[go: up one dir, main page]
More Web Proxy on the site http://driver.im/

CN109671079B - Corrugated board intelligent detection method based on gray level co-occurrence matrix characteristics - Google Patents

Corrugated board intelligent detection method based on gray level co-occurrence matrix characteristics Download PDF

Info

Publication number
CN109671079B
CN109671079B CN201811585653.5A CN201811585653A CN109671079B CN 109671079 B CN109671079 B CN 109671079B CN 201811585653 A CN201811585653 A CN 201811585653A CN 109671079 B CN109671079 B CN 109671079B
Authority
CN
China
Prior art keywords
image
straight line
gray level
follows
occurrence matrix
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
CN201811585653.5A
Other languages
Chinese (zh)
Other versions
CN109671079A (en
Inventor
于兴虎
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.)
Ningbo Intelligent Equipment Research Institute Co.,Ltd.
Original Assignee
Ningbo Intelligent Equipment Research Institute Co ltd
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 Ningbo Intelligent Equipment Research Institute Co ltd filed Critical Ningbo Intelligent Equipment Research Institute Co ltd
Priority to CN201811585653.5A priority Critical patent/CN109671079B/en
Publication of CN109671079A publication Critical patent/CN109671079A/en
Application granted granted Critical
Publication of CN109671079B publication Critical patent/CN109671079B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G06T7/0004—Industrial image inspection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/11—Region-based segmentation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/20—Image preprocessing
    • G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
    • 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
    • G06T2207/10024—Color 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
    • G06T2207/20—Special algorithmic details
    • G06T2207/20036—Morphological image processing
    • 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
    • G06T2207/20048—Transform domain processing
    • G06T2207/20061—Hough transform
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20081—Training; Learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30108—Industrial image inspection
    • G06T2207/30124—Fabrics; Textile; Paper
    • 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
    • G06T2207/30242—Counting objects in image

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Physics & Mathematics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Data Mining & Analysis (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • General Engineering & Computer Science (AREA)
  • Artificial Intelligence (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Multimedia (AREA)
  • Quality & Reliability (AREA)
  • Image Analysis (AREA)

Abstract

The invention relates to an intelligent detection method for a corrugated board based on gray level co-occurrence matrix characteristics. The method mainly solves the problems that the existing corrugated board counting method is not suitable for corrugated board accumulation and has low detection accuracy. An intelligent corrugated board detection method based on gray level co-occurrence matrix characteristics comprises the following steps: firstly, intercepting an interested area of an image; step two, converting the image into a gray image; step three, binaryzation of a gray level image; step four, performing binary image corrosion operation; step five, detecting straight lines; step six, linear screening, clustering and the like. The intelligent detection method for the corrugated board based on the gray level co-occurrence matrix characteristics can adapt to the conditions that the inclined angle exists in the corrugated layer in the corrugated paper stack, the gap exists between the layers and the like, and has high detection accuracy.

