CN107248151A - A kind of LCD panel intelligent detecting method and system based on machine vision - Google Patents
A kind of LCD panel intelligent detecting method and system based on machine vision Download PDFInfo
- Publication number
- CN107248151A CN107248151A CN201710260493.6A CN201710260493A CN107248151A CN 107248151 A CN107248151 A CN 107248151A CN 201710260493 A CN201710260493 A CN 201710260493A CN 107248151 A CN107248151 A CN 107248151A
- Authority
- CN
- China
- Prior art keywords
- lcd panel
- image
- detected
- standard
- binary
- 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
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
- G01—MEASURING; TESTING
- G01D—MEASURING NOT SPECIALLY ADAPTED FOR A SPECIFIC VARIABLE; ARRANGEMENTS FOR MEASURING TWO OR MORE VARIABLES NOT COVERED IN A SINGLE OTHER SUBCLASS; TARIFF METERING APPARATUS; MEASURING OR TESTING NOT OTHERWISE PROVIDED FOR
- G01D21/00—Measuring or testing not otherwise provided for
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/30—Determination of transform parameters for the alignment of images, i.e. image registration
-
- 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
- G06T2207/30148—Semiconductor; IC; Wafer
Abstract
The invention discloses a kind of LCD panel intelligent detecting method based on machine vision and system, method includes obtaining the image of LCD panel and LCD panel to be detected under clamping device undamped state and image in the clamp state of standard respectively, and both images are subjected to registration by projection algorithm, both images are finally made poor, the different zones of both images are drawn, so as to carry out detection judgement to it.System includes standard picture acquiring unit, standard picture binarization unit, image to be detected acquiring unit, image to be detected binarization unit, image registration unit and detection judging unit.The present invention is by carrying out the contrast of pixel to LCD panel binary image to be detected and standard LCD panel binaryzation template image so as to detect the quality of LCD panel, effectively improve the detection efficiency in LCD panel detection process, reduction missing inspection false drop rate simultaneously reduces cost of labor, improves productive temp.It the composite can be widely applied in LCD panel detection.
Description
Technical field
The present invention relates to LCD panel detection technique field, more particularly to a kind of LCD panel Intelligent Measurement based on machine vision
Method and system.
Background technology
Machine vision is a complex art, including image procossing, mechanical engineering technology, control, electric source lighting, optics
Imaging, sensor, simulation and digital video technology, computer hardware technique (image enhaucament and parser, image card, I/O
Card etc.).One typical machine vision applications system includes picture catching, light-source system, image digitazation module, digital picture
Processing module, intelligent decision decision-making module and Mechanical course performing module.
The characteristics of NI Vision Builder for Automated Inspection is the flexibility and automaticity for improving production.It is not suitable for manual work at some
Dangerous work environment or artificial vision be difficult to meet desired occasion, machine in normal service vision substitutes artificial vision；Exist simultaneously
In high-volume industrial processes, manually visual inspection product quality efficiency is low and precision is not high, uses Machine Vision Detection side
Method can greatly improve the automaticity of production efficiency and production.It is to realize and machine vision is easily achieved information integration
The basic technology of computer integrated manufacturing system.
A kind of detection technique that the LCD panel intelligent testing technology modern times are needed badly, in traditional detection method, LCD panel
Quality testing relies primarily on workman's human eye detection, but is due to the subjectivity and visual fatigue of workman itself, allows for this
Detection method has examination criteria to be influenceed by factor and individual subjective factor, and missing inspection false drop rate is high, the shortcomings of cost of labor is high.
The content of the invention
In order to solve the above-mentioned technical problem, detection efficiency can be improved it is an object of the invention to provide one kind, and leakage can be reduced
Examine a kind of the LCD panel intelligent detecting method and system based on machine vision of false drop rate.
The technical solution used in the present invention is：
A kind of LCD panel intelligent detecting method based on machine vision, comprises the following steps：
Image of the acquisition standard LCD panel under clamping device undamped state and image in the clamp state, are marked
Quasi- LCD panel does not clamp image and titer chip clamps image；
Do not clamp image to standard LCD panel and titer chip clamps image and handled, obtain standard LCD panel two-value
Change template image；
Image of the LCD panel to be detected under clamping device undamped state and image in the clamp state are obtained, is obtained
LCD panel to be detected does not clamp image and LCD panel to be detected clamps image；
Do not clamp image to LCD panel to be detected and LCD panel to be detected clamps image and handled, obtain liquid crystal to be detected
Piece binary image；
LCD panel binary image to be detected is matched somebody with somebody with standard LCD panel binaryzation template image by projection algorithm
It is accurate；
LCD panel binary image to be detected and standard LCD panel binaryzation template image are made poor, both images are drawn
Difference, and detection judgement is carried out to it.
