CN110111711A - The detection method and device of screen, computer readable storage medium - Google Patents
The detection method and device of screen, computer readable storage medium Download PDFInfo
- Publication number
- CN110111711A CN110111711A CN201910364098.1A CN201910364098A CN110111711A CN 110111711 A CN110111711 A CN 110111711A CN 201910364098 A CN201910364098 A CN 201910364098A CN 110111711 A CN110111711 A CN 110111711A
- Authority
- CN
- China
- Prior art keywords
- image
- screen
- pixel
- sub
- detected
- 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.)
- Pending
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
- G09—EDUCATION; CRYPTOGRAPHY; DISPLAY; ADVERTISING; SEALS
- G09G—ARRANGEMENTS OR CIRCUITS FOR CONTROL OF INDICATING DEVICES USING STATIC MEANS TO PRESENT VARIABLE INFORMATION
- G09G3/00—Control arrangements or circuits, of interest only in connection with visual indicators other than cathode-ray tubes
- G09G3/006—Electronic inspection or testing of displays and display drivers, e.g. of LED or LCD displays
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10004—Still image; Photographic image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20112—Image segmentation details
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20212—Image combination
- G06T2207/20221—Image fusion; Image merging
-
- 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
- G06T2207/30121—CRT, LCD or plasma display
Abstract
The present invention provides a kind of detection method of screen, device and computer readable storage mediums, are related to field of display technology, the accuracy of the detection process for improving sub-pixel defect point.Wherein the detection method of the screen includes: the image for obtaining screen to be detected；Gabor filtering is carried out to the image of acquired screen to be detected respectively by multiple Gabor filters, obtains multiple gabor filtering images, wherein the gabor filtering parameter of multiple Gabor filters is different；Image co-registration is carried out to multiple gabor filtering images, obtains image co-registration result；According to segmentation threshold range detection image co-registration as a result, obtaining defective sub-pixel in screen to be detected, wherein segmentation threshold range is the intensity value ranges of image slices vegetarian refreshments corresponding to flawless sub-pixel.The detection method of the screen is in the defects detection of screen.
Description
Technical field
The present invention relates to field of display technology more particularly to a kind of detection method and device of screen, computer-readable deposit
Storage media.
Background technique
Display screen may have a defective sub-pixels of some displays after the completion of preparation, brightness it is obvious it is darker even
The sub-pixel not worked is dim spot, and obviously brighter sub-pixel is bright spot for brightness.It therefore, can be to aobvious after the completion of display screen is in preparation
Display screen is made whether the detection of existing defects, and positions the position of defective sub-pixel, in order to it is subsequent to it is above-mentioned have it is scarce
Sunken sub-pixel such as is manually checked and is analyzed at the operation.
Summary of the invention
The embodiment of the present invention provides a kind of detection method and device of screen, computer readable storage medium, to improve
The accuracy in detection of defective sub-pixel in screen.
In order to achieve the above objectives, the embodiment of the present invention adopts the following technical scheme that
In a first aspect, the embodiment provides a kind of detection methods of screen, comprising: obtain screen to be detected
Image；Gabor filtering is carried out to the image of acquired screen to be detected respectively by multiple Gabor filters, obtains multiple gals
Cypress filtering image；Wherein, the gabor filtering parameter of the multiple Gabor filter is different；
Image co-registration is carried out to the multiple gabor filtering image, obtains image co-registration result；According to segmentation threshold range
Described image fusion results are detected, defective sub-pixel in the screen to be detected is obtained；Wherein, the segmentation threshold range
For the intensity value ranges of image slices vegetarian refreshments corresponding to flawless sub-pixel.
In the detection method of above-mentioned screen, gabor filtering is carried out by the image to screen to be detected, and obtain multiple gals
Cypress filtering image separates the textural characteristics of the image of the screen to be detected, and is extracted respectively to multiple gabor filtering figures
As in.Since Gabor filter has preferable effect, the detection of above-mentioned screen in the separation and extraction of textural characteristics
Method is preferable to the separation of the textural characteristics of the image of screen to be detected and extraction effect, to make according to above-mentioned gabor filtering figure
After carrying out image co-registration, obtained fusion results are more accurate, the screen to be detected obtained according to the image co-registration result
Each sub-pixel corresponding to image slices vegetarian refreshments gray value it is also more accurate, and then according to the gray value to screen to be detected
The testing result for carrying out defect point detection is more accurate, namely improves the accuracy in detection of defective sub-pixel.
Based on the above-mentioned technical proposal, in some embodiments, the gabor filtering parameter includes deflection and wavelength；It is described
Gabor filtering parameter difference includes that deflection is different and/or wavelength is different.
In some embodiments, the multiple Gabor filter includes 12 Gabor filters；The multiple gabor filtering
The deflection of device includes 0 °, 45 °, 90 ° and 135 °；The wavelength of the multiple Gabor filter includes N, 2N and 3N；Wherein, N table
Show the sub-pixel column region of the screen to be detected described in the image of the screen to be detected perpendicular to the sub-pixel column area
The number of shared image slices vegetarian refreshments on the extending direction in domain.
