CN106530275A - Element wrong part detection method and system - Google Patents

Element wrong part detection method and system Download PDF

Info

Publication number
CN106530275A
CN106530275A CN201610887939.3A CN201610887939A CN106530275A CN 106530275 A CN106530275 A CN 106530275A CN 201610887939 A CN201610887939 A CN 201610887939A CN 106530275 A CN106530275 A CN 106530275A
Authority
CN
China
Prior art keywords
pixel
image
feature regional
regional images
images
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.)
Granted
Application number
CN201610887939.3A
Other languages
Chinese (zh)
Other versions
CN106530275B (en
Inventor
李红匣
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.)
Guangzhou Shiyuan Electronics Thecnology Co Ltd
Original Assignee
Guangzhou Shiyuan Electronics Thecnology 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 Guangzhou Shiyuan Electronics Thecnology Co Ltd filed Critical Guangzhou Shiyuan Electronics Thecnology Co Ltd
Priority to CN201610887939.3A priority Critical patent/CN106530275B/en
Priority to PCT/CN2016/113146 priority patent/WO2018068415A1/en
Publication of CN106530275A publication Critical patent/CN106530275A/en
Application granted granted Critical
Publication of CN106530275B publication Critical patent/CN106530275B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/22—Matching criteria, e.g. proximity measures
    • 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
    • G06T7/001—Industrial image inspection using an image reference approach
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
    • G06V10/75—Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g. multi-scale approaches; using context analysis; Selection of dictionaries
    • G06V10/751—Comparing pixel values or logical combinations thereof, or feature values having positional relevance, e.g. template matching
    • 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
    • G06T2207/30141—Printed circuit board [PCB]

Landscapes

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

Abstract

The present invention relates to an element wrong part detection method and a system thereof. The method comprises the steps of acquiring the original image of a to-be-detected element on a circuit board, and positioning the characteristic region image of the to-be-detected element in the original image, wherein the characteristic region image contains the characteristic information of the to-be-detected element and the characteristic information is used for distinguishing the to-be-detected element from other elements; respectively comparing the pixel value of each pixel point in the characteristic region image with the pixel value of a corresponding pixel point in a pre-stored characteristic region template image so as to obtain the pixel similarity between the characteristic region image and the characteristic region template image; on the condition that the pixel similarity between the characteristic region image and the characteristic region template image is smaller than a preset similarity threshold, judging the to-be-detected element to be a wrong part. Based on the above element wrong part detection method and the system thereof, the automation of wrong part detection is realized. The detection efficiency and the detection accuracy are effectively improved.