Description

Corrugated board intelligent detection method based on gray level co-occurrence matrix characteristics
Technical Field
The invention relates to the technical field of machine vision, in particular to an intelligent detection method for a corrugated board based on gray level co-occurrence matrix characteristics.
Background
Corrugated board is a multi-layer adhesive body, which is composed of at least one layer of wavy core paper interlayer (commonly called as "hole paper", "corrugated core paper", "corrugated paper core", "corrugated base paper") and one layer of paper board (also called as "box board paper" or "box paper board"). The material has good compressive strength and shock resistance, and can bear certain pressure, impact and vibration; the light-weight and cheap-price printing ink can be produced in various sizes in a large scale, has small storage space before use, and can be printed with various patterns, thereby being widely applied to packaging and transportation of finished products.
In large-scale production, the corrugated board needs to be counted, but a large amount of time is consumed for manually counting, and counting errors can be generated by manual work due to fatigue and the like along with the increase of the working time.
The computer vision identification is based on image processing counting to extract the information of the corrugated paper, but the existing corrugated paper identification algorithm mainly has the following problems: in practice, the corrugated cardboard may not be stacked tightly enough, gaps are easy to occur between every layer of corrugated cardboard, and the stacking is difficult to ensure the level, and the existing method cannot detect accurately in these situations.
Disclosure of Invention
In order to overcome the defects of the background art, the invention provides an intelligent corrugated board detection method based on gray level co-occurrence matrix characteristics, and mainly solves the problems that the existing corrugated board counting method is not suitable for corrugated board stacking and has low detection accuracy.
The technical scheme adopted by the invention is as follows: an intelligent corrugated board detection method based on gray level co-occurrence matrix characteristics comprises the following steps:
step one, intercepting an interested area of an image: intercepting an interested area of a corrugated surface image of a corrugated board collected by an industrial camera, wherein the width of the interested area image is W, and the height of the interested area image is H;
step two, converting into a gray image: if the image of the region of interest obtained in the step one is a multi-channel color image, converting the image of the region of interest into a gray image;
step three, binarization of the gray level image: carrying out local self-adaptive binarization on the gray level image obtained in the step two to obtain a binary image;
step four, binary image corrosion operation: carrying out corrosion operation on the binary image obtained in the step three to obtain a corroded binary image;
step five, straight line detection: carrying out straight line detection on the binary image corroded in the fourth step by using a Hough straight line detection algorithm to obtain a straight line set L0＝{(ρi,θi) 1, 1.., n }, where ρ isiIs the distance of the straight line i from the coordinate point (0,0), θiIs the angle of the line i (0 represents the vertical, pi/2 represents the horizontal), n is the number of lines;
step six, linear screening and clustering: for the straight line set L of the step five0Carrying out rough angle screening and clustering to obtain a straight line set L1The specific process is as follows: traverse L0The straight line of (1) is,
sixthly, initializing i to 1;
sixthly, judging the angle theta of the straight line iiIf theta | |iIf the ratio of the pi to the 2 is less than or equal to 0.05 pi, performing the sixth step, otherwise, performing the sixth step;
step six and three, if set L1Null, add line i to set L1Performing the following steps; if set L1If not, then traverse L1A middle straight line; if there is a straight line j such that | ρj-ρi< 7 and | θj-θiIf | is less than π/120, then update θj＝(θj+θi)/2，ρj＝(ρj+ρi) If not, add the straight line i to the set L1Performing the following steps;
sixthly, updating i to i +1, and repeating the step sixthly until i is larger than n;
step seven, calculating the average value of the straight line angles: set L of calculation steps six1Mean of the angles of the middle straight line
Figure BDA0001919021630000021
n1Is L1The number of middle straight lines; then set L1The angle of the middle straight line satisfies
Figure BDA0001919021630000022
Is put in the set L2And calculate the set L2Mean of the angles of the middle straight line
Figure BDA0001919021630000023
n2Is L2The number of middle straight lines;
step eight, correcting the binary image: rotating the binary image obtained in the step three by using coordinates (0,0) as an origin to rotate anticlockwise for alpha to obtain a rotated binary image, wherein
Figure BDA0001919021630000031
Ninth, image segmentation: dividing the rotated image obtained in the step eight into n21 image blocks, constituting a set SP＝{Pk,k＝1,...,n2-1}, wherein image block PkThe coordinate of the upper left pixel point in the rotated image is (0, ρ)k) Width wkFor the image width after rotation, height hkIs rhok+1-ρk；
Step ten, calculating the gray scale of each pixel in each image block: each image block P in the calculating step ninekMaximum pixel value ofmax,kThen calculate PkThe calculation formula of the gray level of each pixel is as follows:
Figure BDA0001919021630000032
wherein Ik(i, j) is the pixel value of the pixel (i, j) in the image block, NlevelSelecting 16 as the number of gray levels;
