CN108760767B - Large-size liquid crystal display defect detection method based on machine vision - Google Patents
Large-size liquid crystal display defect detection method based on machine vision Download PDFInfo
- Publication number
- CN108760767B CN108760767B CN201810549451.9A CN201810549451A CN108760767B CN 108760767 B CN108760767 B CN 108760767B CN 201810549451 A CN201810549451 A CN 201810549451A CN 108760767 B CN108760767 B CN 108760767B
- Authority
- CN
- China
- Prior art keywords
- liquid crystal
- crystal screen
- image
- crystal display
- camera
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/95—Investigating the presence of flaws or contamination characterised by the material or shape of the object to be examined
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/95—Investigating the presence of flaws or contamination characterised by the material or shape of the object to be examined
- G01N2021/9513—Liquid crystal panels
Abstract
The invention discloses a large-size liquid crystal display defect detection method based on machine vision, which comprises the steps of partitioning a liquid crystal display, arranging a camera for each partition, calibrating the camera to obtain a distortion coefficient matrix, and then obtaining a homography matrix; enabling the liquid crystal screen to display pure-color pictures, shooting liquid crystal screen images of corresponding partitions by each camera, performing distortion correction by adopting a distortion coefficient matrix, transforming all the liquid crystal screen images to the same coordinate system according to a homography matrix, splicing to obtain complete liquid crystal screen images, detecting edges of the liquid crystal screen, extracting liquid crystal screen area images, performing background texture suppression and image enhancement, and then performing defect detection. The invention can effectively realize the defect detection of the large-size liquid crystal screen and has better universality.
Description
Technical Field
The invention belongs to the technical field of liquid crystal display defect detection, and particularly relates to a large-size liquid crystal display defect detection method based on machine vision.
Background
A liquid crystal display is a very common display device at present. Before the liquid crystal display leaves a factory, the liquid crystal display needs to be subjected to defect detection. At present, most relevant liquid crystal display manufacturers rely on the traditional manual detection method to detect the defects of the liquid crystal display. The manual detection method is mainly based on that human eyes observe and feel the liquid crystal screen and judge and quantify the defects of the liquid crystal screen according to the experience of a detector, and the method has obvious defects. Secondly, if the inspector works for a long time, the inspector is easy to get tired, that is, the inspector needs to rest to adjust after working for a while, which reduces the inspection efficiency and increases the production cost.
With the continuous pursuit of high-quality display effect and visual experience of consumers, the current liquid crystal display industry is developing towards high definition, large size and wide viewing angle, so that the defect detection difficulty of the large-size liquid crystal display is further increased.
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provide a large-size liquid crystal screen defect detection method based on machine vision, which can effectively realize defect detection of the large-size liquid crystal screen by combining machine vision and image processing and has better universality.
In order to achieve the above object, the method for detecting defects of a large-size liquid crystal display based on machine vision of the invention comprises the following steps:
s1: partitioning the liquid crystal display according to actual needs, wherein the number of the scoring areas is N; the method comprises the following steps that N cameras are adopted to form a camera array, each camera shoots a partitioned liquid crystal screen image, and the sight line of each camera is perpendicular to the liquid crystal screen;
s2: respectively displaying a calibration image in each subarea of the liquid crystal screen, and respectively calibrating the corresponding cameras to obtainDistortion coefficient matrix M corresponding to each cameraiThen acquiring a homography matrix H corresponding to each camerai，i＝1,2,…,N；
S3: enabling the liquid crystal screen to display pure-color pictures, and respectively shooting corresponding liquid crystal screen partitions by adopting N cameras to obtain liquid crystal screen images of N partitions;
s4: adopting a distortion coefficient matrix M for each liquid crystal screen imageiCarrying out distortion correction to obtain an image p of the liquid crystal displayi(ii) a From N LCD screen images piSelecting a liquid crystal screen image as a reference image according to N homography matrixes HiConverting the other N-1 liquid crystal screen images into a coordinate system of the reference image to obtain N liquid crystal screen images p'i；
S5: for N LCD screen images p'iCarrying out image splicing to obtain a complete liquid crystal screen image;
s6: performing liquid crystal screen edge detection on the complete liquid crystal screen image to extract a liquid crystal screen area image;
s7: carrying out background texture suppression on the liquid crystal screen area image to obtain a liquid crystal screen area image P;
s8: carrying out image enhancement on the liquid crystal screen area image P to obtain an enhanced liquid crystal screen area image P';
s9, carrying out mean value smoothing on the liquid crystal screen area image P ', recording the smoothed liquid crystal screen area image P', traversing each pixel point (α), if P '(β 0, β 1) is not less than P' (β 2, β 3) + lambda or P '(β 4, β 5) < P' (α) -lambda, wherein P '(α) and P' (α) respectively represent the gray values of the pixel point (α) in the liquid crystal screen area image P 'and P', judging that the pixel point is a defect point, enabling the gray value to be 255, otherwise judging that the pixel point is not a defect point, enabling the gray value to be 0, obtaining a binary image, extracting a connected area from the defect pixel point in the binary image, obtaining a defect area, and marking the defect pixel point in the liquid crystal screen area image to obtain a defect detection result.