Description

Element mistake part detection method and system
Technical field
The present invention relates to automatic optics inspection technical field, more particularly to a kind of element mistake part detection method and system.
Background technology
AOI (Automatic Optic Inspection, automatic optics inspection), is circuit board to be welded using optical principle The equipment detected by the common deficiency occurred in producing of delivering a child.For the circuit board of plug-in unit, common defects detection includes Missing part detection, wrong part detection, the detection of anti-part, more than one piece detection etc..Wherein, wrong part detection refers to the feature for extracting element to be detected, And be compared with template, it is whether correct so as to judge the element for inserting circuit board.
At present, the wrong part detection of element is mainly by manually being detected, but, this detection mode is less efficient, and, Testing result easily malfunctions, and detection accuracy is relatively low.
The content of the invention
Based on this, it is necessary to for the problem that prior art detection efficiency is low, accuracy is low, there is provided a kind of element mistake part inspection Survey method and system.
A kind of element mistake part detection method, comprises the following steps：
Element under test original image on circuit boards is obtained, the spy of the element under test is positioned from the original image Levy area image；Wherein, characteristic information of the feature regional images comprising the element under test, it is right that the characteristic information is used for The element under test is made a distinction with other elements；
By the pixel value of each pixel in the feature regional images respectively with the characteristic area template image for prestoring in The pixel value of corresponding pixel points is compared, and obtains the pixel phase of the feature regional images and the characteristic area template image Like degree；
If the pixel similarity is less than default similarity threshold, the element under test mistake part is judged.
A kind of element mistake part detecting system, including：
Locating module, for obtaining element under test original image on circuit boards, positions institute from the original image State the feature regional images of element under test；Wherein, characteristic information of the feature regional images comprising the element under test, described Characteristic information is for making a distinction to the element under test and other elements；
Comparison module, for by the pixel value of each pixel in the feature regional images respectively with the characteristic area for prestoring In the template image of domain, the pixel value of corresponding pixel points is compared, and obtains the feature regional images and the characteristic area template The pixel similarity of image；
Judge module, if being less than default similarity threshold for the pixel similarity, judges that the element under test is wrong Part.
Said elements mistake part detection method and system, by detecting the characteristic area of element under test, work as feature regional images When less with the pixel similarity of characteristic area template image, judge element under test mistake part, realize element mistake part detection from Dynamicization, can effectively improve detection efficiency and accuracy.
Description of the drawings
Element mistake part detection method flow charts of the Fig. 1 for one embodiment；
Character zone images of the Fig. 2 for Jing contours extracts；
Fig. 3 is the character zone image after correction；
Structural representations of the Fig. 4 for the element mistake part detecting system of one embodiment.
Specific embodiment
Below in conjunction with the accompanying drawings technical scheme is illustrated.
Element mistake part detection method flow charts of the Fig. 1 for one embodiment.As shown in figure 1, the element mistake part detection side Method may include following steps：
S1, obtains element under test original image on circuit boards, positions the element under test from the original image Feature regional images；Wherein, characteristic information of the feature regional images comprising the element under test, the characteristic information are used In making a distinction to the element under test and other elements；
Feature of the present invention can include the word in the body region of the color, shape and element under test of element under test Symbol information etc. is easy to the information for making element under test distinguish with other elements.Below by it is described be characterized as character information as a example by carry out Explanation.The character information can be word, symbol, pattern etc..
Before detection, the original image of the element under test can be oriented first from the image of whole circuit board, then The feature regional images of the element under test are oriented from the original image.When having multiple element under tests on one piece of circuit board When being required for carrying out wrong part detection, the original image of each element under test can be obtained respectively, then respectively from each original image In orient the feature regional images of each element under test.Can be according to each element under test position on circuit boards to each The corresponding feature regional images of element under test carry out sequential storage, in order to the execution of subsequent detection operation.A reality wherein Apply in example, can also be each feature regional images serial number, in order to the execution of subsequent detection operation.
S2, by the pixel value of each pixel in the feature regional images respectively with the characteristic area template image for prestoring The pixel value of middle corresponding pixel points is compared, and obtains the pixel of the feature regional images and the characteristic area template image Similarity；
In this step, in order to prevent characteristic area too little, it is unfavorable for subsequent operation, before being compared, can be with Feature regional images are amplified with process.The enhanced processing refers to the amplification of size, will the length and width of image amplify respectively To original n times, n can be arranged according to actual demand, ordinary circumstance n=2.
In order to eliminate impact of the factors such as the stain on circuit board and background color and pattern to testing result, can be right The feature regional images carry out noise reduction process.Specifically, gray proces can be carried out to the feature regional images, obtains ash Degree image, and binary conversion treatment is carried out to the gray level image according to default pixel threshold.Shown gray proces can basis Equation below is carried out：
Gray=0.299*R+0.587*G+0.114*B；
In formula, R, G and B are respectively three color components of RGB color, and Gray is the gray value after binaryzation.Institute State binaryzation and the pixel that gray value is more than default gray threshold can be set to a certain gray value, pixel value is less than or is waited Another gray value is set in the pixel of default gray threshold.Wherein, the gray threshold can make following object function The maximum gray value of functional value：
G (t)=ω0*(μ0-μ)2+ω1*(μ1-μ)2；
Wherein, μ=ω0*μ0+ω1*μ1；
In formula, ω0For ratio of the corresponding pixel of the characteristic information in the feature regional images, μ0For described The average of the pixel value of the corresponding pixel of characteristic information, ω1For background image pixel in the feature regional images Ratio, μ1For the average of the pixel value of the pixel of background image.In this way, can be by the gray value of gray level image Be divided into that two grey value differences between part, and two parts are maximum, the gray difference between each part it is minimum.