step eleven, calculating a gray level co-occurrence matrix of each image block in the horizontal direction: calculating a gray level co-occurrence matrix in the horizontal direction according to the gray level of the pixel in each image block obtained in the step ten, wherein the image block PkHas a gray level co-occurrence matrix of GkThe calculation formula is as follows:
Figure BDA0001919021630000033
wherein G isk(p, q) is a matrix GkThe value of the p-th row and q-th column, I [. cndot.)]To indicate a function, if the equation in parentheses is true, the function value is 1, otherwise 0, and the resulting matrix G iskSize Nlevel×Nlevel；
Step twelve, extracting the characteristics of the gray level co-occurrence matrix: firstly, normalizing the gray level co-occurrence matrix of each image block obtained in the step eleven, wherein the formula is as follows:
G'k(p,q)＝Gk(p,q)/Gk,max
wherein G isk,maxIs a matrix GkThe maximum value of the medium element, and four characteristics of energy, entropy, contrast and inverse difference moment are calculated, wherein the energy characteristic Fk,energyThe calculation formula is as follows:
Figure BDA0001919021630000041
entropy feature Fk,entropyThe calculation formula is as follows:
Figure BDA0001919021630000042
contrast characteristic Fk,contrastThe calculation formula is as follows:
Figure BDA0001919021630000043
inverse difference moment characteristic Fk,momentThe calculation formula is as follows:
Figure BDA0001919021630000044
four features constitute a feature vector
Figure BDA0001919021630000045
Step thirteen, detecting and judging the corrugated paper layer by a classifier: using a pre-trained support vector machine classifier to classify each image block PkCharacteristic F ofkMaking class prediction if feature FkIs between the intervals (1-epsilon, 1+ epsilon), the image block P is consideredkIs a corrugated layer, otherwise, is not, wherein e is 10-4。
Step four, binary image corrosion operation: and carrying out corrosion operation on the binary image obtained in the step three to obtain a corroded binary image, wherein the width of a corrosion element is 31, and the height of the corrosion element is 3.
The method for realizing the local self-adaptive binarization in the third step comprises the following steps: the threshold t (i, j) of the pixel with coordinate (i, j) is calculated as follows:
Figure BDA0001919021630000046
wherein I (I, j) is the pixel value of the gray scale image coordinate (I, j);
the binarization formula is as follows:
Figure BDA0001919021630000047
wherein IB(i, j) is the pixel value of the binarized image coordinate (i, j).
In the step ten, NlevelAnd is selected to be 16.
The training method of the pre-trained support vector machine classifier in the step thirteen is as follows: and for a plurality of corrugated paper images acquired by the industrial camera, segmenting the corrugated paper images in the steps from one to nine to obtain a plurality of image blocks, manually marking each image block as a positive sample and a negative sample, then extracting a feature vector for each image block in the steps from ten to twelve, and finally training a support vector machine classifier by using the feature vector and the mark of each image block.
The invention has the beneficial effects that:
selecting an interesting region of an original image, carrying out pretreatment such as binarization and corrosion, then carrying out Hough line detection on the image, screening and clustering according to line angles, and calculating the position and angle average value of the line; then, correcting the image and dividing the image into a plurality of image blocks; finally, extracting the gray level co-occurrence matrix characteristics of each image block, and judging the corrugated layer by using an offline-trained support vector machine classifier; the method can be suitable for the conditions that the corrugated layer in the corrugated paper stack has an inclination angle, and gaps exist between layers, and the like, and has high detection accuracy.
Detailed Description
The following examples of the invention are further illustrated:
an intelligent corrugated board detection method based on gray level co-occurrence matrix characteristics comprises the following steps:
step one, intercepting an interested area of an image: intercepting an interested area of a corrugated surface image of a corrugated board collected by an industrial camera, wherein the width of the interested area image is W, and the height of the interested area image is H;
step two, converting into a gray image: if the image of the region of interest obtained in the step one is a multi-channel color image, converting the image of the region of interest into a gray image;
step three, binarization of the gray level image: carrying out local self-adaptive binarization on the gray level image obtained in the step two to obtain a binary image;
step four, binary image corrosion operation: carrying out corrosion operation on the binary image obtained in the step three to obtain a corroded binary image;
step five, straight line detection: carrying out straight line detection on the binary image corroded in the fourth step by using a Hough straight line detection algorithm to obtain a straight line set L0＝{(ρi,θi) 1, 1.., n }, where ρ isiIs the distance of the straight line i from the coordinate point (0,0), θiIs the angle of the line i (0 represents the vertical, pi/2 represents the horizontal), n is the number of lines;
step six, linear screening and clustering: for the straight line set L of the step five0Carrying out rough angle screening and clustering to obtain a straight line set L1The specific process is as follows: traverse L0The straight line of (1) is,
sixthly, initializing i to 1;
sixthly, judging the angle theta of the straight line iiIf theta | |iIf the ratio of the pi to the 2 is less than or equal to 0.05 pi, performing the sixth step, otherwise, performing the sixth step;
step six and three, if set L1Null, add line i to set L1Performing the following steps; if set L1If not, then traverse L1A middle straight line; if there is a straight line j such that | ρj-ρi< 7 and | θj-θiIf | is less than π/120, then update θj＝(θj+θi)/2，ρj＝(ρj+ρi) If not, add the straight line i to the set L1Performing the following steps;