The invention relates to a defect detection method of a large-size liquid crystal screen based on machine vision, which comprises the steps of partitioning the liquid crystal screen, setting a camera for each partition, calibrating the camera to obtain a distortion coefficient matrix, and then obtaining a homography matrix; enabling the liquid crystal screen to display pure-color pictures, shooting liquid crystal screen images of corresponding partitions by each camera, performing distortion correction by adopting a distortion coefficient matrix, transforming all the liquid crystal screen images to the same coordinate system according to a homography matrix, splicing to obtain complete liquid crystal screen images, detecting edges of the liquid crystal screen, extracting liquid crystal screen area images, performing background texture suppression and image enhancement, and then performing defect detection. The invention can effectively realize the defect detection of the large-size liquid crystal screen and has better universality.
Drawings
FIG. 1 is a flow chart of an embodiment of a method for detecting defects of a large-size liquid crystal display based on machine vision according to the present invention;
FIG. 2 is a view showing the arrangement of the liquid crystal panel partition and the camera in the present embodiment;
FIG. 3 is a calibration image of a standard LCD panel used in the present embodiment;
FIG. 4 is a diagram showing an example of distortion correction of an image of the upper left divisional liquid crystal panel in the present embodiment;
FIG. 5 is an image after the image distortion of the six subarea liquid crystal screens is corrected in the embodiment;
FIG. 6 is a schematic diagram of feature point matching of an upper left-hand segmented LCD screen image in the present embodiment;
FIG. 7 is a diagram showing 6 LC panel images of a solid color picture in the present embodiment;
FIG. 8 is an image mask diagram in the present embodiment;
FIG. 9 is a complete LCD screen image obtained in the present embodiment;
FIG. 10 is an image of the area of the liquid crystal panel in the present embodiment;
FIG. 11 is an image of an area of a liquid crystal panel after background texture suppression in the present embodiment;
FIG. 12 is a region image of the liquid crystal panel after image enhancement in the present embodiment;
FIG. 13 is a binarized image of the image shown in FIG. 12;
fig. 14 is a defect detection result of the present embodiment.
Detailed Description
The following description of the embodiments of the present invention is provided in order to better understand the present invention for those skilled in the art with reference to the accompanying drawings. It is to be expressly noted that in the following description, a detailed description of known functions and designs will be omitted when it may obscure the subject matter of the present invention.
Examples
FIG. 1 is a flowchart of an embodiment of a method for detecting defects of a large-size liquid crystal display based on machine vision. As shown in FIG. 1, the method for detecting defects of a large-size liquid crystal display based on machine vision mainly comprises two parts: the system comprises a calibration part and a detection part, wherein the calibration part is mainly used for acquiring distortion parameters and a homography matrix of a plurality of cameras for shooting the liquid crystal screen before detecting the defects of the liquid crystal screen so as to be used for multi-image splicing of a subsequent detection part, and the detection part is used for detecting the defects of the actual liquid crystal screen. The invention relates to a large-size liquid crystal display defect detection method based on machine vision, which comprises the following specific steps:
s101: partitioning the liquid crystal screen:
because the large-size liquid crystal screen is large in size, when defect detection is carried out by adopting machine vision, clear liquid crystal screen images are difficult to completely shoot by one camera, so that the liquid crystal screen is partitioned according to actual needs, and the number of the scoring partitions is N. The method comprises the following steps that N cameras are adopted to form a camera array, each camera shoots a partitioned liquid crystal screen image, and the sight line of each camera is perpendicular to the liquid crystal screen. In practical applications, the projection of the camera on the liquid crystal screen is preferably located on the center point of the corresponding partition. In this embodiment, the liquid crystal panel is divided into 2 rows and 3 columns, and 6 partitions are provided in total. Fig. 2 is a view showing the arrangement of the liquid crystal panel partitions and the camera in the present embodiment.