It is described to mark as a kind of further improvement of described LCD panel intelligent detecting method based on machine vision
Quasi- LCD panel does not clamp image and titer chip clamps image and handled, and obtains standard LCD panel binaryzation template image,
The step for specifically include：
Do not clamp image to standard LCD panel and titer chip clamps image and carries out gaussian filtering；
Two images obtained after gaussian filtering process are carried out at binaryzation by compartmentalization automatic threshold segmentation algorithm
Reason, the standard LCD panel of obtaining does not clamp binary image and titer chip clamps binary image；
Standard LCD panel is not clamped into binary image and titer chip clamps binary image and makees poor, titer is obtained
Chip binaryzation template image.
As a kind of further improvement of described LCD panel intelligent detecting method based on machine vision, described treats
Detection LCD panel does not clamp image and LCD panel to be detected clamps image and handled, and obtains LCD panel binary picture to be detected
Picture, the step for specifically include：
Do not clamp image to LCD panel to be detected and LCD panel to be detected clamps image and carries out gaussian filtering；
Two images obtained after gaussian filtering process are carried out at binaryzation by compartmentalization automatic threshold segmentation algorithm
Reason, obtains LCD panel to be detected and does not clamp binary image and LCD panel to be detected clamping binary image；
LCD panel to be detected is not clamped into binary image and LCD panel to be detected clamps binary image and makees poor, is treated
Detect LCD panel binary image.
As a kind of further improvement of described LCD panel intelligent detecting method based on machine vision, described will treat
Detect LCD panel binary image it is registering by projection algorithm progress with standard LCD panel binaryzation template image, the step for have
Body includes：
Gaussian pyramid is set up respectively to LCD panel binary image to be detected and standard LCD panel binaryzation template image,
Obtain LCD panel multilayer pyramid diagram picture to be detected and standard LCD panel multilayer pyramid diagram picture；
To LCD panel multilayer pyramid diagram picture to be detected and standard LCD panel multilayer pyramid diagram picture by projection algorithm from
Top to bottm successively carries out registration.
As a kind of further improvement of described LCD panel intelligent detecting method based on machine vision, described will treat
Detect that LCD panel binary image makees poor with standard LCD panel binaryzation template image, draw the difference of both images, and to it
Carry out detection judgement, the step for specifically include：
LCD panel binary image to be detected and standard LCD panel binaryzation template image are made poor, show that difference is counted
Amount, and shown；
Judge whether discrepancy quantity is more than default detection threshold value, if, then it represents that current LCD panel to be detected is not
It is qualified；Otherwise, it means that LCD panel to be detected is qualified.
It is used as a kind of further improvement of described LCD panel intelligent detecting method based on machine vision, described region
The calculation formula of optimal threshold is in change automatic threshold segmentation algorithm：
Wherein, t represents the threshold value of segmentation, w0For background ratio, u0For background mean value, w1For prospect ratio, u1It is equal for prospect
Value, u is the average of entire image, when above transition formula evaluation maximum t, as segmentation figure picture optimal threshold.
Another technical scheme of the present invention is：
A kind of LCD panel intelligent checking system based on machine vision, including：
Standard picture acquiring unit, for obtaining image of the standard LCD panel under clamping device undamped state and in folder
Image under tight state, the standard LCD panel of obtaining does not clamp image and titer chip clamps image；
Standard picture binarization unit, clamps image and carries out for not clamping image and titer chip to standard LCD panel
Processing, obtains standard LCD panel binaryzation template image；
Image to be detected acquiring unit, for obtain image of the LCD panel to be detected under clamping device undamped state and
Image in the clamp state, obtains LCD panel to be detected and does not clamp image and LCD panel to be detected clamping image；
Image to be detected binarization unit, for not clamping image and LCD panel clamping figure to be detected to LCD panel to be detected
As being handled, LCD panel binary image to be detected is obtained；
Image registration unit, for LCD panel binary image to be detected to be led to standard LCD panel binaryzation template image
Cross projection algorithm and carry out registration；
Judging unit is detected, for LCD panel binary image to be detected and standard LCD panel binaryzation template image to be made
Difference, draws the difference of both images, and carry out detection judgement to it.
It is used as a kind of further improvement of described LCD panel intelligent checking system based on machine vision, described standard
Image binaryzation unit includes：
Filter unit, clamps image and carries out gaussian filtering for not clamping image and titer chip to standard LCD panel；
Automatic threshold unit, compartmentalization automatic threshold segmentation is passed through for two images to being obtained after gaussian filtering process
Algorithm carries out binary conversion treatment, and the standard LCD panel of obtaining does not clamp binary image and titer chip clamps binary image；
Make poor unit, for standard LCD panel not to be clamped into binary image and titer chip clamping binary image work
Difference, obtains standard LCD panel binaryzation template image.
It is used as a kind of further improvement of described LCD panel intelligent checking system based on machine vision, described image
Registration unit includes：
Pyramid sets up unit, for LCD panel binary image to be detected and standard LCD panel binaryzation template image
Gaussian pyramid is set up respectively, obtains LCD panel multilayer pyramid diagram picture to be detected and standard LCD panel multilayer pyramid diagram picture；
Registration unit is projected, for LCD panel multilayer pyramid diagram picture to be detected and standard LCD panel multilayer pyramid diagram
As successively carrying out registration from top to bottom by projection algorithm.