sixthly, updating i to i +1, and repeating the step sixthly until i is larger than n;
step seven, calculating the average value of the straight line angles: calculating stepSet of six steps L1Mean of the angles of the middle straight line
Figure BDA0001919021630000061
n1Is L1The number of middle straight lines; then set L1The angle of the middle straight line satisfies
Figure BDA0001919021630000062
Is put in the set L2And calculate the set L2Mean of the angles of the middle straight line
Figure BDA0001919021630000063
n2Is L2The number of middle straight lines;
step eight, correcting the binary image: rotating the binary image obtained in the step three by using coordinates (0,0) as an origin to rotate anticlockwise for alpha to obtain a rotated binary image, wherein
Figure BDA0001919021630000064
Ninth, image segmentation: dividing the rotated image obtained in the step eight into n21 image blocks, constituting a set SP＝{Pk,k＝1,...,n2-1}, wherein image block PkThe coordinate of the upper left pixel point in the rotated image is (0, ρ)k) Width wkFor the image width after rotation, height hkIs rhok+1-ρk；
Step ten, calculating the gray scale of each pixel in each image block: each image block P in the calculating step ninekMaximum pixel value ofmax,kThen calculate PkThe calculation formula of the gray level of each pixel is as follows:
Figure BDA0001919021630000071
wherein Ik(i, j) is the pixel value of the pixel (i, j) in the image block, NlevelSelecting 16 as the number of gray levels;
step eleven, calculating a gray level co-occurrence matrix of each image block in the horizontal direction: calculating a gray level co-occurrence matrix in the horizontal direction according to the gray level of the pixel in each image block obtained in the step ten, wherein the image block PkHas a gray level co-occurrence matrix of GkThe calculation formula is as follows:
Figure BDA0001919021630000072
wherein G isk(p, q) is a matrix GkThe value of the p-th row and q-th column, I [. cndot.)]To indicate a function, if the equation in parentheses is true, the function value is 1, otherwise 0, and the resulting matrix G iskSize Nlevel×Nlevel；
Step twelve, extracting the characteristics of the gray level co-occurrence matrix: firstly, normalizing the gray level co-occurrence matrix of each image block obtained in the step eleven, wherein the formula is as follows:
G'k(p,q)＝Gk(p,q)/Gk,max
wherein G isk,maxIs a matrix GkThe maximum value of the medium element, and four characteristics of energy, entropy, contrast and inverse difference moment are calculated, wherein the energy characteristic Fk,energyThe calculation formula is as follows:
Figure BDA0001919021630000073
entropy feature Fk,entropyThe calculation formula is as follows:
Figure BDA0001919021630000074
contrast characteristic Fk,contrastThe calculation formula is as follows:
Figure BDA0001919021630000075
moment of adverse differenceCharacteristic Fk,momentThe calculation formula is as follows:
Figure BDA0001919021630000081
four features constitute a feature vector
Figure BDA0001919021630000082
Step thirteen, detecting and judging the corrugated paper layer by a classifier: using a pre-trained support vector machine classifier to classify each image block PkCharacteristic F ofkMaking class prediction if feature FkIs between the intervals (1-epsilon, 1+ epsilon), the image block P is consideredkIs a corrugated layer, otherwise, is not, wherein e is 10-4。
Step four, binary image corrosion operation: and carrying out corrosion operation on the binary image obtained in the step three to obtain a corroded binary image, wherein the width of a corrosion element is 31, and the height of the corrosion element is 3.
The method for realizing the local self-adaptive binarization in the third step comprises the following steps: the threshold t (i, j) of the pixel with coordinate (i, j) is calculated as follows:
Figure BDA0001919021630000083
wherein I (I, j) is the pixel value of the gray scale image coordinate (I, j);
the binarization formula is as follows:
Figure BDA0001919021630000084
wherein IB(i, j) is the pixel value of the binarized image coordinate (i, j).
In the step ten, NlevelAnd is selected to be 16.
The training method of the pre-trained support vector machine classifier in the step thirteen is as follows: and for a plurality of corrugated paper images acquired by the industrial camera, segmenting the corrugated paper images in the steps from one to nine to obtain a plurality of image blocks, manually marking each image block as a positive sample and a negative sample, then extracting a feature vector for each image block in the steps from ten to twelve, and finally training a support vector machine classifier by using the feature vector and the mark of each image block.
The invention has the beneficial effects that:
selecting an interesting region of an original image, carrying out pretreatment such as binarization and corrosion, then carrying out Hough line detection on the image, screening and clustering according to line angles, and calculating the position and angle average value of the line; then, correcting the image and dividing the image into a plurality of image blocks; finally, extracting the gray level co-occurrence matrix characteristics of each image block, and judging the corrugated layer by using an offline-trained support vector machine classifier; the method can be suitable for the conditions that the corrugated layer in the corrugated paper stack has an inclination angle, and gaps exist between layers, and the like, and has high detection accuracy.
The skilled person should understand that: although the invention has been described in terms of the above specific embodiments, the inventive concept is not limited thereto and any modification applying the inventive concept is intended to be included within the scope of the patent claims.