S102: acquiring camera parameters:
respectively displaying a calibration image in each subarea of the liquid crystal display screen, respectively calibrating the corresponding cameras to obtain a distortion coefficient matrix M corresponding to each cameraiThen acquiring a homography matrix H corresponding to each camerai，i＝1,2,…,N。
In this embodiment, a zhangying friend checkerboard calibration method is adopted for calibration, a standard liquid crystal screen calibration image is displayed in a liquid crystal screen, and the standard liquid crystal screen calibration image is set to display one checkerboard image in each partition. Fig. 3 is a calibration image of a standard lcd panel used in the present embodiment. As shown in fig. 3, since the liquid crystal panel in this embodiment is divided into 6 partitions on average, the calibration image of the standard liquid crystal panel is composed of 6 identical checkerboard images, and the calibration method is as follows:
each camera respectively shoots an image of the liquid crystal screen, and the calibration is carried out by referring to a standard liquid crystal screen calibration image to obtain a distortion coefficient matrix Mi. Then according to the distortion coefficient matrix MiAnd carrying out distortion correction on the corresponding ith liquid crystal screen subarea image. Fig. 4 is a diagram showing an example of distortion correction of the image of the upper left divisional liquid crystal panel in the present embodiment. Fig. 5 is an image after the image distortion of the six subarea liquid crystal screens of the embodiment is corrected.
And for the corrected ith liquid crystal screen image, extracting the coordinate position of the inner corner point of the checkerboard image, and performing characteristic point matching with the standard liquid crystal screen calibration image to obtain a characteristic point matching pair, thereby obtaining the characteristic point matching relation of the corresponding camera and the liquid crystal screen. Fig. 6 is a schematic diagram of feature point matching of the upper left partitioned liquid crystal screen image in the present embodiment. And then substituting the coordinates of the matched pair of characteristic points into the following formula:
h represents a homography matrix, (X, Y) represents the coordinates of the characteristic points in the calibration image of the liquid crystal screen in the characteristic point matching pair, and (X, Y) represents the coordinates of the characteristic points in the calibration image of the liquid crystal screen in the characteristic point matching pair subarea. The homography matrix H corresponding to each camera can be obtained by respectively substituting the characteristic point matching pairs of each liquid crystal screen image into the formulai。
The characteristic point matching pair obtained by adopting the inner corner points of the checkerboard has the advantages of rotation invariance and no change along with the change of illumination conditions, the obtained homography matrix can be more accurate, the subsequent image splicing effect is better, and the influence of external factors can be effectively reduced.
S103: acquiring an image of the liquid crystal screen:
and displaying pure-color pictures on the liquid crystal screen, and respectively shooting the corresponding liquid crystal screen subareas by adopting N cameras to obtain liquid crystal screen images of the N subareas. The color of the pure color picture can be set according to the requirement, and the four colors of red, green, blue and white are generally selected. In practical application, only one color can be selected, and four colors can be adopted to perform defect detection once respectively to synthesize detection results, so that the detection results are more accurate. Fig. 7 shows 6 lcd images of a solid color screen in this embodiment.
S104: preprocessing an image of the liquid crystal screen:
adopting a distortion coefficient matrix M for each liquid crystal screen imageiCarrying out distortion correction to obtain an image p of the liquid crystal displayi. From N LCD screen images piSelecting a liquid crystal screen image as a reference image according to N homography matrixes HiConverting the other N-1 liquid crystal screen images into a coordinate system of the reference image to obtain N liquid crystal screen images p'i。
The expression for coordinate transformation of the liquid crystal screen image may be expressed as follows:
namely:
X＝a11u+a12v+a13
Y＝a21u+a22v+a23
Z＝a31u+a32v+a33
wherein, (u, v) represents the pixel point coordinates of the liquid crystal screen subarea image before coordinate transformation, (X, Y, Z) represents the three-dimensional point coordinates of the pixel point of the liquid crystal screen subarea image after coordinate transformation,
the perspective transformation matrix is a homography matrix.
Since the points required in the present invention are all on a plane, the above formula can be simplified to
Wherein, (x, y) represents the pixel point coordinate after the liquid crystal screen partition image coordinate transformation.