In some embodiments, described that image co-registration is carried out to the multiple gabor filtering image, obtain image co-registration knot
The step of fruit, comprising: use following formula (1), to the feature vector f of the multiple gabor filtering imagei,j(x, y) is returned
One changes operation, obtains normalized feature vector pi,j(x, y):
Wherein, i, j are integer, and 1≤i≤m, 1≤j≤n, m indicate the value of the wavelength of the multiple Gabor filter
Number, n indicate the value number of the deflection of the multiple Gabor filter, and minF (x, y) indicates the screen to be detected
Image image slices vegetarian refreshments (x, y) gray value F (x, y) minimum value, maxF (x, y) indicates the figure of the screen to be detected
The maximum value of the gray value F (x, y) of the image slices vegetarian refreshments (x, y) of picture；Using following formula (2), (3), to described normalized
Feature vector pij(x, y) carries out the first fusion operation, obtains the first fusion results Li(x, y):
Bi,j(x, y)=pi,j(x,y)pi,j+1(x,y)+pi,j(x,y)pi,j+2(x,y)+…+pi,j(x,y)pi,n(x,y) (3)；
Using following formula (4), (5), to the first fusion results Li(x, y) carries out the second fusion operation, obtains institute
State image co-registration result N (x, y):
In some embodiments, described according to segmentation threshold range detection described image fusion results, it obtains described to be checked
The step of surveying defective sub-pixel in screen, comprising:
According to described image fusion results, image slices vegetarian refreshments corresponding to each sub-pixel in the screen to be detected is obtained
Intensity value ranges；According to the intensity value ranges of image slices vegetarian refreshments corresponding to the segmentation threshold range and sub-pixel, respectively
Judge each sub-pixel for flawless sub-pixel, dim spot or bright spot: if the segmentation threshold range covers sub-pixel
The intensity value ranges of corresponding image slices vegetarian refreshments, or be overlapped with the intensity value ranges of image slices vegetarian refreshments corresponding to sub-pixel,
Then determine that the sub-pixel is flawless sub-pixel；If the upper limit value of the intensity value ranges of image slices vegetarian refreshments corresponding to sub-pixel
Greater than the upper limit value of the segmentation threshold range, then determine the sub-pixel for dim spot；If image slices vegetarian refreshments corresponding to sub-pixel
Intensity value ranges lower limit value be less than the segmentation threshold range lower limit value, then determine the sub-pixel for bright spot.
In some embodiments, described according to segmentation threshold range detection described image fusion results, obtain it is described to
Before the step of detecting the defective sub-pixel of screen, further include the steps that obtaining the segmentation threshold range；The acquisition
The step of segmentation threshold range, comprising: obtain the image of zero defect screen；By the multiple Gabor filter to described
The image of zero defect screen carries out multiple gabor filtering, obtains multiple gabor filtering reference pictures；To the multiple gabor filtering
Reference picture carries out image co-registration, obtains with reference to blending image and reference picture fusion results；It is merged according to the reference picture
As a result, determining the maximum gradation value and minimum gradation value of image slices vegetarian refreshments corresponding to the sub-pixel of the zero defect screen, obtain
To the segmentation threshold range.
In some embodiments, described according to the reference picture fusion results, determine the sub- picture of the zero defect screen
Before the step of maximum gradation value and minimum gradation value of image slices vegetarian refreshments corresponding to element, further includes: to described with reference to fusion
Image pixel point in the periphery predeterminable area of image is removed processing, and according to removing, treated obtains with reference to blending image
The reference picture fusion results.
In some embodiments, described by multiple Gabor filters respectively to the image of acquired screen to be detected
Before the step of carrying out gabor filtering, obtaining multiple gabor filtering images, further includes: carried out to the image of the screen to be detected
Noise reduction and/or image enhancement processing.
Second aspect, the embodiment of the invention provides a kind of detection devices of screen, comprising: image grabber is configured to
Obtain the image of screen to be detected；The gabor filtering parameter of multiple Gabor filters, the multiple Gabor filter is different, described
Multiple Gabor filters are configured to carry out gabor filtering to the image of the screen to be detected respectively, obtain multiple gabor filtering figures
Picture；Processor is configured to carry out image co-registration to the multiple gabor filtering image, obtains image co-registration result；Detector is matched
It is set to according to segmentation threshold range, detects described image fusion results, obtain defective sub-pixel in the screen to be detected；
Wherein, the segmentation threshold range is the intensity value ranges of image slices vegetarian refreshments corresponding to flawless sub-pixel.
The detection method of screen provided by the beneficial effect and first aspect that the detection device of above-mentioned screen can be generated
Beneficial effect it is identical, details are not described herein again.
The third aspect, the embodiment of the invention provides a kind of computer readable storage medium, the computer-readable storage
Media storage has one or more instructions, one or more steps in the detection method to execute screen as described in relation to the first aspect
Suddenly.
The detection of screen provided by the beneficial effect and first aspect that above-mentioned computer readable storage medium can be generated
The beneficial effect of method is identical, and details are not described herein again.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
Some embodiments of invention for those of ordinary skill in the art without creative efforts, can be with
Other attached drawings are obtained according to these attached drawings.
Fig. 1 is the flow diagram according to a kind of detection method of screen of some embodiments of the invention；
Fig. 2 is the schematic diagram according to a kind of image of screen to be detected of some embodiments of the invention；
Fig. 3 is the flow diagram according to a kind of detection method of screen of some embodiments of the invention；
Fig. 4 is the flow diagram according to a kind of detection method of screen of some embodiments of the invention.
Specific embodiment
To keep goal of the invention of the invention, feature and advantage more obvious and easy to understand, below in conjunction with attached drawing, to this hair
Bright technical solution is clearly and completely described.Obviously, described embodiment is only that a part of the invention is implemented
Example, instead of all the embodiments.Based on the embodiments of the present invention, those of ordinary skill in the art are obtained all other
Embodiment belongs to the scope of protection of the invention.
Referring to Fig. 1, some embodiments of the present invention provide a kind of detection method of screen, comprising the following steps:
S1. the image of screen to be detected is obtained.