Although the noise reduction process of image can reduce the interference of noise, the interference of noise can not be completely eliminated. Under normal circumstances, noise is comparatively fine in the picture, discrete, and character zone is more continuous.So in order to more accurate fixed Position character zone, can carry out Morphological scale-space to the image of binaryzation, character zone is connected together to form a panel region, and Contours extract is carried out, contour area the best part is character area.Character zone image such as Fig. 2 institutes of Jing contours extracts Show.
As character area has the rotation of certain angle, therefore can be to correct image.Specifically, to Jing shapes After the feature regional images of state process carry out contours extract, minimum matrix fitting can also be carried out to the contour images, Obtain fitted figure picture；Obtain the coordinate value on three summits in the fitted figure picture；According to the coordinate value and the original seat for prestoring Scale value calculates the transfer matrix rotated to the contour images；According to the transfer matrix to the picture on the contour images Vegetarian refreshments carries out coordinate transform；The contour images of Jing coordinate transforms are set to into feature regional images.
Wherein, the transfer matrix can be designated as：
Coordinate transform can be carried out to the point on the feature regional images according to equation below：
In formula, the coordinate value of (x, y) for the pixel on the image of Jing contours extracts before coordinate transform, (x', y') is seat The coordinate value of the pixel after mark conversion on the image of Jing contours extracts.Assume the seat in the rectangle upper left corner, the lower left corner and the upper right corner Mark is respectively (x1,y1)、(x2,y2)、(x3,y3), it is corrected after rectangle coordinate for (0,0), (0, h), (w, 0), wherein w With h represent correction after rectangle it is wide and high, and in order to ensure correct after rectangle it is close with original image or identical, should meet：
Thus, if original image transforms to six elements of the affine transformation matrix M of the image after correction.Word after correction Symbol area image is as shown in Figure 3.
When relatively, the similarity can be calculated according to equation below：
In formula, (x, y) represents the coordinate value of the pixel in the feature regional images, and (x+d, y+d) represents the spy Levy in region template image with coordinate value in the feature regional images for the corresponding pixel of the pixel of (x, y) coordinate Value, I (x, y) represent pixel value of the coordinate value for the pixel of (x, y) in the feature regional images, and M (x+d, y+d) represents institute Pixel value of the coordinate value for the pixel of (x+d, y+d) in characteristic area template image is stated, C represents the similarity.
If storing multiple feature regional images in step S1, this step can respectively by each feature regional images with it is right The characteristic area template image answered carries out the comparison of pixel value.In one embodiment, each characteristic area template image can be with It is stored sequentially in advance locally, phase storage order could be arranged to or phase identical with the storage order of each feature regional images Should.Or, can number for each characteristic area template image, its numbering could be arranged to deposit with each feature regional images Storage order is identical or corresponding.By way of sequential storage feature regional images and/or characteristic area template image, can be in order to Concurrently multiple element under tests are compared, so as to improve element mistake part detection efficiency.
S3, if the pixel similarity is less than default similarity threshold, judges the element under test mistake part.
In this step, if the feature regional images of element under test are less than with the pixel similarity of characteristic area reference picture Default similarity threshold, then show that the feature regional images of element under test are differed with the character zone of characteristic area reference picture It is larger, such that it is able to judge the element under test mistake part；If conversely, the feature regional images of element under test are referred to characteristic area The pixel similarity of image is more than or equal to default similarity threshold, then show feature regional images and the feature of element under test The character zone of area reference image is more similar, such that it is able to judge the not wrong part of the element under test.
The similarity threshold can be according to actual conditions sets itself, and in general, the value of the similarity threshold is got over Greatly, detection accuracy is higher.
The element mistake part detection method of the present invention realizes the automation of element mistake part detection, can effectively improve detection effect Rate and accuracy.Especially in color and/or the more similar shape facility of element, by extracting the character on element under test Information, and be compared with formwork element, whether similar can determine well.
With said elements mistake part detection method accordingly, the present invention also provides a kind of element mistake part detecting system.Such as Fig. 2 Shown, the element mistake part detecting system may include：
Locating module 10, for obtaining element under test original image on circuit boards, positions from the original image The feature regional images of the element under test；Wherein, characteristic information of the feature regional images comprising the element under test, institute Characteristic information is stated for making a distinction to the element under test and other elements；
Feature of the present invention can include the word in the body region of the color, shape and element under test of element under test Symbol information etc. is easy to the information for making element under test distinguish with other elements.Below by it is described be characterized as character information as a example by carry out Explanation.The character information can be word, symbol, pattern etc..
Before detection, the original image of the element under test can be oriented first from the image of whole circuit board, then The feature regional images of the element under test are oriented from the original image.When having multiple element under tests on one piece of circuit board When being required for carrying out wrong part detection, the original image of each element under test can be obtained respectively, then respectively from each original image In orient the feature regional images of each element under test.Can be according to each element under test position on circuit boards to each The corresponding feature regional images of element under test carry out sequential storage, in order to the execution of subsequent detection operation.A reality wherein Apply in example, can also be each feature regional images serial number, in order to the execution of subsequent detection operation.
Comparison module 20, for by the pixel value of each pixel in the feature regional images respectively with the feature for prestoring In region template image, the pixel value of corresponding pixel points is compared, and obtains the feature regional images and the characteristic area mould The pixel similarity of plate image；
In order to prevent characteristic area too little, it is unfavorable for subsequent operation, before being compared, can also be to feature regional As being amplified process.The enhanced processing refers to the amplification of size, will the length and width of image be amplified to original n times, n respectively Can be arranged according to actual demand, ordinary circumstance n=2.