It is used as a kind of further improvement of described LCD panel intelligent checking system based on machine vision, described detection
Judging unit includes：
Discrepancy computing unit, for by LCD panel binary image to be detected and standard LCD panel binaryzation template image
It is poor to make, and draws discrepancy quantity, and shown；
Diversity judgement unit, for judging whether discrepancy quantity is more than default detection threshold value, if, then it represents that it is current
LCD panel to be detected it is unqualified；Otherwise, it means that LCD panel to be detected is qualified.
The beneficial effects of the invention are as follows：
The present invention a kind of LCD panel intelligent detecting method and system based on machine vision pass through to LCD panel two to be detected
The contrast that value image carries out pixel with standard LCD panel binaryzation template image effectively carries so as to detect the quality of LCD panel
Detection efficiency in high LCD panel detection process, reduces missing inspection false drop rate and reduces cost of labor, improve productive temp.
Brief description of the drawings
The embodiment to the present invention is described further below in conjunction with the accompanying drawings：
Fig. 1 is a kind of step flow chart of the LCD panel intelligent detecting method based on machine vision of the present invention；
Fig. 2 is a kind of LCD panel intelligent detecting method Plays LCD panel image binaryzation based on machine vision of the present invention
Step flow chart；
Fig. 3 is liquid crystal picture two-value to be detected in a kind of LCD panel intelligent detecting method based on machine vision of the present invention
The step flow chart of change；
Fig. 4 is registering step flow chart in a kind of LCD panel intelligent detecting method based on machine vision of the present invention；
Fig. 5 is the step flow that judgement is detected in a kind of LCD panel intelligent detecting method based on machine vision of the present invention
Figure；
Fig. 6 is a kind of block diagram of the LCD panel intelligent checking system based on machine vision of the present invention.
Embodiment
With reference to Fig. 1, a kind of LCD panel intelligent detecting method based on machine vision of the present invention comprises the following steps：
Image of the acquisition standard LCD panel under clamping device undamped state and image in the clamp state, are marked
Quasi- LCD panel does not clamp image and titer chip clamps image；
Do not clamp image to standard LCD panel and titer chip clamps image and handled, obtain standard LCD panel two-value
Change template image；
Image of the LCD panel to be detected under clamping device undamped state and image in the clamp state are obtained, is obtained
LCD panel to be detected does not clamp image and LCD panel to be detected clamps image；
Do not clamp image to LCD panel to be detected and LCD panel to be detected clamps image and handled, obtain liquid crystal to be detected
Piece binary image；
LCD panel binary image to be detected is matched somebody with somebody with standard LCD panel binaryzation template image by projection algorithm
It is accurate；
LCD panel binary image to be detected and standard LCD panel binaryzation template image are made poor, both images are drawn
Difference, and detection judgement is carried out to it.
With reference to Fig. 2, it is further used as preferred embodiment, it is described that image and titer are not clamped to standard LCD panel
Chip clamps image and handled, and obtains standard LCD panel binaryzation template image, the step for specifically include：
Do not clamp image to standard LCD panel and titer chip clamps image and carries out gaussian filtering；
Two images obtained after gaussian filtering process are carried out at binaryzation by compartmentalization automatic threshold segmentation algorithm
Reason, the standard LCD panel of obtaining does not clamp binary image and titer chip clamps binary image；
Standard LCD panel is not clamped into binary image and titer chip clamps binary image and makees poor, titer is obtained
Chip binaryzation template image.
With reference to Fig. 3, it is further used as preferred embodiment, it is described not clamp image to LCD panel to be detected and to be checked
Survey LCD panel and clamp image and handled, obtain LCD panel binary image to be detected, the step for specifically include：
Do not clamp image to LCD panel to be detected and LCD panel to be detected clamps image and carries out gaussian filtering；
Two images obtained after gaussian filtering process are carried out at binaryzation by compartmentalization automatic threshold segmentation algorithm
Reason, obtains LCD panel to be detected and does not clamp binary image and LCD panel to be detected clamping binary image；
LCD panel to be detected is not clamped into binary image and LCD panel to be detected clamps binary image and makees poor, is treated
Detect LCD panel binary image.
With reference to Fig. 4, it is further used as preferred embodiment, it is described by LCD panel binary image to be detected and standard
LCD panel binaryzation template image by projection algorithm carry out registration, the step for specifically include：
Gaussian pyramid is set up respectively to LCD panel binary image to be detected and standard LCD panel binaryzation template image,
Obtain LCD panel multilayer pyramid diagram picture to be detected and standard LCD panel multilayer pyramid diagram picture；
To LCD panel multilayer pyramid diagram picture to be detected and standard LCD panel multilayer pyramid diagram picture by projection algorithm from
Top to bottm successively carries out registration.
With reference to Fig. 5, it is further used as preferred embodiment, it is described by LCD panel binary image to be detected and standard
It is poor that LCD panel binaryzation template image is made, and draws the difference of both images, and carries out detection judgement to it, the step for specifically wrap
Include：
LCD panel binary image to be detected and standard LCD panel binaryzation template image are made poor, show that difference is counted
Amount, and shown；
Judge whether discrepancy quantity is more than default detection threshold value, if, then it represents that current LCD panel to be detected is not
It is qualified；Otherwise, it means that LCD panel to be detected is qualified.