It should be pointed out that the sub-pixel needs of screen to be detected are clearly shown in the image of screen to be detected, so as to
The subsequent separation and extraction that textural characteristics are carried out to the arrangement of above-mentioned sub-pixel.Illustratively, as above-mentioned acquisition screen to be detected
A kind of achievable mode of the image of curtain, can use industrial camera and shoots to above-mentioned screen to be detected, above-mentioned to obtain
The image of screen to be detected.The image of the screen to be detected can be black white image.
S2. gabor filtering is carried out to the image of acquired screen to be detected respectively by multiple Gabor filters, obtained
Multiple gabor filtering images.Wherein, the gabor filtering parameter of multiple Gabor filters is different.
S3. image co-registration is carried out to multiple gabor filtering images, obtains image co-registration result.
S4. according to segmentation threshold range detection image co-registration as a result, obtaining defective sub-pixel in screen to be detected.Its
In, segmentation threshold range is the intensity value ranges of image slices vegetarian refreshments corresponding to flawless sub-pixel.
Illustratively, in the step S4 of the detection method of above-mentioned screen, it is to be detected that flawless sub-pixel can be this
Flawless sub-pixel in screen, the sub-pixel being also possible in other zero defect screens.
It should be noted that the token state of the brightness of each sub-pixel of screen to be detected is brightness value, by each sub- picture
When in element shooting to the image of screen to be detected, the brightness of each sub-pixel can be with the gray value of corresponding image slices vegetarian refreshments come table
Sign.For example, if the brightness value of some sub-pixel of screen to be detected is higher, on the image of screen to be detected with the sub-pixel pair
The gray value for the image slices vegetarian refreshments answered is lower；If the brightness value of some sub-pixel of screen to be detected is lower, screen to be detected
Image on image slices vegetarian refreshments corresponding with the sub-pixel gray value it is higher.It may also be noted that " image slices vegetarian refreshments " refers to
Be for forming the pixel of the image on image, and " sub-pixel " refers to the sub- picture for display in screen to be detected
Element.
In the detection method of above-mentioned screen, gabor filtering is carried out by the image to screen to be detected, and obtain multiple
Gabor filtering image separates the textural characteristics of the image of the screen to be detected, and separated textural characteristics are distinguished
It extracts into multiple gabor filtering images.It, can be by separated line after carrying out image co-registration to multiple gabor filtering images
Reason Fusion Features and the textural characteristics for obtaining the more completely image of screen to be detected, to realize the image to screen to be detected
Textural characteristics extraction.Then according to image co-registration result each sub-pixel of screen to be detected can be obtained corresponding to image slices
The gray value of vegetarian refreshments, and by detecting to obtain defective sub-pixel compared with segmentation threshold range.Due to Gabor filter
There is preferable effect in the separation and extraction of textural characteristics, therefore in step s 2 to the texture of the image of screen to be detected
The separation of feature and extraction effect are preferable, to make each sub-pixel institute according to the obtained screen to be detected of image co-registration result
The gray value of corresponding image slices vegetarian refreshments is more accurate, therefore, the carry out defect point inspection according to the gray value to screen to be detected
The testing result of survey is more accurate, namely improves the accuracy in detection of defective sub-pixel.
In addition, the detection method of above-mentioned screen needs not move through during carrying out defect point detection to screen to be detected
It is a large amount of to calculate, therefore the detection efficiency of the detection method of the screen is higher, can be measured in real time to screen to be detected.
Based on the above-mentioned technical proposal, in some embodiments of the invention, as a kind of achievable mode, in above-mentioned steps
In S2, gabor filtering parameter may include deflection and wavelength, gabor filtering parameter difference may include deflection it is different and/or
Wavelength is different.
By setting different directions angle and/or the Gabor filter of wavelength, more can accurately realize to screen to be detected
The separation and extraction of the textural characteristics of the image of curtain, to improve the accuracy in detection of the sub-pixel of defective point.
Illustratively, above-mentioned multiple Gabor filters may include 12 Gabor filters, 12 Gabor filters
Deflection includes 0 °, 45 °, 90 ° and 135 °, and wavelength includes N, 2N and 3N.
Referring to fig. 2, wherein N indicates that in the image 100 of screen to be detected, the sub-pixel column region 1 of screen to be detected exists
The number of image slices vegetarian refreshments shared on perpendicular direction with the extending direction uv in sub-pixel column region 1.That is, Fig. 2
The width d of sub-pixel column region 1 is N times of the width of image slices vegetarian refreshments.It is noted that the width in sub-pixel column region 1
DegreeWherein, d1Indicate that (namely sub-pixel 10 is in prolonging perpendicular to sub-pixel column region 1 for the width of sub-pixel 10
Stretch the size on the uv of direction), d2Indicate the size of space between adjacent two sub-pixels 10.
It should be noted that above-mentioned 12 Gabor filters are by arranging aforementioned four deflection and three wavelength
Column combination obtains, namely a deflection is selected from aforementioned four deflection, while one is selected from above three wavelength
Gabor filtering parameter of the wavelength as a Gabor filter can get altogether the different gabor filtering of 12 gabor filtering parameters
Device:
First Gabor filter: deflection is 0 °, wavelength N；
Second Gabor filter: deflection is 0 °, wavelength 2N；
Third Gabor filter: deflection is 0 °, wavelength 3N；
4th Gabor filter: deflection is 45 °, wavelength N；
5th Gabor filter: deflection is 45 °, wavelength 2N；
6th Gabor filter: deflection is 45 °, wavelength 3N；
7th Gabor filter: deflection is 90 °, wavelength N；
8th Gabor filter: deflection is 90 °, wavelength N；
9th Gabor filter: deflection is 90 °, wavelength 2N；
Tenth Gabor filter: deflection is 135 °, wavelength 3N；
11st Gabor filter: deflection is 135 °, wavelength 3N；
12nd Gabor filter: deflection is 135 °, wavelength 3N.