In order to eliminate impact of the factors such as the stain on circuit board and background color and pattern to testing result, can be right The feature regional images carry out noise reduction process.Specifically, gray proces can be carried out to the feature regional images, obtains ash Degree image, and binary conversion treatment is carried out to the gray level image according to default pixel threshold.Shown gray proces can basis Equation below is carried out：
Gray=0.299*R+0.587*G+0.114*B；
In formula, R, G and B are respectively three color components of RGB color, and Gray is the gray value after binaryzation.Institute State binaryzation and the pixel that gray value is more than default gray threshold can be set to a certain gray value, pixel value is less than or is waited Another gray value is set in the pixel of default gray threshold.Wherein, the gray threshold can make following object function The maximum gray value of functional value：
G (t)=ω0*(μ0-μ)2+ω1*(μ1-μ)2；
Wherein, μ=ω0*μ0+ω1*μ1；
In formula, ω0For ratio of the corresponding pixel of the characteristic information in the feature regional images, μ0For described The average of the pixel value of the corresponding pixel of characteristic information, ω1For background image pixel in the feature regional images Ratio, μ1For the average of the pixel value of the pixel of background image.In this way, can be by the gray value of gray level image Be divided into that two grey value differences between part, and two parts are maximum, the gray difference between each part it is minimum.
Although the noise reduction process of image can reduce the interference of noise, the interference of noise can not be completely eliminated. Under normal circumstances, noise is comparatively fine in the picture, discrete, and character zone is more continuous.So in order to more accurate fixed Position character zone, can carry out Morphological scale-space to the image of binaryzation, character zone is connected together to form a panel region, and Contours extract is carried out, contour area the best part is character area.Character zone image such as Fig. 2 institutes of Jing contours extracts Show.
As character area has the rotation of certain angle, therefore can be to correct image.Specifically, to Jing shapes After the feature regional images of state process carry out contours extract, minimum matrix fitting can also be carried out to the contour images, Obtain fitted figure picture；Obtain the coordinate value on three summits in the fitted figure picture；According to the coordinate value and the original seat for prestoring Scale value calculates the transfer matrix rotated to the contour images；According to the transfer matrix to the picture on the contour images Vegetarian refreshments carries out coordinate transform；The contour images of Jing coordinate transforms are set to into feature regional images.
Wherein, the transfer matrix can be designated as：
Coordinate transform can be carried out to the point on the feature regional images according to equation below：
In formula, the coordinate value of (x, y) for the pixel on the image of Jing contours extracts before coordinate transform, (x', y') is seat The coordinate value of the pixel after mark conversion on the image of Jing contours extracts.Assume the seat in the rectangle upper left corner, the lower left corner and the upper right corner Mark is respectively (x1,y1)、(x2,y2)、(x3,y3), it is corrected after rectangle coordinate for (0,0), (0, h), (w, 0), wherein w With h represent correction after rectangle it is wide and high, and in order to ensure correct after rectangle it is close with original image or identical, should meet：
Thus, if original image transforms to six elements of the affine transformation matrix M of the image after correction.Word after correction Symbol area image is as shown in Figure 3.
When relatively, the similarity can be calculated according to equation below：
In formula, (x, y) represents the coordinate value of the pixel in the feature regional images, and (x+d, y+d) represents the spy Levy in region template image with coordinate value in the feature regional images for the corresponding pixel of the pixel of (x, y) coordinate Value, I (x, y) represent pixel value of the coordinate value for the pixel of (x, y) in the feature regional images, and M (x+d, y+d) represents institute Pixel value of the coordinate value for the pixel of (x+d, y+d) in characteristic area template image is stated, C represents the similarity.
If storing multiple feature regional images in locating module 10, comparison module 20 can respectively by each characteristic area Image carries out the comparison of pixel value with corresponding characteristic area template image.In one embodiment, each characteristic area template Image can be stored sequentially in locally in advance, and phase storage order could be arranged to the storage order phase with each feature regional images It is same or corresponding.Or, can number for each characteristic area template image, its numbering could be arranged to and each feature regional The storage order of picture is identical or corresponding.By way of sequential storage feature regional images and/or characteristic area template image, can In order to concurrently be compared to multiple element under tests, so as to improve element mistake part detection efficiency.
Judge module 30, if being less than default similarity threshold for the pixel similarity, judges the element under test Wrong part.
If the feature regional images of element under test are to the pixel similarity of characteristic area reference picture less than default similar Degree threshold value, then show that the feature regional images of element under test differ larger with the character zone of characteristic area reference picture, so as to Can be determined that the element under test mistake part；If conversely, the picture of the feature regional images of element under test and characteristic area reference picture Plain similarity is more than or equal to default similarity threshold, then the feature regional images and characteristic area for showing element under test are referred to The character zone of image is more similar, such that it is able to judge the not wrong part of the element under test.
The similarity threshold can be according to actual conditions sets itself, and in general, the value of the similarity threshold is got over Greatly, detection accuracy is higher.
The element mistake part detecting system of the present invention realizes the automation of element mistake part detection, can effectively improve detection effect Rate and accuracy.Especially in color and/or the more similar shape facility of element, by extracting the character on element under test Information, and be compared with formwork element, whether similar can determine well.
The element mistake part detecting system of the present invention is corresponded with the element mistake part detection method of the present invention, in said elements Technology character and its advantage that the embodiment of wrong part detection method is illustrated are applied to the enforcement of element mistake part detecting system In example, hereby give notice that.
Each technology character of embodiment described above arbitrarily can be combined, for making description succinct, not to above-mentioned reality Apply all possible combination of each technology character in example to be all described, as long as however, the combination of these technology characters is not deposited In contradiction, the scope of this specification record is all considered to be.
Embodiment described above only expresses the several embodiments of the present invention, and its description is more concrete and detailed, but and Therefore can not be construed as limiting the scope of the patent.It should be pointed out that for one of ordinary skill in the art comes Say, without departing from the inventive concept of the premise, some deformations and improvement can also be made, these belong to the protection of the present invention Scope.Therefore, the protection domain of patent of the present invention should be defined by claims.

