering is finished, setting a threshold value E according to a clustering resultT，ET2-10 parts of ═ a; if the entropy value E of any sub-region of the magnetic material is more than or equal to ETIf the magnetic material is defective, otherwise, the magnetic material is genuine. Thereby realizing the detection of the appearance defects of the magnetic material.
According to a first aspect of the present invention, a region segmentation for an algorithm for detecting apparent defects of a magnetic material is disclosed, the magnetic material being equally segmented into an integer number of sub-regions.
According to a second aspect of the invention, a specific flow of a gray level co-occurrence matrix algorithm for detecting appearance defects of magnetic materials is disclosed. The method mainly comprises the following steps: and calculating the gray level co-occurrence matrix of each region, then calculating the entropy value of each gray level co-occurrence matrix, distinguishing the entropy values of the defect region and the normal region by a clustering method, and setting a threshold value to screen out the defect region, thereby detecting the appearance defect of the magnetic material.
The method combines the characteristics of the magnetic material, fuses the characteristics of the gray level co-occurrence matrix, and can accurately detect the appearance defects of the magnetic material. The method has strong universality and high algorithm robustness, can be applied to the detection of the appearance defects of various magnetic materials, and has strong applicability and better detection effect.
Detailed Description
The present invention will be further described with reference to the following embodiments.
A magnetic material appearance defect detection method based on a gray level co-occurrence matrix comprises the following specific steps:
step (1), f (x, y) represents a collected magnetic material horizontal projection gray level image with the size of H multiplied by V, and the magnetic material gray level image is subjected to binarization processing:
Figure BDA0002032186460000041
where x, y represent the abscissa and ordinate of the image, η represents the binarization threshold, and b (x, y) represents the binarized image. In this example, η is set to 200.
Step (2), positioning the center point of the magnetic material:
(2-1) accessing each pixel point of the binary image b (x, y) in a scanning mode from top to bottom, and accumulating the black pixel points in a row until the row of the black pixel points are accumulated and Countup≥η1Stopping, recording the Row number as Rowup；
Accessing each pixel point of the binary image b (x, y) in a scanning mode from bottom to top, and accumulating the black pixel points according to a line until the black pixel points of the line are accumulated and Countdown≥η1Stop, record the lineRow line numberdown；
RowupIndicating the ordinate, Row, of the upper edge of the magnetic materialdownRepresenting the ordinate, eta, of the lower edge of the magnetic material1A threshold value representing the upper and lower edges of the magnetic material to be screened;
(2-2) calculating the ordinate Row of the center point of the magnetic materialmiddle：
Figure BDA0002032186460000042
(2-3) accessing each pixel point of the binary image b (x, y) in a scanning mode from left to right, and accumulating the black pixel points in a row until the row of black pixel points are accumulated and Countleft≥η2Stopping, recording the column number as Colleft；
Accessing each pixel point of the binary image b (x, y) in a scanning mode from right to left, and accumulating the black pixel points in a row until the row of black pixel points are accumulated and Countright≥η2Stopping, recording the column number as Colright；
ColleftThe abscissa representing the left edge of the magnetic material, ColrightAbscissa, η, representing right edge of magnetic material2A threshold value representing the left and right edges of the screened magnetic material;