It should be noted that carrying out the filter of gal cypress to the image of screen to be detected respectively by above-mentioned 12 Gabor filters
Wave, available 12 gabor filtering images.
By the way that the deflection of multiple Gabor filters is set as 0 °, 45 °, 90 ° and 135 °, textural characteristics can be made to separate
With the deflection of extraction is equiangularly spaced is uniformly distributed, further promote separation and extraction effect to textural characteristics.
Illustratively, can make the extending direction of the sub-pixel column or rows in the image of screen to be detected with it is multiple
One of deflection included by Gabor filter is in the same direction, for example, making the extension side of the sub-pixel column in the image of screen to be detected
To being 90 °.Sub-pixel in screen to be detected is arranged in matrix, and the angle between sub-pixel column and rows is in 90 °, this
When, multiple Gabor filters can be along the extending directions (0 ° of direction) of rows, the extending direction (90 ° of directions) of sub-pixel column
Textural characteristics are separated and extracted with two angular bisector directions (45 ° and 135 ° of directions) of the two, are promoted to texture spy
The separation of sign and extraction effect.
In addition, the wavelength of multiple Gabor filters is set as N, 2N and 3N, it can be at multiple wavelengths to above-mentioned texture
Feature is separated, and the extraction of the textural characteristics under multiple wavelength is realized, and is further promoted the separation to textural characteristics and is mentioned
The accuracy taken.
Referring to Fig. 3, in some embodiments of the invention, as a kind of achievable mode, to more in above-mentioned steps S3
A gabor filtering image carries out image co-registration may comprise steps of when obtaining image co-registration result:
S31. following formula (1) is used, to the feature vector f of multiple gabor filtering imagesi,jFortune is normalized in (x, y)
It calculates, obtains normalized feature vector pi,j(x, y):
Wherein, i, j are integer, 1≤i≤m, 1≤j≤n.M indicates the value number of the wavelength of multiple Gabor filters,
N indicates the value number of the deflection of multiple Gabor filters.MinF (x, y) indicates the image pixel of the image of screen to be detected
The minimum value of the gray value F (x, y) of point (x, y), maxF (x, y) indicate the image slices vegetarian refreshments (x, y) of the image of screen to be detected
Gray value F (x, y) maximum value.It should be noted that image slices vegetarian refreshments (x, y) indicates to be detected in above-mentioned formula (1)
The gray value of the xth row of the image of screen, y column image slices vegetarian refreshments.
It should also be noted that, since the value number of the wavelength of multiple Gabor filters is m, multiple Gabor filters
The value number n of deflection can obtain m × n gabor filtering parameter by carrying out permutation and combination to wavelength and deflection altogether
Different Gabor filters.Gabor filtering is carried out to the image of screen to be detected respectively with the m × n Gabor filter, then altogether
Available m × n gabor filtering image, wherein each gabor filtering image corresponds to a filter result, the filter result
Feature vector f can be usedi,j(x, y) is indicated.
Illustratively, the image of screen to be detected includes R image slices vegetarian refreshments (x, y) altogether, then features described above vector fi,j
(x, y) is 1 × R dimensional vector, this feature vector fi,j(x, y) includes the gray value F of R element and R image slices vegetarian refreshments
(x, y) maps one by one.
S32. following formula (2), (3) are used, to normalized feature vector pi,j(x, y) carries out the first fusion operation, obtains
To the first fusion results Li(x, y):
Bi,j(x, y)=pi,j(x,y)pi,j+1(x,y)+pi,j(x,y)pi,j+2(x,y)+…+pi,j(x,y)pi,n(x,y) (3)。
It should be noted that the first fusion operation is the fusion operation of direction dimension in above-mentioned steps S32, that is,
It says, the gabor filtering image that the Gabor filter of Same Wavelength, different directions angle filters is merged, obtain m
First fusion results, are expressed as Li(x,y).The direction of the m the first fusion results is identical.
S33. following formula (4), (5) are used, to the first fusion results Li(x, y) carries out the second fusion operation, obtains figure
As fusion results N (x, y):
It should be noted that the second fusion operation is the fusion operation of wavelength dimension in above-mentioned steps S33, that is,
It says, m the first fusion results of same direction angle, different wave length is merged, image co-registration result N (x, y) is obtained.
Illustratively, below by obtained gabor filtering image after the Gabor filter filtering with 12 different parameters
For fusion, above-mentioned steps S31~S33 is illustrated, wherein the value number of the wavelength of 12 Gabor filters is 3
(i.e. m=3), the value number 4 (i.e. n=4) of the deflection of multiple Gabor filters.
After carrying out 12 filtering to the image of screen to be detected respectively by above-mentioned 12 Gabor filters, it can be obtained 12
Gabor filtering image, 12 gabor filtering images respectively correspond a feature vector, respectively indicate are as follows: f1,1(x,y)、f1,2
(x,y)、f1,3(x,y)、f1,4(x,y)、f2,1(x,y)、f2,2(x,y)、f2,3(x,y)、f2,4(x,y)、f3,1(x,y)、f3,2(x,
y)、f3,3(x, y) and f3,4(x,y).Wherein, in the footmark of above-mentioned 12 feature vectors, first digit represents the volume of wavelength
Number, second digit represents deflection number.