Claims (10)

1. a kind of element mistake part detection method, it is characterised in that comprise the following steps：
Element under test original image on circuit boards is obtained, the characteristic area of the element under test is positioned from the original image Area image；Wherein, characteristic information of the feature regional images comprising the element under test, the characteristic information is for described Element under test is made a distinction with other elements；
The pixel value of each pixel in the feature regional images is corresponding with the characteristic area template image for prestoring respectively The pixel value of pixel is compared, and obtains the feature regional images similar to the pixel of the characteristic area template image Degree；
If the pixel similarity is less than default similarity threshold, the element under test mistake part is judged.
2. element according to claim 1 mistake part detection method, it is characterised in that will be each in the feature regional images Before the pixel value of individual pixel is compared with the pixel value of corresponding pixel points in the characteristic area template image for prestoring respectively, It is further comprising the steps of：
Process is amplified to the feature regional images.
3. element according to claim 1 mistake part detection method, it is characterised in that position from the original image described The step of feature regional images of element under test, includes：
Morphological scale-space is carried out to the feature regional images；
Contours extract is carried out to the feature regional images of Jing Morphological scale-spaces, contour images are obtained；
Contour area the best part in the contour images is set to into feature regional images.
4. element according to claim 3 mistake part detection method, it is characterised in that shape is carried out to the feature regional images The step of state process, includes：
Gray proces are carried out to the feature regional images, gray level image is obtained；
Binary conversion treatment is carried out to the gray level image according to default pixel threshold；
Morphological scale-space is carried out to the gray level image of binaryzation.
5. element according to claim 4 mistake part detection method, it is characterised in that the default gray threshold for make with The maximum gray value of the functional value of lower object function：
G (t)=ω0*(μ0-μ)2+ω1*(μ1-μ)2；
Wherein, μ=ω0*μ0+ω1*μ1；
In formula, ω0For ratio of the corresponding pixel of the characteristic information in the feature regional images, μ0For the feature The average of the pixel value of the corresponding pixel of information, ω1For the ratio of the pixel in the feature regional images of background image Example, μ1For the average of the pixel value of the pixel of background image.
6. element according to claim 3 mistake part detection method, it is characterised in that in the characteristic area to Jing Morphological scale-spaces It is after area image carries out contours extract, further comprising the steps of：
Minimum matrix fitting is carried out to the contour images, fitted figure picture is obtained；
Obtain the coordinate value on three summits in the fitted figure picture；
The transfer matrix rotated to the contour images is calculated according to the coordinate value and the original coordinates value for prestoring；
Coordinate transform is carried out to the pixel on the contour images according to the transfer matrix；
The contour images of Jing coordinate transforms are set to into feature regional images.
7. element according to claim 6 mistake part detection method, it is characterised in that according to the transfer matrix to the wheel The step of pixel on wide image carries out coordinate transform includes：
Coordinate transform is carried out to the point on the feature regional images according to equation below：
x ′ y ′ = M * ( x , y , 1 ) T ;
In formula, M is transfer matrix, and (x, y) is coordinate value of the pixel before coordinate transform on the image of Jing contours extracts, (x', Y' it is) coordinate value of the pixel on the image of Jing contours extracts after coordinate transform.
8. element according to claim 1 mistake part detection method, it is characterised in that obtain the feature regional images and institute The step of pixel similarity for stating characteristic area template image, includes：
The similarity is calculated according to equation below：
C = Σ x Σ y | I ( x , y ) - M ( x + d , y + d ) |
In formula, (x, y) represents the coordinate value of the pixel in the feature regional images, and (x+d, y+d) represents the characteristic area The coordinate value of the pixel corresponding with the pixel that coordinate value in the feature regional images is (x, y), I in the template image of domain (x, y) represents pixel value of the coordinate value for the pixel of (x, y) in the feature regional images, and M (x+d, y+d) represents the spy Pixel value of the coordinate value for the pixel of (x+d, y+d) in region template image is levied, C represents the similarity.
9. element according to claim 1 mistake part detection method, it is characterised in that the characteristic information is character information.
10. a kind of element mistake part detecting system, it is characterised in that include：
Locating module, for obtaining element under test original image on circuit boards, treats described in positioning from the original image Survey the feature regional images of element；Wherein, characteristic information of the feature regional images comprising the element under test, the feature Information is for making a distinction to the element under test and other elements；
Comparison module, for by the pixel value of each pixel in the feature regional images respectively with the characteristic area mould for prestoring In plate image, the pixel value of corresponding pixel points is compared, and obtains the feature regional images and the characteristic area template image Pixel similarity；
Judge module, if being less than default similarity threshold for the pixel similarity, judges the element under test mistake part.
CN201610887939.3A 2016-10-11 2016-10-11 Element mistake part detection method and system Active CN106530275B (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
CN201610887939.3A CN106530275B (en) 2016-10-11 2016-10-11 Element mistake part detection method and system
PCT/CN2016/113146 WO2018068415A1 (en) 2016-10-11 2016-12-29 Detection method and system for wrong part