It is further used as preferred embodiment, the calculating of optimal threshold in described compartmentalization automatic threshold segmentation algorithm
Formula is：
Wherein, t represents the threshold value of segmentation, w0For background ratio, u0For background mean value, w1For prospect ratio, u1It is equal for prospect
Value, u is the average of entire image, when above transition formula evaluation maximum t, as segmentation figure picture optimal threshold.
With reference to Fig. 6, a kind of LCD panel intelligent checking system based on machine vision of the invention, including：
Standard picture acquiring unit, for obtaining image of the standard LCD panel under clamping device undamped state and in folder
Image under tight state, the standard LCD panel of obtaining does not clamp image and titer chip clamps image；
Standard picture binarization unit, clamps image and carries out for not clamping image and titer chip to standard LCD panel
Processing, obtains standard LCD panel binaryzation template image；
Image to be detected acquiring unit, for obtain image of the LCD panel to be detected under clamping device undamped state and
Image in the clamp state, obtains LCD panel to be detected and does not clamp image and LCD panel to be detected clamping image；
Image to be detected binarization unit, for not clamping image and LCD panel clamping figure to be detected to LCD panel to be detected
As being handled, LCD panel binary image to be detected is obtained；
Image registration unit, for LCD panel binary image to be detected to be led to standard LCD panel binaryzation template image
Cross projection algorithm and carry out registration；
Judging unit is detected, for LCD panel binary image to be detected and standard LCD panel binaryzation template image to be made
Difference, draws the difference of both images, and carry out detection judgement to it.
It is further used as preferred embodiment, described standard picture binarization unit includes：
Filter unit, clamps image and carries out gaussian filtering for not clamping image and titer chip to standard LCD panel；
Automatic threshold unit, compartmentalization automatic threshold segmentation is passed through for two images to being obtained after gaussian filtering process
Algorithm carries out binary conversion treatment, and the standard LCD panel of obtaining does not clamp binary image and titer chip clamps binary image；
Make poor unit, for standard LCD panel not to be clamped into binary image and titer chip clamping binary image work
Difference, obtains standard LCD panel binaryzation template image.
It is further used as preferred embodiment, described image registration unit includes：
Pyramid sets up unit, for LCD panel binary image to be detected and standard LCD panel binaryzation template image
Gaussian pyramid is set up respectively, obtains LCD panel multilayer pyramid diagram picture to be detected and standard LCD panel multilayer pyramid diagram picture；
Registration unit is projected, for LCD panel multilayer pyramid diagram picture to be detected and standard LCD panel multilayer pyramid diagram
As successively carrying out registration from top to bottom by projection algorithm.
It is further used as preferred embodiment, described detection judging unit includes：
Discrepancy computing unit, for by LCD panel binary image to be detected and standard LCD panel binaryzation template image
It is poor to make, and draws discrepancy quantity, and shown；
Diversity judgement unit, for judging whether discrepancy quantity is more than default detection threshold value, if, then it represents that it is current
LCD panel to be detected it is unqualified；Otherwise, it means that LCD panel to be detected is qualified.
In the specific embodiment of the invention, build by detecting box, man-machine interaction display, camera underframe, 1# cameras, 2# phases
The detection platform that machine, camera frame, 1# liquid crystal plate clamp and 2# liquid crystal plate clamp are constituted.Camera and liquid crystal plate clamp are individually fixed in
On platform, and relative position is kept to fix, the fixed camera underframe on detecting box, camera support；Fixer wafer chuck makes
Two fixtures are obtained to align with detecting box center line；In camera underframe end stationary computer man-machine interaction display；In camera branch
1# cameras and 2# cameras are installed on frame, camera and computer communication connecting line and fixture gas circuit, adjustment camera position and Jiao are installed
Away from causing 1# camera perspectives that 1# fixtures are completely covered, LCD panel detection electricity below 2# fixtures, fixture is completely covered in 2# camera perspectives
Lu Shang electricity, and make it that image is most clear, then fixed camera position.
S1, a piece of titer wafer product is taken, be put into 1# fixtures, pneumatic circuit closes holding jig undamped state, drives
Dynamic 1# cameras obtain this state hypograph, and the standard that obtains does not clamp image, then open pneumatic circuit, and 1# fixtures keep clamping shape
State, while detecting power on circuitry, 1# drives camera during LCD panel display image, obtains standard and clamps image.