Using above-mentioned formula (1), operation is normalized to above-mentioned 12 feature vectors respectively, 12 is obtained and returns accordingly
One feature vector changed: p1,1(x,y)、p1,2(x,y)、p1,3(x,y)、p1,4(x,y)、p2,1(x,y)、p2,2(x,y)、p2,3(x,
y)、p2,4(x,y)、p3,1(x,y)、p3,2(x,y)、p3,3(x, y) and p3,4(x,y)。
Using above-mentioned formula (2), (3), the first fusion operation, Ye Jijin are carried out to above-mentioned 12 normalized feature vectors
The fusion of line direction dimension,
To calculate L1For (x, y),
L1(x, y)=[p1,1(x,y)+p1,2(x,y)+p1,3(x,y)+p1,4(x,y)]-[p1,1(x,y)p1,2(x,y)+p1,1
(x,y)p1,3(x,y)+p1,1(x,y)p1,4(x,y)+p1,2(x,y)p1,3(x,y)+p1,2(x,y)p1,4(x,y)+p1,3(x,y)p1,4
(x,y)]
, 3 the first fusion results: L can be obtained by the first fusion operation1(x,y)、L2(x, y) and L3(x,y)。
Using above-mentioned formula (4), (5), second of fusion operation is carried out to above-mentioned 3 the first fusion results, namely carry out
The fusion operation of wavelength dimension:
Then have:
Referring to Fig. 3, in some embodiments of the invention, as a kind of achievable mode, the basis in above-mentioned steps S4
Segmentation threshold range detection described image fusion results can wrap when obtaining defective sub-pixel in the screen to be detected
Include following steps:
S41. according to image co-registration as a result, obtaining the ash of image slices vegetarian refreshments corresponding to each sub-pixel in screen to be detected
Angle value range.
S42. the intensity value ranges of the image slices vegetarian refreshments according to corresponding to segmentation threshold range and sub-pixel, judgement is every respectively
A sub-pixel is flawless sub-pixel, dim spot or bright spot:
If segmentation threshold range covers the intensity value ranges of image slices vegetarian refreshments corresponding to sub-pixel, or right with sub-pixel institute
The intensity value ranges for the image slices vegetarian refreshments answered are overlapped, then determine that the sub-pixel is flawless sub-pixel.
If the upper limit value of the intensity value ranges of image slices vegetarian refreshments corresponding to sub-pixel is greater than the upper limit of segmentation threshold range
Value, then determine the sub-pixel for dim spot.
If the lower limit value of the intensity value ranges of image slices vegetarian refreshments corresponding to sub-pixel is less than the lower limit of segmentation threshold range
Value, then determine the sub-pixel for bright spot.
It should be noted that above-mentioned, " segmentation threshold range covers the gray value model of image slices vegetarian refreshments corresponding to sub-pixel
Enclose " the case where include: one in the upper limit or lower limit of the intensity value ranges of image slices vegetarian refreshments corresponding to sub-pixel with
The upper limit or lower limit of segmentation threshold range is overlapped, another is within the scope of segmentation threshold.
It should also be noted that, the brightness value of screen to be detected and the variation of the gray value in the image of screen to be detected become
Gesture is opposite, that is to say, that the brightness value of some pixel is higher in screen to be detected, image of the pixel in screen to be detected
In gray value it is lower；Conversely, the brightness value of some pixel is lower in screen to be detected, the pixel is in screen to be detected
Gray value in image is higher.
In some embodiments of the invention, as a kind of achievable mode, in step s 4 according to segmentation threshold range
Detection image fusion results, before the step of obtaining the defective sub-pixel of screen to be detected, the detection method of above-mentioned screen
Further include: the step of obtaining segmentation threshold range.
Referring to fig. 4, the step of acquisition segmentation threshold range the following steps are included:
A1. the image of zero defect screen is obtained.
Illustratively, it can use industrial camera to shoot above-mentioned zero defect screen, to obtain above-mentioned zero defect screen
The image of curtain.
A2. multiple gabor filtering is carried out by image of multiple Gabor filters to zero defect screen, obtains multiple gal cypresses
Filtered reference image.
A3. image co-registration is carried out to multiple gabor filtering reference pictures respectively, is obtained with reference to blending image and reference picture
Fusion results.
A4. according to reference picture fusion results, determine that image slices vegetarian refreshments is most corresponding to the sub-pixel of zero defect screen
High-gray level value and minimum gradation value obtain segmentation threshold range.
In the detection method of above-mentioned screen, since the texture of the image of the image and screen to be detected of zero defect screen is special
Used Gabor filter is identical when the separation and extraction of sign, therefore separation for the textural characteristics of above-mentioned two image and mentions
Take the difference of effect smaller, thus it is more accurate according to the segmentation threshold range that above-mentioned reference picture fusion results determine, into one
Step improves the accuracy in detection of defective sub-pixel.
Further, in some embodiments of the invention, nothing is determined according to reference picture fusion results in step B4
Before the step of maximum gradation value and minimum gradation value of image slices vegetarian refreshments corresponding to the sub-pixel of defect screen, further includes:
Processing is removed to the image pixel point in the periphery predeterminable area with reference to blending image, treated joins according to removing
It examines blending image and obtains the reference picture fusion results.Pass through the image slices in the periphery predeterminable area to reference blending image
Vegetarian refreshments is removed processing, and maximum gradation value and minimum gradation value can will likely be caused to determine the image slices vegetarian refreshments row of inaccuracy
It removes, to improve the accuracy of segmentation threshold range, further improves the accuracy in detection of defective sub-pixel.
As an example, the width dimensions of above-mentioned periphery predeterminable area can be according to the size of multiple Gabor filter cores
It determines.