(2-4) calculating the abscissa Col of the center point of the magnetic materialmiddle：
Figure BDA0002032186460000051
(2-5) obtaining the coordinate of the center point of the magnetic material as (Col)middle,Rowmiddle)。
Step (3), equally dividing a magnetic material region in a source image f (x, y) into N sub-regions with the size of M multiplied by N:
(3-1) calculating the Width and Height of the magnetic material:
Figure BDA0002032186460000052
(3-2) selecting size M of sub-regionN, satisfies:
Figure BDA0002032186460000053
in this example, mxn is 46 × 46, N1＝9,n2＝5；
(3-3) traversing all the sub-regions in a scanning mode from left to right and from top to bottom, wherein the coordinate (col) of the center point of the ith sub-regioni,rowi) Comprises the following steps:
Figure BDA0002032186460000054
p denotes a column number traversing the sub-region in a left-to-right scanning manner, and q denotes a row number traversing the sub-region in a top-to-bottom scanning manner.
Step (4), calculating the gray level co-occurrence matrix of the n sub-regions:
(4-1) selecting a proper gray level k, and carrying out gray level conversion on a source image f (x, y):
Figure BDA0002032186460000055
wherein, [ f (x, y)/T]The expression f (x, y) is divided by T to obtain an integer part, T represents the maximum difference value of different gray values of the same gray level, the range of the gray level k is 1-256 levels, and k is set to be 8 in the example;
(4-2) calculating a gray level co-occurrence matrix G of the ith sub-regioni：
Figure BDA0002032186460000061
Gray level co-occurrence matrix GiItem g in (1)rsIndicates that a pixel point pair s satisfying the following condition appears in the sub-area1(x1,y1) And s2(x2,y2) The number of times.
Figure BDA0002032186460000062
direct＝0°；
Figure BDA0002032186460000063
direct＝45°；
Figure BDA0002032186460000064
direct＝90°；
Figure BDA0002032186460000065
direct＝135°；
Wherein, the pixel point s1(x1,y1) Has a gray value of (r-1) and a pixel point s2(x2,y2) Has a gray value of (s-1), d represents s1And s2Pixel pitch between, direct denotes s1And s2The relative orientation between d and d is 1 and direct is 0 ° in this example. (ii) a
(4-3) computing a gray level co-occurrence matrix GiEntropy value E ofi：
Figure BDA0002032186460000066
arsIs a gray level co-occurrence matrix GiItem g in (1)rsA normalized value;
and (4-4) repeatedly iterating the step (4-2) and the step (4-3) until all the n sub-regions are calculated.
And (5) manufacturing entropy values of gray level co-occurrence matrixes of all defect regions and normal regions into data samples, and carrying out K-Means clustering:
(5-1) randomly selecting w entropy values in a data sample as initial objects, wherein each object represents a clustering center;
(5-2) for all entropy values, according to Euclidean distances between the entropy values and the clustering centers, allocating the entropy values to the classes corresponding to the clustering centers (most similar) closest to the entropy values according to the closest criterion;
(5-3) updating the clustering center: taking the mean value corresponding to all the objects in each class as the clustering center of the class, and calculating the value of an objective function, wherein the objective function is as follows:
Figure BDA0002032186460000067
wherein u isjRepresenting the value of the jth one of the w cluster centers, eiRepresenting the m objects of the classi object values;
and (5-4) judging whether the values of the clustering center and the objective function are changed, if not, successfully clustering, otherwise, starting iteration from the step (5-2) again.
Step (6) after the clustering is finished, setting a threshold value E according to a clustering resultT，ET2-10, the threshold value is set to be 5.98 in the example; if the entropy value E of any sub-region of the magnetic material is more than or equal to ETIf the magnetic material is defective, otherwise, the magnetic material is genuine. Thereby realizing the detection of the appearance defects of the magnetic material.