As can be seen, from the N homography matrices HiN liquid crystal screen images can be converted into the same coordinate system to obtain N liquid crystal screen images p'i。
S105: splicing images of the liquid crystal screen:
for N LCD screen images p'iAnd carrying out image splicing to obtain a complete liquid crystal screen image. The specific method for splicing the images of the liquid crystal display can be set according to actual needs, and the method adopted in the embodiment is as follows: setting an image mask of each liquid crystal screen in a complete liquid crystal screen image for each liquid crystal screen subarea, and then carrying out p 'on N liquid crystal screen images'iAnd extracting the subarea images, and splicing the N extracted subarea images to obtain a complete liquid crystal screen image. Fig. 8 is an image mask diagram in the present embodiment. The size of the image mask is determined according to the size of the liquid crystal screen image after the homography matrix transformation, because the size of the liquid crystal screen image after the homography matrix transformation is theoretically the same as that of the calibration image. However, in practice, the transformed liquid crystal screen image may have slight deviation, so the transformed liquid crystal screen image is set to be slightly larger than the calibration image, and the size of the image mask is also slightly larger than the calibration image. Then, setting the mask of each subarea according to the condition of the previous liquid crystal screen subareas. Fig. 9 is a complete liquid crystal screen image obtained by the present embodiment. As shown in fig. 8 and 9, by using the image mask, the corresponding subarea image can be obtained, thereby realizing the stitching.
S106: extracting an area image of the liquid crystal screen:
and carrying out liquid crystal screen edge detection on the complete liquid crystal screen image to extract a liquid crystal screen area image. The specific method for detecting the edge of the liquid crystal display can be set according to actual needs, and in this embodiment, a scanning line-based method for detecting the edge of the liquid crystal display is adopted, and the specific method is as follows:
candidate areas of the edges of the liquid crystal screen in four directions are set in the complete liquid crystal screen image in advance, and the candidate areas can be set according to experience. The candidate area is then scanned with scan lines parallel to the edges of the image, i.e. the horizontal edges with scan lines parallel to the x-axis and the vertical edges with scan lines parallel to the y-axis. And scanning lines to obtain maximum gray level jump points, and fitting the maximum gray level jump points as points to be fitted corresponding to the edges of the liquid crystal display to obtain the edges of the liquid crystal display. The specific method for fitting the edge of the liquid crystal display comprises the following steps: let the linear equation of the edge of the liquid crystal screen be y ═ a0+a1And x, obtaining the edge of the liquid crystal display screen according to the point to be fitted by adopting a least square method. By adopting the method, four edges of the upper part, the lower part, the left part and the right part of the liquid crystal screen area in the complete liquid crystal screen image can be obtained, and the liquid crystal screen area image is extracted according to the four edges, so that the background information part is filtered. Fig. 10 is an image of the liquid crystal panel region in the present embodiment.
S107: background texture suppression:
due to the physical structure of the liquid crystal screen, the shot image may present periodic horizontal and vertical texture stripes, which greatly affect the subsequent detection and need to be filtered by using background texture suppression. Therefore, the background texture suppression is performed on the liquid crystal screen area image extracted in step S106 to obtain a liquid crystal screen area image P. The specific algorithm for background texture suppression can be selected according to actual needs, and the background texture suppression method based on gaussian filtering is adopted in the embodiment, and the specific method is as follows:
and carrying out Fourier transform on the liquid crystal screen area image to obtain a spectrogram G. Two-dimensional fourier transform of an image is a common technique in the field of image processing, and is not described herein again. And filtering the spectrogram G by adopting a frequency domain Gaussian low-pass filter to obtain a filtered spectrogram G'. And performing inverse Fourier transform on the spectrogram G' to obtain an image P of the liquid crystal screen area, wherein the background texture is suppressed.
The method is simple to implement and consumes short time, and most importantly, the method has very little influence on the defect information of the liquid crystal screen image and provides favorable precondition for the subsequent defect detection work. Fig. 11 is an image of the liquid crystal panel region after the background texture suppression in the present embodiment.
S108: image enhancement:
and carrying out image enhancement on the liquid crystal screen area image P to obtain an enhanced liquid crystal screen area image P'. The specific algorithm of image enhancement can be determined according to actual needs, and the embodiment adopts an image enhancement method based on gaussian difference, and the specific method is as follows: and performing Gaussian difference on the liquid crystal screen region image P by adopting two Gaussian kernels with the same size and different standard deviations, and taking the obtained difference image as a liquid crystal screen region image P'. And the Gaussian difference is adopted for image enhancement, so that the defects of the liquid crystal display screen are more prominent. Fig. 12 is an image of the liquid crystal panel region after image enhancement in the present embodiment.