In some embodiments of the invention, pass through multiple Gabor filters in above-mentioned steps S2 respectively to acquired
The image of screen to be detected carries out gabor filtering, and before obtaining multiple gabor filtering images, the detection method of above-mentioned screen is also wrapped
It includes: noise reduction and/or image enhancement processing is carried out to the image of screen to be detected.
Illustratively, it may include: the figure to screen to be detected that the above-mentioned image to screen to be detected, which carries out noise reduction process,
As carrying out median filter process.By median filter process, it can be effectively reduced the noise of the sensor for obtaining image, mention
The high subsequent step of separation and extraction effect in to(for) the textural characteristics of the image of screen to be detected.
As another example, it may include: to be checked that the above-mentioned image to screen to be detected, which carries out image enhancement processing,
Survey screen image carry out based on CLAHE (Contrast Limited Adaptive Histogram Equalization, it is right
Be limited histogram equalization than degree) image enhancement processing.
A kind of detection device of screen, including image grabber, Duo Gejia are additionally provided in some embodiments of the present invention
Cypress filter, processor and detector.
Wherein, image grabber is configured to obtain the image of screen to be detected.The gabor filtering of multiple Gabor filters is joined
Number is different, and is configured to carry out gabor filtering to the image of screen to be detected respectively, obtains multiple gabor filtering images.Processor
It is configured to carry out image co-registration to multiple gabor filtering images, obtains image co-registration result.Detector is configured to according to segmentation threshold
It is worth range, detection image fusion results obtain defective sub-pixel in screen to be detected, wherein segmentation threshold range is nothing
The intensity value ranges of image slices vegetarian refreshments corresponding to the sub-pixel of defect.
The detection device of the screen has beneficial effect identical with the detection method of above-mentioned screen, and details are not described herein again.
A kind of computer readable storage medium is additionally provided in some embodiments of the present invention, the computer-readable storage medium
Matter is stored with one or more instructions, with one or more steps in the detection method of above-mentioned screen.
The computer readable storage medium has beneficial effect identical with the detection method of above-mentioned screen, no longer superfluous herein
It states.
The foregoing is merely a specific embodiment of the invention, but scope of protection of the present invention is not limited thereto, any
In the technical scope disclosed by the present invention, any changes or substitutions that can be easily thought of by those familiar with the art, all answers
It is included within the scope of the present invention.Therefore, protection scope of the present invention should be with the scope of protection of the claims
It is quasi-.
Claims (10)
1. a kind of detection method of screen characterized by comprising
Obtain the image of screen to be detected；
Gabor filtering is carried out to the image of acquired screen to be detected respectively by multiple Gabor filters, obtains multiple gal cypresses
Filtering image；Wherein, the gabor filtering parameter of the multiple Gabor filter is different；
Image co-registration is carried out to the multiple gabor filtering image, obtains image co-registration result；
According to segmentation threshold range detection described image fusion results, defective sub-pixel in the screen to be detected is obtained；
Wherein, the segmentation threshold range is the intensity value ranges of image slices vegetarian refreshments corresponding to flawless sub-pixel.
2. the detection method of screen according to claim 1, which is characterized in that the gabor filtering parameter includes deflection
And wavelength；The gabor filtering parameter difference includes that deflection is different and/or wavelength is different.
3. the detection method of screen according to claim 2, which is characterized in that the multiple Gabor filter includes 12
Gabor filter；
The deflection of the multiple Gabor filter includes 0 °, 45 °, 90 ° and 135 °；
The wavelength of the multiple Gabor filter includes N, 2N and 3N；Wherein, N indicates the institute in the image of the screen to be detected
State the sub-pixel column region of screen to be detected image pixel shared on the extending direction perpendicular to the sub-pixel column region
The number of point.
4. the detection method of screen according to claim 2, which is characterized in that described to the multiple gabor filtering image
The step of carrying out image co-registration, obtaining image co-registration result, comprising:
Using following formula (1), to the feature vector f of the multiple gabor filtering imagei,jOperation is normalized in (x, y), obtains
To normalized feature vector pi,j(x, y):
Wherein, i, j are integer, and 1≤i≤m, 1≤j≤n, m indicate the value number of the wavelength of the multiple Gabor filter,
N indicates the value number of the deflection of the multiple Gabor filter；MinF (x, y) indicates the image of the screen to be detected
The minimum value of the gray value F (x, y) of image slices vegetarian refreshments (x, y), maxF (x, y) indicate the image of the image of the screen to be detected
The maximum value of the gray value F (x, y) of pixel (x, y)；
Using following formula (2), (3), to the normalized feature vector pi,j(x, y) carries out the first fusion operation, obtains the
One fusion results Li(x, y):
Bi,j(x, y)=pi,j(x,y)pi,j+1(x,y)+pi,j(x,y)pi,j+2(x,y)+…+pi,j(x,y)pi,n(x,y) (3)；
Using following formula (3), (4), to the first fusion results Li(x, y) carries out the second fusion operation, obtains described image
Fusion results N (x, y):
5. the detection method of screen according to claim 1, which is characterized in that described according to segmentation threshold range detection institute
Image co-registration is stated as a result, the step of obtaining defective sub-pixel in the screen to be detected, comprising:
According to described image fusion results, the ash of image slices vegetarian refreshments corresponding to each sub-pixel in the screen to be detected is obtained
Angle value range；
According to the intensity value ranges of image slices vegetarian refreshments corresponding to the segmentation threshold range and sub-pixel, judge respectively described every
A sub-pixel is flawless sub-pixel, dim spot or bright spot:
If the intensity value ranges of image slices vegetarian refreshments corresponding to the segmentation threshold range covering sub-pixel, or it is right with sub-pixel institute
The intensity value ranges for the image slices vegetarian refreshments answered are overlapped, then determine that the sub-pixel is flawless sub-pixel；
If the upper limit value of the intensity value ranges of image slices vegetarian refreshments corresponding to sub-pixel is greater than the upper limit of the segmentation threshold range
Value, then determine the sub-pixel for dim spot；
If the lower limit value of the intensity value ranges of image slices vegetarian refreshments corresponding to sub-pixel is less than the lower limit of the segmentation threshold range
Value, then determine the sub-pixel for bright spot.