Claims (3)

1. A magnetic material appearance defect detection method based on gray level co-occurrence matrix is characterized by comprising the following specific steps:
step (1) f (x, y) represents an acquired horizontal projection gray level image of the magnetic material with the size of H multiplied by V, and the gray level image of the magnetic material is subjected to binarization processing:
Figure FDA0002735467900000011
wherein x, y represent the abscissa and ordinate of the image, η represents the binarization threshold, and b (x, y) represents the binarization image;
positioning the center point of the magnetic material:
(2-1) accessing each pixel point of the binary image b (x, y) in a scanning mode from top to bottom, and accumulating the black pixel points in a row until the row of the black pixel points are accumulated and Countup≥η1Stopping, recording the Row number as Rowup；
Accessing each pixel point of the binary image b (x, y) in a scanning mode from bottom to top, and accumulating the black pixel points according to a line until the black pixel points of the line are accumulated and Countdown≥η1Stopping, recording the Row number as Rowdown；
RowupIndicating the ordinate, Row, of the upper edge of the magnetic materialdownRepresenting the ordinate, eta, of the lower edge of the magnetic material1A threshold value representing the upper and lower edges of the magnetic material to be screened;
(2-2) calculation of magnetic MaterialOrdinate Row of the center pointmiddle：
Figure FDA0002735467900000012
(2-3) accessing each pixel point of the binary image b (x, y) in a scanning mode from left to right, and accumulating the black pixel points in a row until the row of black pixel points are accumulated and Countleft≥η2Stopping, recording the column number as Colleft；
Accessing each pixel point of the binary image b (x, y) in a scanning mode from right to left, and accumulating the black pixel points in a row until the row of black pixel points are accumulated and Countright≥η2Stopping, recording the column number as Colright；
ColleftThe abscissa representing the left edge of the magnetic material, ColrightAbscissa, η, representing right edge of magnetic material2A threshold value representing the left and right edges of the screened magnetic material;
(2-4) calculating the abscissa Col of the center point of the magnetic materialmiddle：
Figure FDA0002735467900000013
(2-5) obtaining the coordinate of the center point of the magnetic material as (Col)middle,Rowmiddle)；
Step (3) equally dividing the magnetic material region in the source image f (x, y) into N sub-regions with the size of M multiplied by N:
(3-1) calculating the Width Width and Height of the magnetic material:
Figure FDA0002735467900000021
(3-2) selecting a size of the sub-region, MxN, that satisfies:
Figure FDA0002735467900000029
(3-3) traversing all the sub-regions in a scanning mode from left to right and from top to bottom, wherein the center point of the ith sub-region is locatedLabel (col)i,rowi) Comprises the following steps:
Figure FDA0002735467900000024
p represents a column number for traversing the sub-region in a left-to-right scanning manner, and q represents a row number for traversing the sub-region in a top-to-bottom scanning manner;
step (4), calculating the gray level co-occurrence matrix of n sub-regions:
(4-1) selecting proper gray level k, and carrying out gray level conversion on the source image f (x, y):
Figure FDA0002735467900000025
wherein, [ f (x, y)/T]The expression f (x, y) is divided by T to obtain an integer part, and T represents the maximum difference value of different gray values of the same gray level;
(4-2) calculating a gray level co-occurrence matrix G of the ith sub-regioni：
Figure FDA0002735467900000026
Gray level co-occurrence matrix GiItem g in (1)rsIndicates that a pixel point pair s satisfying the following condition appears in the sub-area1(x1,y1) And s2(x2,y2) The number of times of (2):
Figure FDA0002735467900000027
Figure FDA0002735467900000028
wherein, the pixel point s1(x1,y1) Has a gray value of (r-1) and a pixel point s2(x2,y2) Has a gray value of (s-1), d represents s1And s2Pixel pitch between, direct denotes s1And s2The relative orientation therebetween;
(4-3) calculating the gray level co-occurrence matrix GiEntropy value E ofi：
Figure FDA0002735467900000031
arsIs a gray level co-occurrence matrix GiItem g in (1)rsA normalized value;
(4-4) repeating the step (4-2) and the step (4-3) until all the n sub-regions are calculated;
and (5) making entropy values of gray level co-occurrence matrixes of all defect regions and normal regions into data samples, and carrying out K-Means clustering:
(5-1) randomly selecting w entropy values in a data sample as initial objects, wherein each object represents a clustering center;
(5-2) for all entropy values, according to Euclidean distances between the entropy values and the clustering centers, allocating the entropy values to the class corresponding to the clustering center closest to the entropy values according to the closest criterion;
(5-3) updating the clustering center: taking the mean value corresponding to all the objects in each class as the clustering center of the class, and calculating the value of an objective function, wherein the objective function is as follows:
Figure FDA0002735467900000032
wherein u isjRepresenting the value of the jth one of the w cluster centers, eiRepresenting the ith object value in the m objects of the class;
(5-4) judging whether the values of the clustering center and the objective function are changed, if not, successfully clustering, otherwise, starting iteration from the step (5-2) again;
after the clustering in the step (6) is finished, setting a threshold value E according to a clustering resultT(ii) a If the entropy value E of any sub-region of the magnetic material is more than or equal to ETIf the magnetic material is a defective product, otherwise, the magnetic material is a good product, and the detection of the appearance defects of the magnetic material is realized.
2. The method for detecting the appearance defects of the magnetic material based on the gray level co-occurrence matrix according to claim 1, wherein the method comprises the following steps: the range of the gray level k in the step (4) is 1-256 levels.
3. The method for detecting the appearance defects of the magnetic material based on the gray level co-occurrence matrix according to claim 1, wherein the method comprises the following steps: the threshold E in step (6)T＝2～10。
CN201910313134.1A 2019-04-18 2019-04-18 Magnetic material appearance defect detection method based on gray level co-occurrence matrix Active CN110189297B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910313134.1A CN110189297B (en) 2019-04-18 2019-04-18 Magnetic material appearance defect detection method based on gray level co-occurrence matrix