Claims (4)

1. A corrugated board intelligent detection method based on gray level co-occurrence matrix characteristics is characterized in that: the method comprises the following steps:
step one, intercepting an interested area of an image: intercepting an interested area of a corrugated surface image of a corrugated board collected by an industrial camera, wherein the width of the interested area image is W, and the height of the interested area image is H;
step two, converting into a gray image: if the image of the region of interest obtained in the step one is a multi-channel color image, converting the image of the region of interest into a gray image;
step three, binarization of the gray level image: carrying out local self-adaptive binarization on the gray level image obtained in the step two to obtain a binary image;
step four, binary image corrosion operation: carrying out corrosion operation on the binary image obtained in the step three to obtain a corroded binary image;
step five, straight line detection: carrying out straight line detection on the binary image corroded in the fourth step by using a Hough straight line detection algorithm to obtain a straight line set L0＝{(ρi，θi) 1., n }, where ρ isiIs the distance of the straight line i from the coordinate point (0,0), θiIs the angle of the straight line i, θiWhen 0, it represents a vertical line, θiWhen pi/2 represents a horizontal line, n is the number of straight lines;
step six, linear screening and clustering: for the straight line set L of the step five0Carrying out rough angle screening and clustering to obtain a straight line set L1The specific process is as follows: traverse L0The straight line of (1) is,
sixthly, initializing i to 1;
sixthly, judging the angle theta of the straight line iiIf theta | |iIf the ratio of the pi to the 2 is less than or equal to 0.05 pi, performing the sixth step, otherwise, performing the sixth step;
step six and three, if set L1Null, add line i to set L1Performing the following steps; if set L1If not, then traverse L1A middle straight line; if there is a straight line j such that | ρj-ρi< 7 and | θj-θiIf | is less than π/120, then update θj＝(θj+θi)/2，ρj＝(ρj+ρi) If not, add the straight line i to the set L1Performing the following steps;
sixthly, updating i to i +1, and repeating the step sixthly until i is larger than n;
step seven, calculating the average value of the straight line angles: set L of calculation steps six1Mean of the angles of the middle straight line
Figure FDA0002841229860000021
n1Is L1The number of middle straight lines; then set L1The angle of the middle straight line satisfies
Figure FDA0002841229860000022
Is put in the set L2And calculate the set L2Mean of the angles of the middle straight line
Figure FDA0002841229860000023
n2Is L2The number of middle straight lines;
step eight, correcting the binary image: rotating the binary image obtained in the step three by using coordinates (0,0) as an origin to rotate anticlockwise for alpha to obtain a rotated binary image, wherein
Figure FDA0002841229860000024
Ninth, image segmentation: dividing the rotated image obtained in the step eight into n21 image blocks, constituting a set SP＝{Pk|k＝1，...，n2-1}, wherein image block PkThe coordinate of the upper left pixel point in the rotated image is (0, ρ)k) Width wkFor the image width after rotation, height hkIs rhok+1-ρk；
Step ten, calculating the gray scale of each pixel in each image block: each image block P in the calculating step ninekMaximum pixel value ofmax，kThen calculate PkThe calculation formula of the gray level of each pixel is as follows:
Figure FDA0002841229860000025
wherein Ik(i, j) is the pixel value of the pixel (i, j) in the image block, NlevelSelecting 16 as the number of gray levels;