6. the detection method of screen according to claim 1, which is characterized in that described according to segmentation threshold range detection
Described image fusion results further include described in acquisition before the step of obtaining the defective sub-pixel of the screen to be detected
The step of segmentation threshold range；
The step of acquisition segmentation threshold range, comprising:
Obtain the image of zero defect screen；
Multiple gabor filtering is carried out by image of the multiple Gabor filter to the zero defect screen, obtains multiple gal cypresses
Filtered reference image；
Image co-registration is carried out to the multiple gabor filtering reference picture, is obtained with reference to blending image and reference picture fusion knot
Fruit；
According to the reference picture fusion results, image slices vegetarian refreshments corresponding to the sub-pixel of the zero defect screen is determined most
High-gray level value and minimum gradation value obtain the segmentation threshold range.
7. the detection method of screen according to claim 6, which is characterized in that described merged according to the reference picture is tied
Fruit, the step of determining the maximum gradation value and minimum gradation value of image slices vegetarian refreshments corresponding to the sub-pixel of the zero defect screen
Before, further includes: processing is removed to the image pixel point in the periphery predeterminable area with reference to blending image, according to going
The reference picture fusion results are obtained with reference to blending image except treated.
8. the detection method of screen according to claim 1, which is characterized in that pass through multiple Gabor filters point described
Before the step of other image to acquired screen to be detected carries out gabor filtering, obtains multiple gabor filtering images, also wrap
It includes: noise reduction and/or image enhancement processing is carried out to the image of the screen to be detected.
9. a kind of detection device of screen characterized by comprising
Image grabber is configured to obtain the image of screen to be detected；
The gabor filtering parameter of multiple Gabor filters, the multiple Gabor filter is different, and the multiple Gabor filter is matched
It is set to and gabor filtering is carried out to the image of the screen to be detected respectively, obtain multiple gabor filtering images；
Processor is configured to carry out image co-registration to the multiple gabor filtering image, obtains image co-registration result；
Detector is configured to detect described image fusion results according to segmentation threshold range, obtain having in the screen to be detected
The sub-pixel of defect；Wherein, the segmentation threshold range is the gray value of image slices vegetarian refreshments corresponding to flawless sub-pixel
Range.
10. a kind of computer readable storage medium, which is characterized in that the computer-readable recording medium storage has one or more
A instruction, to execute one or more steps in the detection method such as screen according to any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910364098.1A CN110111711A (en) | 2019-04-30 | 2019-04-30 | The detection method and device of screen, computer readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910364098.1A CN110111711A (en) | 2019-04-30 | 2019-04-30 | The detection method and device of screen, computer readable storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110111711A true CN110111711A (en) | 2019-08-09 |
Family
ID=67487923
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910364098.1A Pending CN110111711A (en) | 2019-04-30 | 2019-04-30 | The detection method and device of screen, computer readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110111711A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110675789A (en) * | 2019-09-29 | 2020-01-10 | 联想(北京)有限公司 | Information processing method, system and computer storage medium |
| CN111340752A (en) * | 2019-12-04 | 2020-06-26 | 京东方科技集团股份有限公司 | Screen detection method and device, electronic equipment and computer readable storage medium |
| CN111445421A (en) * | 2020-04-08 | 2020-07-24 | Oppo广东移动通信有限公司 | Image filtering method and device, storage medium and mobile equipment |
| CN111524102A (en) * | 2020-04-10 | 2020-08-11 | 京东方科技集团股份有限公司 | Method and device for detecting screen contamination of liquid crystal display |
| CN111612757A (en) * | 2020-05-18 | 2020-09-01 | 苏州精濑光电有限公司 | Screen crack detection method, device, equipment and storage medium |
| CN111882520A (en) * | 2020-06-16 | 2020-11-03 | 歌尔股份有限公司 | Screen defect detection method and device and head-mounted display equipment |
| WO2022062812A1 (en) * | 2020-09-28 | 2022-03-31 | 歌尔股份有限公司 | Screen defect detection method, apparatus, and electronic device |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102200519A (en) * | 2010-03-26 | 2011-09-28 | 郭上鲲 | Inspection system |
| CN102706881A (en) * | 2012-03-19 | 2012-10-03 | 天津工业大学 | Cloth defect detecting method based on machine vision |
| WO2012157835A1 (en) * | 2011-05-16 | 2012-11-22 | 동국대학교 산학협력단 | Method for managing a medical vascular image by using an image fusing fusion technique |
| US20130004028A1 (en) * | 2011-06-28 | 2013-01-03 | Jones Michael J | Method for Filtering Using Block-Gabor Filters for Determining Descriptors for Images |
| CN104134211A (en) * | 2014-07-23 | 2014-11-05 | 西安工程大学 | Band steel defect detection method based on multi-resolution filtering of Gaussian filter |