S2：Obtained two images are subjected to gaussian filtering, then the progress pair of deployment area automatic threshold segmentation algorithm
Picture carries out binary conversion treatment and obtains binaryzation picture, finally obtains standard LCD panel two-value as difference to the two images after processing
Change template image；
To adapt to the requirement of image procossing, the noise being mixed into during image digitazation is eliminated.In the image gathered
First have to carry out gaussian filtering to image, two-dimensional Gaussian function can be expressed as：
Wherein μ is peak value (peak value correspondence position), and σ represents standard deviation, and (variable x and variable y respectively have an average, also respectively have
One standard deviation)；
, it is necessary to carry out binary conversion treatment to image, concrete methods of realizing is automatic threshold segmentation after image filtering is handled
Algorithm, because in industrial actual application, the image collected may be because of light source irradiation inequality, external environmental interference
The automatic threshold segmentation algorithm in brightness disproportionation, the present invention is caused to be incited somebody to action according to the difference of brightness of image etc. the influence of factor
The image collected is automatically divided into N number of region, then asks each region respectively in the algorithm using automatic threshold segmentation
Threshold value, and then realize the binaryzation of image；
If gray level image gray level is L, then tonal range is [0, L-1], and image is calculated using automatic threshold segmentation algorithm
Optimal threshold be：
Wherein, t represents the threshold value of segmentation, w0For background ratio, u0For background mean value, w1For prospect ratio, u1It is equal for prospect
Value, u is the average of entire image, when above transition formula evaluation maximum t, as segmentation figure picture optimal threshold.
S3：Liquid crystal flake products to be detected are taken, 1# fixtures are put into, pneumatic circuit, which is closed, keeps 1# fixture undamped states, drives
Dynamic 1# cameras acquisition is to be detected not to clamp image, pneumatic circuit is then opened, 1# fixtures keep clamped condition, while detecting circuit
It is powered, 1# drives camera during LCD panel display image, obtains clamping image to be detected；Carry out identical process be applied to 2# fixtures and
2# cameras.Acquired image digitazation is stored in computer, binary conversion treatment is passed through by computer, has obtained to be detected
LCD panel binary image, the process is identical with step S3 process.
S4：LCD panel binary image to be detected is carried out with standard LCD panel binaryzation template image by projection algorithm
Registration；
S41：Set up Gauss gold respectively to LCD panel binary image to be detected and standard LCD panel binaryzation template image
Word tower, obtains LCD panel multilayer pyramid diagram picture to be detected and standard LCD panel multilayer pyramid diagram picture, obtains low resolution figure
Picture.In order to improve the discrimination of information in the present embodiment, using the characteristic of gaussian pyramid model multi-scale expression, to input
The image that picture sets up in the gaussian pyramid model that three layers and smoothing factor are 0.5, gaussian pyramid in different groups time has
Different sizes and resolution ratio, the picture size close to bottom is relatively large, reflects image Small and Medium Sized details；And with layer
Secondary to move up, size and resolution ratio all relative reductions of image so only describe the main information of target in image；
S42：LCD panel multilayer pyramid diagram picture to be detected and standard LCD panel multilayer pyramid diagram picture are calculated by projecting
Method successively carries out registration from top to bottom.It is most rough layer first to searching for the pyramidal the superiors, most rough layer is matched,
The XY gradient maps and difference between the two of mobile image and still image are generated, is then fitted using least square method registration
To optimal varied coefficient.Current search result is recorded, is entered when being scanned for next tomographic image centered on this result
Row search, while not stopping to correct the structure of preceding layer Gaussian image, repeats the step to original image i.e. out to out layer.
S5：LCD panel binary image to be detected and standard LCD panel binaryzation template image are made poor, discrepancy is drawn
Quantity, and it is shown in man-machine interaction display interfaces.Rational detection threshold value is preset according to the difference of the required precision of detection, sentenced
Whether offset dissimilarity quantity is more than default detection threshold value, if, then it represents that current LCD panel to be detected is unqualified；Conversely,
Then represent that LCD panel to be detected is qualified.
From the foregoing it can be that the present invention a kind of LCD panel intelligent detecting method and system based on machine vision pass through it is right
LCD panel binary image to be detected carries out the contrast of pixel to detect liquid crystal with standard LCD panel binaryzation template image
The quality of piece, effectively improves the detection efficiency in LCD panel detection process, reduces missing inspection false drop rate and reduces cost of labor, improves
Productive temp.
Above is the preferable implementation to the present invention is illustrated, but the invention is not limited to the implementation
Example, those skilled in the art can also make a variety of equivalent variations or replace on the premise of without prejudice to spirit of the invention
Change, these equivalent deformations or replacement are all contained in the application claim limited range.
Claims (10)
1. a kind of LCD panel intelligent detecting method based on machine vision, it is characterised in that comprise the following steps：
Image of the acquisition standard LCD panel under clamping device undamped state and image in the clamp state, obtain titer
Chip does not clamp image and titer chip clamps image；
Do not clamp image to standard LCD panel and titer chip clamps image and handled, obtain standard LCD panel binaryzation mould
Plate image；
Image of the LCD panel to be detected under clamping device undamped state and image in the clamp state are obtained, obtains to be checked
Survey LCD panel and do not clamp image and LCD panel to be detected clamping image；
Do not clamp image to LCD panel to be detected and LCD panel to be detected clamps image and handled, obtain LCD panel two to be detected
Value image；
LCD panel binary image to be detected is registering by projection algorithm progress with standard LCD panel binaryzation template image；
LCD panel binary image to be detected and standard LCD panel binaryzation template image are made poor, both images are drawn not
Together, and to it detection judgement is carried out.