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910313134.1A CN110189297B (en) 2019-04-18 2019-04-18 Magnetic material appearance defect detection method based on gray level co-occurrence matrix

Publications (2)

Publication Number Publication Date
CN110189297A CN110189297A (en) 2019-08-30
CN110189297B true CN110189297B (en) 2021-02-19

Family

ID=67714702

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910313134.1A Active CN110189297B (en) 2019-04-18 2019-04-18 Magnetic material appearance defect detection method based on gray level co-occurrence matrix

Country Status (1)

Country Link
CN (1) CN110189297B (en)

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112508922B (en) * 2020-12-14 2021-12-21 深圳精智达技术股份有限公司 Mura detection method, device, terminal equipment and storage medium
CN113577422B (en) * 2021-07-12 2023-05-26 健帆生物科技集团股份有限公司 Quality inspection method and assembly method of dialyzer
CN114049489B (en) * 2022-01-13 2022-04-12 山东鹰联光电科技股份有限公司 Early warning method for breakage of optical fiber cable sheath
CN117173160B (en) * 2023-10-31 2024-01-26 山东远通汽车智能科技有限公司 Image processing-based method for detecting quality of caravan component
CN117197128B (en) * 2023-11-02 2024-01-30 宁波明禾新能源科技有限公司 Product image detection method and system of bending machine and bending machine

Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104794491A (en) * 2015-04-28 2015-07-22 重庆大学 Fuzzy clustering steel plate surface defect detection method based on pre classification
CN104866862A (en) * 2015-04-27 2015-08-26 中南大学 Strip steel surface area type defect identification and classification method
CN106981073A (en) * 2017-03-31 2017-07-25 中南大学 A kind of ground moving object method for real time tracking and system based on unmanned plane
CN107038432A (en) * 2017-05-12 2017-08-11 西安电子科技大学 Fingerprint image orientation extracting method based on frequency information
CN107491763A (en) * 2017-08-24 2017-12-19 歌尔科技有限公司 Finger areas dividing method and device based on depth image
CN107664644A (en) * 2017-09-30 2018-02-06 电子科技大学 A kind of apparent automatic detection device of object based on machine vision and method
CN107680086A (en) * 2017-09-27 2018-02-09 电子科技大学 A kind of existing arc-shaped side has the material profile defect inspection method of straight line again
CN107784657A (en) * 2017-09-29 2018-03-09 西安因诺航空科技有限公司 A kind of unmanned aerial vehicle remote sensing image partition method based on color space classification
CN108447050A (en) * 2018-03-07 2018-08-24 湘潭大学 A kind of Surface Flaw dividing method based on super-pixel

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106683076B (en) * 2016-11-24 2019-08-09 南京航空航天大学 The method of train wheel tread damage detection based on textural characteristics cluster