step eleven, calculating a gray level co-occurrence matrix of each image block in the horizontal direction: calculating a gray level co-occurrence matrix in the horizontal direction according to the gray level of the pixel in each image block obtained in the step ten, wherein the image block PkHas a gray level co-occurrence matrix of GkThe calculation formula is as follows:
Figure FDA0002841229860000026
wherein G isk(p, q) is a matrix GkThe value of the p-th row and q-th column, I [. cndot.)]To indicate a function, if the equation in parentheses is true, the function value is 1, otherwise 0, and the resulting matrix G iskSize Nlevel×Nlevel；
Step twelve, extracting the characteristics of the gray level co-occurrence matrix: firstly, normalizing the gray level co-occurrence matrix of each image block obtained in the step eleven, wherein the formula is as follows:
G′k(p，q)＝Gk(p，q)/Gk，max
wherein G isk，maxIs a matrix GkThe maximum value of the medium element, and four characteristics of energy, entropy, contrast and inverse difference moment are calculated, wherein the energy characteristic Fk，energyThe calculation formula is as follows:
Figure FDA0002841229860000031
entropy feature Fk，entropyThe calculation formula is as follows:
Figure FDA0002841229860000032
contrast characteristic Fk，contrastThe calculation formula is as follows:
Figure FDA0002841229860000033
inverse difference moment characteristic Fk，momentThe calculation formula is as follows:
Figure FDA0002841229860000034
four features constitute a feature vector
Figure FDA0002841229860000035
Step thirteen, detecting and judging the corrugated paper layer by a classifier: using a pre-trained support vector machine classifier to classify each image block PkCharacteristic F ofkMaking class prediction if feature FkIs between the intervals (1-epsilon, 1+ epsilon), the image block P is consideredkIs a corrugated layer, otherwise, is not, wherein e is 10-4。
2. The intelligent corrugated board detection method based on the gray level co-occurrence matrix characteristic according to claim 1, characterized in that: step four, binary image corrosion operation: and carrying out corrosion operation on the binary image obtained in the step three to obtain a corroded binary image, wherein the width of a corrosion element is 31, and the height of the corrosion element is 3.
3. The intelligent corrugated board detection method based on the gray level co-occurrence matrix characteristic according to claim 1, characterized in that: the method for realizing the local self-adaptive binarization in the third step comprises the following steps: the threshold t (i, j) of the pixel with coordinate (i, j) is calculated as follows:
Figure FDA0002841229860000041
wherein I (I, j) is the pixel value of the gray scale image coordinate (I, j);
the binarization formula is as follows:
Figure FDA0002841229860000042
wherein IB(i, j) is the pixel value of the binarized image coordinate (i, j).
4. The intelligent corrugated board detection method based on the gray level co-occurrence matrix characteristic according to claim 1, characterized in that: the training method of the pre-trained support vector machine classifier in the step thirteen is as follows: and for a plurality of corrugated paper images acquired by the industrial camera, segmenting the corrugated paper images in the steps from one to nine to obtain a plurality of image blocks, manually marking each image block as a positive sample and a negative sample, then extracting a feature vector for each image block in the steps from ten to twelve, and finally training a support vector machine classifier by using the feature vector and the mark of each image block.
CN201811585653.5A 2018-12-25 2018-12-25 Corrugated board intelligent detection method based on gray level co-occurrence matrix characteristics Active CN109671079B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201811585653.5A CN109671079B (en) 2018-12-25 2018-12-25 Corrugated board intelligent detection method based on gray level co-occurrence matrix characteristics