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201610887939.3A CN106530275B (en) 2016-10-11 2016-10-11 Element mistake part detection method and system

Publications (2)

Publication Number Publication Date
CN106530275A true CN106530275A (en) 2017-03-22
CN106530275B CN106530275B (en) 2019-06-11

Family

ID=58331466

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201610887939.3A Active CN106530275B (en) 2016-10-11 2016-10-11 Element mistake part detection method and system

Country Status (2)

Country Link
CN (1) CN106530275B (en)
WO (1) WO2018068415A1 (en)

Cited By (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107330879A (en) * 2017-06-21 2017-11-07 广州视源电子科技股份有限公司 A kind of inspection method of device, device, equipment and storage medium
WO2018068417A1 (en) * 2016-10-11 2018-04-19 广州视源电子科技股份有限公司 Component defect detection method and system
CN109870117A (en) * 2017-12-05 2019-06-11 英业达科技有限公司 Binaryzation outline detection system and its method
CN110544640A (en) * 2018-05-28 2019-12-06 长鑫存储技术有限公司 method and apparatus for inspecting semiconductor substrate
CN111209944A (en) * 2019-12-31 2020-05-29 上海索广电子有限公司 Self-adaptive image detection system and image detection method
CN111738247A (en) * 2020-05-15 2020-10-02 上海望友信息科技有限公司 Identification method and identification device of polarity identification, electronic equipment and storage medium
CN112750116A (en) * 2021-01-15 2021-05-04 北京市商汤科技开发有限公司 Defect detection method and device, computer equipment and storage medium
CN112801987A (en) * 2021-02-01 2021-05-14 上海悦易网络信息技术有限公司 Mobile phone part abnormity detection method and equipment
CN115131287A (en) * 2022-05-16 2022-09-30 奥蒂玛光学科技（深圳）有限公司 Image processing method, area detection method, detection device and storage medium
CN115564779A (en) * 2022-12-08 2023-01-03 东莞先知大数据有限公司 Part defect detection method and device and storage medium
CN116152251A (en) * 2023-04-20 2023-05-23 成都数之联科技股份有限公司 Television backboard detection method, model training method, device, equipment and medium
CN117152145A (en) * 2023-10-31 2023-12-01 威海天拓合创电子工程有限公司 Board card process detection method and device based on image
WO2024044913A1 (en) * 2022-08-29 2024-03-07 Siemens Aktiengesellschaft Method, apparatus, electronic device, storage medium and computer program product for detecting circuit board assembly defect