Patent Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104866862A (en) * 2015-04-27 2015-08-26 中南大学 Strip steel surface area type defect identification and classification method
CN104794491A (en) * 2015-04-28 2015-07-22 重庆大学 Fuzzy clustering steel plate surface defect detection method based on pre classification
CN106981073A (en) * 2017-03-31 2017-07-25 中南大学 A kind of ground moving object method for real time tracking and system based on unmanned plane
CN107038432A (en) * 2017-05-12 2017-08-11 西安电子科技大学 Fingerprint image orientation extracting method based on frequency information
CN107491763A (en) * 2017-08-24 2017-12-19 歌尔科技有限公司 Finger areas dividing method and device based on depth image
CN107680086A (en) * 2017-09-27 2018-02-09 电子科技大学 A kind of existing arc-shaped side has the material profile defect inspection method of straight line again
CN107784657A (en) * 2017-09-29 2018-03-09 西安因诺航空科技有限公司 A kind of unmanned aerial vehicle remote sensing image partition method based on color space classification
CN107664644A (en) * 2017-09-30 2018-02-06 电子科技大学 A kind of apparent automatic detection device of object based on machine vision and method
CN108447050A (en) * 2018-03-07 2018-08-24 湘潭大学 A kind of Surface Flaw dividing method based on super-pixel

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
Automatic defect recognition of TFT array process using gray level co-occurrence matrix;Shih-Wei Yang等;《Optik》;20140630;全文 *
Defect Detection on Air Bearing Surface with Gray Level Co-occurrence Matrix;PICHATE KUNAKORNVONG等;《The 4th Joint International Conference on Information and Communication Technology, Electronic and Electrical Engineering (JICTEE-2014)》;20140424;全文 *
基于图像的踏面表面缺陷特征分析识别;杨颖;《中国优秀硕士学位论文全文数据库工程科技Ⅱ辑》;20180315;全文 *

Also Published As

Publication number Publication date
CN110189297A (en) 2019-08-30

Similar Documents

Publication Publication Date Title
CN110189297B (en) Magnetic material appearance defect detection method based on gray level co-occurrence matrix
CN114445387A (en) Fiberboard quality classification method based on machine vision
CN105894036B (en) A kind of characteristics of image template matching method applied to mobile phone screen defects detection
CN115239704B (en) Accurate detection and repair method for wood surface defects
CN104990925B (en) One kind is based on gradient multi thresholds optimization defect inspection method
CN115351598A (en) Numerical control machine tool bearing detection method
CN115100199B (en) Method for detecting wafer low-texture defects
CN114219794B (en) Method and system for evaluating surface quality of shaving board based on machine vision
CN108764361B (en) Working condition identification method of indicator diagram of beam-pumping unit based on integrated learning
CN114820625B (en) Automobile top block defect detection method
CN113506246B (en) Concrete 3D printing component fine detection method based on machine vision
CN116309577B (en) Intelligent detection method and system for high-strength conveyor belt materials
CN112907561A (en) Notebook appearance flaw detection method based on deep learning
CN116523913B (en) Intelligent detection method for quality of screw rod
CN115797361B (en) Aluminum template surface defect detection method
CN111402236A (en) Hot-rolled strip steel surface defect grading method based on image gray value
CN116664565A (en) Hidden crack detection method and system for photovoltaic solar cell
CN110648330A (en) Defect detection method for camera glass
CN117152119A (en) Profile flaw visual detection method based on image processing
CN114119603A (en) Image processing-based snack box short shot defect detection method
CN110458809B (en) Yarn evenness detection method based on sub-pixel edge detection
CN117635609B (en) Visual inspection method for production quality of plastic products
CN114565607A (en) Fabric defect image segmentation method based on neural network
CN117237747B (en) Hardware defect classification and identification method based on artificial intelligence
CN108805854B (en) Method for rapidly counting tablets and detecting completeness of tablets in complex environment

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