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201811585653.5A CN109671079B (en) 2018-12-25 2018-12-25 Corrugated board intelligent detection method based on gray level co-occurrence matrix characteristics

Publications (2)

Publication Number Publication Date
CN109671079A CN109671079A (en) 2019-04-23
CN109671079B true CN109671079B (en) 2021-04-16

Family

ID=66146126

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201811585653.5A Active CN109671079B (en) 2018-12-25 2018-12-25 Corrugated board intelligent detection method based on gray level co-occurrence matrix characteristics

Country Status (1)

Country Link
CN (1) CN109671079B (en)

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110490207B (en) * 2019-08-27 2023-07-18 河北科技大学 Bar end face character image recognition method based on binary image gray level co-occurrence matrix
CN111882562A (en) * 2020-07-08 2020-11-03 谢超奇 Carton whole humidity analysis system
CN113850808B (en) * 2021-12-01 2022-03-22 武汉泰盛包装材料有限公司 Multilayer corrugated paper arrangement defect detection method and device based on image processing
CN113935996B (en) * 2021-12-16 2022-03-15 武汉市菲利纸业有限责任公司 Corrugated paper corrugation penetration phenomenon detection method and system based on image processing

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101609555A (en) * 2009-07-27 2009-12-23 浙江工商大学 A kind of gray-scale template matching method based on gray level co-occurrence matrixes
CN104715240A (en) * 2015-03-25 2015-06-17 华中科技大学 Hong Kong dollar texture image version classifying method based on gray-level co-occurrence matrix
CN108090511A (en) * 2017-12-15 2018-05-29 泰康保险集团股份有限公司 Image classification method, device, electronic equipment and readable storage medium storing program for executing

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6970592B2 (en) * 2000-09-04 2005-11-29 Fujitsu Limited Apparatus and method for correcting distortion of input image
JP5117211B2 (en) * 2008-02-05 2013-01-16 株式会社リコー Image forming apparatus

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101609555A (en) * 2009-07-27 2009-12-23 浙江工商大学 A kind of gray-scale template matching method based on gray level co-occurrence matrixes
CN104715240A (en) * 2015-03-25 2015-06-17 华中科技大学 Hong Kong dollar texture image version classifying method based on gray-level co-occurrence matrix
CN108090511A (en) * 2017-12-15 2018-05-29 泰康保险集团股份有限公司 Image classification method, device, electronic equipment and readable storage medium storing program for executing

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
GLCM texture based fractal method for evaluating fabric surface roughness;Xin Wang et al.;《2009 Canadian Conference on Electrical and Computer Engineering》;20090619;第104-107页 *
灰度共生矩阵纹理特征提取算法的优化与实现;吴文涵 等;《数字技术与应用》;20150630;第1-3页 *

Also Published As

Publication number Publication date
CN109671079A (en) 2019-04-23

Similar Documents

Publication Publication Date Title
CN109671079B (en) Corrugated board intelligent detection method based on gray level co-occurrence matrix characteristics
CN112465809B (en) Mold defect detection method based on image recognition, computer readable storage medium and device
CN106557764A (en) A kind of water level recognition methodss based on binary-coded character water gauge and image procossing
CN104866862A (en) Strip steel surface area type defect identification and classification method
CN111340875B (en) Space moving target detection method based on three-dimensional laser radar
CN111145228A (en) Heterogeneous image registration method based on local contour point and shape feature fusion
CN108573511B (en) Point-distributed cooperative coding mark and identification and positioning method thereof
CN113689429B (en) Wood board defect detection method based on computer vision
CN111127417B (en) Printing defect detection method based on SIFT feature matching and SSD algorithm improvement
CN109658404A (en) A kind of corrugated board intelligent detecting method extracted based on image outline with cluster
CN110223310B (en) Line structure light center line and box edge detection method based on deep learning
CN112233116A (en) Concave-convex mark visual detection method based on neighborhood decision and gray level co-occurrence matrix description
CN106803245B (en) Railway roadbed state evaluation method based on ground penetrating radar periodic detection
CN111161260A (en) Hot-rolled strip steel surface defect detection method and device based on deep learning
CN114219794A (en) Method and system for evaluating surface quality of shaving board based on machine vision
CN114022439A (en) Flexible circuit board defect detection method based on morphological image processing
CN109685791B (en) Corrugated board robust detection method based on multi-window autocorrelation texture detection
CN104112123A (en) Defect characteristic extraction and identification method of AOI system used for bullet apparent defect detection
CN115063620B (en) Bit layering based Roots blower bearing wear detection method
CN109711405B (en) Non-supervised extraction method of corrugated board based on skeleton self-correction
CN115147418A (en) Compression training method and device for defect detection model
CN114612458A (en) Flexo printing first piece detection method based on electronic sample manuscript
CN112381844B (en) Self-adaptive ORB feature extraction method based on image blocking
CN105868794B (en) A kind of Ship Target fuzzy recognition method based on ISAR image
CN113516123A (en) Detection and identification method for tire embossed characters

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
TA01 Transfer of patent application right

Effective date of registration: 20210326

Address after: 315200 No.189 Guangming Road, Zhuangshi street, Zhenhai District, Ningbo City, Zhejiang Province

Applicant after: Ningbo Intelligent Equipment Research Institute Co.,Ltd.

Address before: 264006 No. 32, Yantai economic and Technological Development Zone, Shandong, Zhujianglu Road

Applicant before: Yu Xinghu

TA01 Transfer of patent application right
GR01 Patent grant
GR01 Patent grant