Families Citing this family (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109614879B (en) * 2018-11-19 2022-12-02 温州大学 Hopper particle detection method based on image recognition
CN109685781B (en) * 2018-12-17 2022-11-29 江苏蜂奥生物科技有限公司 Multi-target rapid identification method based on certain rules and applied to propolis soft capsules
CN112308814A (en) * 2019-07-26 2021-02-02 北京四方继保自动化股份有限公司 Method and system for automatically identifying switch on-off position state of disconnecting link of power system
CN111553376B (en) * 2019-12-24 2023-05-23 西安元智系统技术有限责任公司 Cultural relic contour monitoring method
CN111462035B (en) * 2020-01-21 2024-03-08 北京明略软件系统有限公司 Picture detection method and device
CN111476782B (en) * 2020-04-10 2023-04-07 南通大学 Automobile rear camera extension line terminal image detection system
CN111539933B (en) * 2020-04-22 2023-06-06 大连日佳电子有限公司 Direct-insert element detection method and system
CN113034488B (en) * 2021-04-13 2024-04-19 荣旗工业科技(苏州)股份有限公司 Visual inspection method for ink-jet printed matter
CN114332069B (en) * 2022-01-05 2024-02-20 合肥工业大学 Connector detection method and device based on machine vision
CN114441499B (en) * 2022-04-11 2022-07-12 天津美腾科技股份有限公司 Grade detection method and device, identification equipment, ore pulp grade instrument and storage medium
CN115482234B (en) * 2022-10-08 2023-09-15 浙江花园药业有限公司 High-precision defect detection method and system for aluminum-plastic blister medicines
CN115797361B (en) * 2023-02-13 2023-06-27 山东淼泰建筑科技有限公司 Aluminum template surface defect detection method
CN116205919B (en) * 2023-05-05 2023-06-30 深圳市智宇精密五金塑胶有限公司 Hardware part production quality detection method and system based on artificial intelligence

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102222328A (en) * 2011-07-01 2011-10-19 杭州电子科技大学 Edge-preserving self-adaptive weighted filtering method for natural scene images
CN103810478A (en) * 2014-02-21 2014-05-21 广东小天才科技有限公司 Sitting posture detection method and device
CN104363449A (en) * 2014-10-31 2015-02-18 华为技术有限公司 Method and relevant device for predicting pictures
CN104931907A (en) * 2015-05-13 2015-09-23 中国矿业大学 Digital display electrical measurement instrument quality group-detection system based on robot visual sense
CN105957059A (en) * 2016-04-20 2016-09-21 广州视源电子科技股份有限公司 Electronic component missing detection method and system

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101055620B (en) * 2006-04-12 2011-04-06 富士通株式会社 Shape comparison device and method
CN102831381B (en) * 2011-06-15 2016-05-04 罗普特（厦门）科技集团有限公司 image difference comparison system and method
CN104463178A (en) * 2014-12-29 2015-03-25 广州视源电子科技股份有限公司 Electronic component recognizing method and system

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102222328A (en) * 2011-07-01 2011-10-19 杭州电子科技大学 Edge-preserving self-adaptive weighted filtering method for natural scene images
CN103810478A (en) * 2014-02-21 2014-05-21 广东小天才科技有限公司 Sitting posture detection method and device
CN104363449A (en) * 2014-10-31 2015-02-18 华为技术有限公司 Method and relevant device for predicting pictures
CN104931907A (en) * 2015-05-13 2015-09-23 中国矿业大学 Digital display electrical measurement instrument quality group-detection system based on robot visual sense
CN105957059A (en) * 2016-04-20 2016-09-21 广州视源电子科技股份有限公司 Electronic component missing detection method and system