| CN106780486A (en) * | 2017-01-16 | 2017-05-31 | 大连海事大学 | A kind of Surface Defects in Steel Plate image extraction method |
| CN107220962A (en) * | 2017-04-07 | 2017-09-29 | 北京工业大学 | A kind of image detecting method and device of tunnel crackle |
| CN107845086A (en) * | 2017-09-19 | 2018-03-27 | 佛山缔乐视觉科技有限公司 | A kind of detection method, system and the device of leather surface conspicuousness defect |
| CN108765402A (en) * | 2018-05-30 | 2018-11-06 | 武汉理工大学 | Non-woven fabrics defects detection and sorting technique |
-
2019
- 2019-04-30 CN CN201910364098.1A patent/CN110111711A/en active Pending
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102200519A (en) * | 2010-03-26 | 2011-09-28 | 郭上鲲 | Inspection system |
| US20110235868A1 (en) * | 2010-03-26 | 2011-09-29 | Kuo Cooper S K | Inspection System |
| WO2012157835A1 (en) * | 2011-05-16 | 2012-11-22 | 동국대학교 산학협력단 | Method for managing a medical vascular image by using an image fusing fusion technique |
| US20130004028A1 (en) * | 2011-06-28 | 2013-01-03 | Jones Michael J | Method for Filtering Using Block-Gabor Filters for Determining Descriptors for Images |
| CN102706881A (en) * | 2012-03-19 | 2012-10-03 | 天津工业大学 | Cloth defect detecting method based on machine vision |
| CN104134211A (en) * | 2014-07-23 | 2014-11-05 | 西安工程大学 | Band steel defect detection method based on multi-resolution filtering of Gaussian filter |
| CN106780486A (en) * | 2017-01-16 | 2017-05-31 | 大连海事大学 | A kind of Surface Defects in Steel Plate image extraction method |
| CN107220962A (en) * | 2017-04-07 | 2017-09-29 | 北京工业大学 | A kind of image detecting method and device of tunnel crackle |
| CN107845086A (en) * | 2017-09-19 | 2018-03-27 | 佛山缔乐视觉科技有限公司 | A kind of detection method, system and the device of leather surface conspicuousness defect |
| CN108765402A (en) * | 2018-05-30 | 2018-11-06 | 武汉理工大学 | Non-woven fabrics defects detection and sorting technique |
Non-Patent Citations (1)
| Title |
|---|
| 陈廉政: "基于图像背景重构和水平集的TFT-LCD缺陷检测算法研究与应用", 《中国优秀博硕士学位论文全文数据库(硕士)信息科技辑》 * |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110675789A (en) * | 2019-09-29 | 2020-01-10 | 联想(北京)有限公司 | Information processing method, system and computer storage medium |
| CN111340752A (en) * | 2019-12-04 | 2020-06-26 | 京东方科技集团股份有限公司 | Screen detection method and device, electronic equipment and computer readable storage medium |
| CN111445421A (en) * | 2020-04-08 | 2020-07-24 | Oppo广东移动通信有限公司 | Image filtering method and device, storage medium and mobile equipment |
| CN111524102A (en) * | 2020-04-10 | 2020-08-11 | 京东方科技集团股份有限公司 | Method and device for detecting screen contamination of liquid crystal display |
| CN111524102B (en) * | 2020-04-10 | 2024-04-16 | 京东方科技集团股份有限公司 | Screen dirt detection method and device of liquid crystal display |
| CN111612757A (en) * | 2020-05-18 | 2020-09-01 | 苏州精濑光电有限公司 | Screen crack detection method, device, equipment and storage medium |
| CN111882520A (en) * | 2020-06-16 | 2020-11-03 | 歌尔股份有限公司 | Screen defect detection method and device and head-mounted display equipment |
| CN111882520B (en) * | 2020-06-16 | 2023-10-17 | 歌尔光学科技有限公司 | Screen defect detection method and device and head-mounted display equipment |
| WO2022062812A1 (en) * | 2020-09-28 | 2022-03-31 | 歌尔股份有限公司 | Screen defect detection method, apparatus, and electronic device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110111711A (en) | The detection method and device of screen, computer readable storage medium | |
| JP6159298B2 (en) | Method for detecting and removing ghost artifacts in HDR image processing using multi-scale normalized cross-correlation | |
| CN105453153B (en) | Traffic lights detects | |
| CN105279372B (en) | A kind of method and apparatus of determining depth of building | |
| CN101971611B (en) | Method of selecting best image and image selection apparatus | |
| CN106934803A (en) | The detection method and device of electronic device surface defect | |
| KR101662846B1 (en) | Apparatus and method for generating bokeh in out-of-focus shooting | |
| CN106770362A (en) | Gross imperfection detection means and method based on AOI | |
| CN106663322B (en) | The method and apparatus of feature in image for identification | |
| CN105741231B (en) | The skin makeup treating method and apparatus of image | |
| CN110909750B (en) | Image difference detection method and device, storage medium and terminal | |
| CN106228541A (en) | The method and device of screen location in vision-based detection | |
| CN105740874B (en) | Determine the method and device of operation coordinate when automatic test script playback | |
| CN104749801B (en) | High Precision Automatic optical detecting method and system | |
| CN109801322A (en) | A kind of light leak test method and device | |
| CN112583999A (en) | Lens contamination detection method for camera module | |
| JP2009036582A (en) | Inspection method, inspection device and inspection program of plane display panel | |
| TWI751869B (en) | Computation apparatus and method of detecting defects for near-eye display | |
| CN105229665A (en) | To the enhancing analysis of the snakelike belt wear assessment based on image | |
| JP4520880B2 (en) | Blot inspection method and blot inspection apparatus | |
| US9536162B2 (en) | Method for detecting an invisible mark on a card | |
| JP2020145594A5 (en) | ||
| CN105894068B (en) | FPAR card design and rapid identification and positioning method | |
| JP2005326323A (en) | Image quality inspection device | |
| CN107451990A (en) | A kind of photograph image altering detecting method using non-linear guiding filtering |
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 |