2. a kind of LCD panel intelligent detecting method based on machine vision according to claim 1, it is characterised in that：It is described
Do not clamped to standard LCD panel image and titer chip clamp image and handle, obtain standard LCD panel binaryzation template
Image, the step for specifically include：
Do not clamp image to standard LCD panel and titer chip clamps image and carries out gaussian filtering；
Two images to being obtained after gaussian filtering process are obtained by compartmentalization automatic threshold segmentation algorithm progress binary conversion treatment
Binary image is not clamped to standard LCD panel and titer chip clamps binary image；
Standard LCD panel is not clamped into binary image and titer chip clamps binary image and makees poor, standard LCD panel is obtained
Binaryzation template image.
3. a kind of LCD panel intelligent detecting method based on machine vision according to claim 1, it is characterised in that：It is described
Do not clamped to LCD panel to be detected image and LCD panel to be detected clamp image and handle, obtain LCD panel two-value to be detected
Change image, the step for specifically include：
Do not clamp image to LCD panel to be detected and LCD panel to be detected clamps image and carries out gaussian filtering；
Two images to being obtained after gaussian filtering process are obtained by compartmentalization automatic threshold segmentation algorithm progress binary conversion treatment
Binary image is not clamped to LCD panel to be detected and LCD panel to be detected clamps binary image；
LCD panel to be detected is not clamped into binary image and LCD panel to be detected clamps binary image and makees poor, obtains to be detected
LCD panel binary image.
4. a kind of LCD panel intelligent detecting method based on machine vision according to claim 1, it is characterised in that：It is described
LCD panel binary image to be detected is registering by projection algorithm progress with standard LCD panel binaryzation template image, this
Step is specifically included：
Gaussian pyramid is set up respectively to LCD panel binary image to be detected and standard LCD panel binaryzation template image, obtained
LCD panel multilayer pyramid diagram picture to be detected and standard LCD panel multilayer pyramid diagram picture；
To LCD panel multilayer pyramid diagram picture to be detected and standard LCD panel multilayer pyramid diagram picture by projection algorithm on to
Under successively carry out registration.
5. a kind of LCD panel intelligent detecting method based on machine vision according to claim 1, it is characterised in that：It is described
LCD panel binary image to be detected and standard LCD panel binaryzation template image are made poor, draw the difference of both images,
And carry out detection judgement to it, the step for specifically include：
LCD panel binary image to be detected and standard LCD panel binaryzation template image are made poor, discrepancy quantity is drawn, and
Shown；
Judge whether discrepancy quantity is more than default detection threshold value, if, then it represents that current LCD panel to be detected is unqualified；
Otherwise, it means that LCD panel to be detected is qualified.
6. a kind of LCD panel intelligent detecting method based on machine vision according to Claims 2 or 3, it is characterised in that：
The calculation formula of optimal threshold is in described compartmentalization automatic threshold segmentation algorithm：
Wherein, t represents the threshold value of segmentation, w0For background ratio, u0For background mean value, w1For prospect ratio, u1For prospect average, u
For the average of entire image.
7. a kind of LCD panel intelligent checking system based on machine vision, it is characterised in that including：
Standard picture acquiring unit, for obtaining image of the standard LCD panel under clamping device undamped state and clamping shape
Image under state, the standard LCD panel of obtaining does not clamp image and titer chip clamps image；
Standard picture binarization unit, is clamped at image for not clamping image and titer chip to standard LCD panel
Reason, obtains standard LCD panel binaryzation template image；
Image to be detected acquiring unit, for obtaining image of the LCD panel to be detected under clamping device undamped state and in folder
Image under tight state, obtains LCD panel to be detected and does not clamp image and LCD panel to be detected clamping image；
Image to be detected binarization unit, clamps image and enters for not clamping image and LCD panel to be detected to LCD panel to be detected
Row processing, obtains LCD panel binary image to be detected；
Image registration unit, for LCD panel binary image to be detected and standard LCD panel binaryzation template image to be passed through into throwing
Shadow algorithm carries out registration；
Judging unit is detected, for LCD panel binary image to be detected and standard LCD panel binaryzation template image to be made into poor,
The difference of both images is drawn, and detection judgement is carried out to it.
8. a kind of LCD panel intelligent checking system based on machine vision according to claim 7, it is characterised in that：It is described
Standard picture binarization unit include：
Filter unit, clamps image and carries out gaussian filtering for not clamping image and titer chip to standard LCD panel；
Automatic threshold unit, compartmentalization automatic threshold segmentation algorithm is passed through for two images to being obtained after gaussian filtering process
Binary conversion treatment is carried out, the standard LCD panel of obtaining does not clamp binary image and titer chip clamps binary image；
Make poor unit, for standard LCD panel not being clamped into binary image and titer chip clamps binary image and makees poor,
Obtain standard LCD panel binaryzation template image.