Cited By (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2018068417A1 (en) * 2016-10-11 2018-04-19 广州视源电子科技股份有限公司 Component defect detection method and system
CN107330879A (en) * 2017-06-21 2017-11-07 广州视源电子科技股份有限公司 A kind of inspection method of device, device, equipment and storage medium
CN109870117A (en) * 2017-12-05 2019-06-11 英业达科技有限公司 Binaryzation outline detection system and its method
CN110544640A (en) * 2018-05-28 2019-12-06 长鑫存储技术有限公司 method and apparatus for inspecting semiconductor substrate
CN111209944B (en) * 2019-12-31 2023-09-01 上海索广映像有限公司 Self-adaptive image detection system and image detection method
CN111209944A (en) * 2019-12-31 2020-05-29 上海索广电子有限公司 Self-adaptive image detection system and image detection method
CN111738247A (en) * 2020-05-15 2020-10-02 上海望友信息科技有限公司 Identification method and identification device of polarity identification, electronic equipment and storage medium
CN112750116A (en) * 2021-01-15 2021-05-04 北京市商汤科技开发有限公司 Defect detection method and device, computer equipment and storage medium
CN112750116B (en) * 2021-01-15 2023-08-11 北京市商汤科技开发有限公司 Defect detection method, device, computer equipment and storage medium
CN112801987A (en) * 2021-02-01 2021-05-14 上海悦易网络信息技术有限公司 Mobile phone part abnormity detection method and equipment
CN115131287A (en) * 2022-05-16 2022-09-30 奥蒂玛光学科技（深圳）有限公司 Image processing method, area detection method, detection device and storage medium
WO2024044913A1 (en) * 2022-08-29 2024-03-07 Siemens Aktiengesellschaft Method, apparatus, electronic device, storage medium and computer program product for detecting circuit board assembly defect
CN115564779A (en) * 2022-12-08 2023-01-03 东莞先知大数据有限公司 Part defect detection method and device and storage medium
CN116152251A (en) * 2023-04-20 2023-05-23 成都数之联科技股份有限公司 Television backboard detection method, model training method, device, equipment and medium
CN117152145A (en) * 2023-10-31 2023-12-01 威海天拓合创电子工程有限公司 Board card process detection method and device based on image
CN117152145B (en) * 2023-10-31 2024-02-23 威海天拓合创电子工程有限公司 Board card process detection method and device based on image

Also Published As

Publication number Publication date
CN106530275B (en) 2019-06-11
WO2018068415A1 (en) 2018-04-19

Similar Documents

Publication Publication Date Title
CN106530275A (en) Element wrong part detection method and system
CN111243032B (en) Full-automatic detection method for checkerboard corner points
EP1870835B1 (en) Two-dimensional code
US7031511B2 (en) Hole inspection apparatus and method
CN108596166A (en) A kind of container number identification method based on convolutional neural networks classification
Wang et al. Robust license plate recognition based on dynamic projection warping
CN105976354B (en) Element localization method and system based on color and gradient
WO2018068417A1 (en) Component defect detection method and system
JP2001133418A (en) Method and apparatus for defect detection based on shape feature
CN111259891B (en) Method, device, equipment and medium for identifying identity card in natural scene
CN104331695A (en) Robust round identifier shape quality detection method
CN108709500B (en) Circuit board element positioning and matching method
CN103150725B (en) Based on SUSAN edge detection method and the system of non-local mean
CN1898555B (en) Substrate inspection device
CN110288040A (en) A kind of similar evaluation method of image based on validating topology and equipment
CN108182707B (en) Chessboard grid calibration template under incomplete collection condition and automatic identification method thereof
CN111814778A (en) Text line region positioning method, layout analysis method and character recognition method
JPH05281154A (en) Inspection apparatus for defect of pattern
JPH07249119A (en) Method for recognizing picture and device therefor
CN114187294B (en) Regular wafer positioning method based on prior information
CN111798429B (en) Visual inspection method for defects of printed matter
JP2676990B2 (en) Wiring pattern inspection equipment
JP2964594B2 (en) Mark inspection method
JP2805476B2 (en) Character recognition method
JPH0254375A (en) Method and device for checking pattern defect

Legal Events

Date Code Title Description
C06 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