9. a kind of LCD panel intelligent checking system based on machine vision according to claim 7, it is characterised in that：It is described
Image registration unit include：
Pyramid sets up unit, for distinguishing LCD panel binary image to be detected and standard LCD panel binaryzation template image
Gaussian pyramid is set up, LCD panel multilayer pyramid diagram picture to be detected and standard LCD panel multilayer pyramid diagram picture is obtained；
Registration unit is projected, for logical to LCD panel multilayer pyramid diagram picture to be detected and standard LCD panel multilayer pyramid diagram picture
Cross projection algorithm and successively carry out registration from top to bottom.
10. a kind of LCD panel intelligent checking system based on machine vision according to claim 7, it is characterised in that：Institute
The detection judging unit stated includes：
Discrepancy computing unit, for LCD panel binary image to be detected and standard LCD panel binaryzation template image to be made
Difference, draws discrepancy quantity, and shown；
Diversity judgement unit, for judging whether discrepancy quantity is more than default detection threshold value, if, then it represents that current treats
Detect that LCD panel is unqualified；Otherwise, it means that LCD panel to be detected is qualified.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710260493.6A CN107248151B (en) | 2017-04-20 | 2017-04-20 | Intelligent liquid crystal display detection method and system based on machine vision |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710260493.6A CN107248151B (en) | 2017-04-20 | 2017-04-20 | Intelligent liquid crystal display detection method and system based on machine vision |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107248151A true CN107248151A (en) | 2017-10-13 |
| CN107248151B CN107248151B (en) | 2020-12-22 |
Family
ID=60016405
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710260493.6A Active CN107248151B (en) | 2017-04-20 | 2017-04-20 | Intelligent liquid crystal display detection method and system based on machine vision |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107248151B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108960169A (en) * | 2018-07-12 | 2018-12-07 | 杭州电子科技大学 | Instrument and equipment state on_line monitoring method and system based on computer vision |
| CN113552136A (en) * | 2021-07-30 | 2021-10-26 | 广州中国科学院先进技术研究所 | High-temperature forging visual detection system with vibration isolation capability |
| CN113570605A (en) * | 2021-09-28 | 2021-10-29 | 深圳市绘晶科技有限公司 | Defect detection method and system based on liquid crystal display panel |
Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5038048A (en) * | 1988-12-23 | 1991-08-06 | Hitachi, Ltd. | Defect detection system and method for pattern to be inspected utilizing multiple-focus image signals |
| CN1678036A (en) * | 2004-03-31 | 2005-10-05 | 株式会社岛津制作所 | Radiographic apparatus, and radiation detection signal processing method |
| CN1801433A (en) * | 2006-01-11 | 2006-07-12 | 彩虹集团电子股份有限公司 | Plasma display screen fault checking method |
| JP2010071892A (en) * | 2008-09-19 | 2010-04-02 | Toray Ind Inc | Pattern inspection method and pattern inspection device |
| CN102053093A (en) * | 2010-11-08 | 2011-05-11 | 北京大学深圳研究生院 | Method for detecting surface defects of chip cut from wafer surface |
| CN102095733A (en) * | 2011-03-02 | 2011-06-15 | 上海大学 | Machine vision-based intelligent detection method for surface defect of bottle cap |
| CN102509300A (en) * | 2011-11-18 | 2012-06-20 | 深圳市宝捷信科技有限公司 | Defect detection method and system |
| CN103439348A (en) * | 2013-08-16 | 2013-12-11 | 中国科学院半导体研究所 | Remote controller key defect detection method based on difference image method |
| CN104360501A (en) * | 2014-10-15 | 2015-02-18 | 西安交通大学 | Visual detection method and device for defects of liquid crystal display screen |
| CN104792794A (en) * | 2015-04-28 | 2015-07-22 | 武汉工程大学 | Machine vision based optical film surface defect detecting method |
| CN105069790A (en) * | 2015-08-06 | 2015-11-18 | 潍坊学院 | Rapid imaging detection method for gear appearance defect |
| CN105976389A (en) * | 2016-05-20 | 2016-09-28 | 南京理工大学 | Mobile phone baseboard connector defect detection method |
| CN106169431A (en) * | 2016-07-06 | 2016-11-30 | 江苏维普光电科技有限公司 | Mask plate based on GPU and wafer defect detection method |
| CN106204614A (en) * | 2016-07-21 | 2016-12-07 | 湘潭大学 | A kind of workpiece appearance defects detection method based on machine vision |
-
2017
- 2017-04-20 CN CN201710260493.6A patent/CN107248151B/en active Active
Patent Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5038048A (en) * | 1988-12-23 | 1991-08-06 | Hitachi, Ltd. | Defect detection system and method for pattern to be inspected utilizing multiple-focus image signals |
| CN1678036A (en) * | 2004-03-31 | 2005-10-05 | 株式会社岛津制作所 | Radiographic apparatus, and radiation detection signal processing method |
| CN1801433A (en) * | 2006-01-11 | 2006-07-12 | 彩虹集团电子股份有限公司 | Plasma display screen fault checking method |
| JP2010071892A (en) * | 2008-09-19 | 2010-04-02 | Toray Ind Inc | Pattern inspection method and pattern inspection device |
| CN102053093A (en) * | 2010-11-08 | 2011-05-11 | 北京大学深圳研究生院 | Method for detecting surface defects of chip cut from wafer surface |
| CN102095733A (en) * | 2011-03-02 | 2011-06-15 | 上海大学 | Machine vision-based intelligent detection method for surface defect of bottle cap |
| CN102509300A (en) * | 2011-11-18 | 2012-06-20 | 深圳市宝捷信科技有限公司 | Defect detection method and system |
| CN103439348A (en) * | 2013-08-16 | 2013-12-11 | 中国科学院半导体研究所 | Remote controller key defect detection method based on difference image method |
| CN104360501A (en) * | 2014-10-15 | 2015-02-18 | 西安交通大学 | Visual detection method and device for defects of liquid crystal display screen |
| CN104792794A (en) * | 2015-04-28 | 2015-07-22 | 武汉工程大学 | Machine vision based optical film surface defect detecting method |
| CN105069790A (en) * | 2015-08-06 | 2015-11-18 | 潍坊学院 | Rapid imaging detection method for gear appearance defect |
| CN105976389A (en) * | 2016-05-20 | 2016-09-28 | 南京理工大学 | Mobile phone baseboard connector defect detection method |
| CN106169431A (en) * | 2016-07-06 | 2016-11-30 | 江苏维普光电科技有限公司 | Mask plate based on GPU and wafer defect detection method |
| CN106204614A (en) * | 2016-07-21 | 2016-12-07 | 湘潭大学 | A kind of workpiece appearance defects detection method based on machine vision |
Non-Patent Citations (4)
| Title |
|---|
| LEI WANG: "《Design of Machine Vision Applications in Detection of Defects in High-Speed Bar Copper》", 《2010 INTERNATIONAL CONFERENCE ON E-PRODUCT E-SERVICE AND E-ENTERTAINMENT》 * |
| VISHWANATH A. SINDAGI等: "《OLED panel defect detection using local inlier-outlier ratios and modified LBP》", 《2015 14TH IAPR INTERNATIONAL CONFERENCE ON MACHINE VISION APPLICATIONS (MVA)》 * |
| 叶亭等: "《一种基于线阵CCD技术印刷电路板胶片的尺寸及缺陷在线检测方法》", 《光学与光电技术》 * |
| 周江等: "《基于机器视觉的磁钢片缺陷检测研究》", 《机电工程》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108960169A (en) * | 2018-07-12 | 2018-12-07 | 杭州电子科技大学 | Instrument and equipment state on_line monitoring method and system based on computer vision |
| CN113552136A (en) * | 2021-07-30 | 2021-10-26 | 广州中国科学院先进技术研究所 | High-temperature forging visual detection system with vibration isolation capability |
| CN113570605A (en) * | 2021-09-28 | 2021-10-29 | 深圳市绘晶科技有限公司 | Defect detection method and system based on liquid crystal display panel |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107248151B (en) | 2020-12-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110659660B (en) | Automatic optical detection classification equipment using deep learning system and training equipment thereof | |
| CN104749184B (en) | Automatic optical detection method and system | |
| CN109754442B (en) | Gear pitting detection system based on machine vision | |
| CN105067639B (en) | The eyeglass defect automatic detection device and method of a kind of Grating Modulation | |
| CN110216080A (en) | A kind of Quality Monitoring Control System of the PCB processing producing line based on image comparison | |
| CN106248686A (en) | Glass surface defects based on machine vision detection device and method | |
| CN111612737B (en) | Artificial board surface flaw detection device and detection method | |
| CN107966454A (en) | A kind of end plug defect detecting device and detection method based on FPGA | |
| CN109840900A (en) | A kind of line detection system for failure and detection method applied to intelligence manufacture workshop | |
| CN107884413A (en) | A kind of device and detection method of automatic detection bearing roller missing | |
| CN110403232A (en) | A kind of cigarette quality detection method based on second level algorithm | |
| CN110108712A (en) | Multifunctional visual sense defect detecting system | |
| CN107248151A (en) | A kind of LCD panel intelligent detecting method and system based on machine vision | |
| CN111583114A (en) | Automatic measuring device and measuring method for pipeline threads | |
| CN107664644A (en) | A kind of apparent automatic detection device of object based on machine vision and method | |
| CN109358067A (en) | Motor ring varistor defect detecting system based on computer vision and method | |
| CN109374632A (en) | A kind of detection method and system of display panel | |
| CN110956627A (en) | Intelligent optical detection sample characteristic and flaw intelligent lighting image capturing method and device | |
| WO2023134286A1 (en) | Online automatic quality testing and classification method for cathode copper | |
| CN107782744A (en) | A kind of eyeglass defect automatic detection device of Grating Modulation | |
| Fu et al. | Medicine glass bottle defect detection based on machine vision | |
| CN207081666U (en) | A kind of zipper detecting device based on machine vision | |
| CN105391998B (en) | Automatic detection method and apparatus for resolution of low-light night vision device | |
| CN110412055A (en) | A kind of lens white haze defect inspection method based on multiple light courcess dark-ground illumination | |
| CN109387524A (en) | Thread defect detection method and device based on linearly polarized photon |
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 |