S109: and (3) defect detection:
the method comprises the steps of conducting mean smoothing on a liquid crystal screen region image P ', recording a smoothed liquid crystal screen region image P'. traversing each pixel point (α), if P '(β 0, β 1) ≥ P' (β 2, β 3) + lambda or P '(β 4, β 5) is less than P' (α) -lambda, wherein P '(α) and P' (α) respectively represent gray values of the pixel (α) in the liquid crystal screen region image P 'and P', lambda represents a preset threshold, setting according to an actual environment, judging that the pixel point is a defect point, enabling the gray value to be 255, otherwise, judging that the pixel point is not the defect point, enabling the gray value to be 0, and accordingly obtaining a binary image.
The defect segmentation is carried out based on the local threshold, so that the defect detection method has a good detection effect on various low-contrast defects, is high in detection precision, and can achieve the detection precision of one pixel level. Fig. 13 is a binarized image of the image shown in fig. 12. Fig. 14 is a defect detection result of the present embodiment. As shown in fig. 14, in the present embodiment, a defective area is indicated in the liquid crystal panel area image P obtained by suppressing the background texture. As can be seen from fig. 14, the defect detection of the large-sized liquid crystal panel can be effectively performed by the present invention.
Although illustrative embodiments of the present invention have been described above to facilitate the understanding of the present invention by those skilled in the art, it should be understood that the present invention is not limited to the scope of the embodiments, and various changes may be made apparent to those skilled in the art as long as they are within the spirit and scope of the present invention as defined and defined by the appended claims, and all matters of the invention which utilize the inventive concepts are protected.
Claims (8)
1. A large-size liquid crystal display defect detection method based on machine vision is characterized by comprising the following steps:
s1: partitioning the liquid crystal display according to actual needs, wherein the number of the scoring areas is N; the method comprises the following steps that N cameras are adopted to form a camera array, each camera shoots a partitioned liquid crystal screen image, and the sight line of each camera is perpendicular to the liquid crystal screen;
s2: respectively displaying a calibration image in each subarea of the liquid crystal display screen, respectively calibrating the corresponding cameras to obtain a distortion coefficient matrix M corresponding to each cameraiThen acquiring a homography matrix H corresponding to each camerai，i＝1,2,…,N；
S3: enabling the liquid crystal screen to display pure-color pictures, and respectively shooting corresponding liquid crystal screen partitions by adopting N cameras to obtain liquid crystal screen images of N partitions;
s4: adopting a distortion coefficient matrix M for each liquid crystal screen imageiCarrying out distortion correction to obtain an image p of the liquid crystal displayi(ii) a From N LCD screen images piSelecting a liquid crystal screen image as a reference image according to N homography matrixes HiConverting the other N-1 liquid crystal screen images into a coordinate system of the reference image to obtain N liquid crystal screen images p'i；
S5: for N LCD screen images p'iCarrying out image splicing to obtain a complete liquid crystal screen image;
s6: performing liquid crystal screen edge detection on the complete liquid crystal screen image to extract a liquid crystal screen area image;
s7: carrying out background texture suppression on the liquid crystal screen area image to obtain a liquid crystal screen area image P;
s8: carrying out image enhancement on the liquid crystal screen area image P to obtain an enhanced liquid crystal screen area image P';
s9, performing mean value smoothing on the liquid crystal screen region image P ', recording the smoothed liquid crystal screen region image as P', traversing each pixel point (α), if P '(β 0, β 1) is more than or equal to P' (β 2, β 3) + lambda or P '(β 4, β 5) < P' (α) -lambda, wherein P '(α) and P' (α) respectively represent gray values of the pixel (α) in the liquid crystal screen region image P ', P', and lambda represents a preset threshold, judging that the pixel point is a defect point, enabling the gray value to be 255, otherwise, judging that the pixel point is not the defect point, enabling the gray value to be 0, thereby obtaining a binary image, extracting a connected region from the defect pixel point in the binary image, thereby obtaining a defect region, and marking the defect region image P to obtain a defect detection result.
2. The method for detecting the defect of the large-size liquid crystal screen according to claim 1, wherein the projection of the camera on the liquid crystal screen is located on the center point of the corresponding subarea.
3. The method for detecting defects of a large-size liquid crystal display according to claim 1, wherein the calibration method of the camera in the step S2 is as follows: displaying a standard liquid crystal screen calibration image in a liquid crystal screen, wherein the standard liquid crystal screen calibration image is set to be required to display a checkerboard image in each partition of the liquid crystal screen, each camera respectively shoots a liquid crystal screen image, calibration is carried out by referring to the standard liquid crystal screen calibration image, and a distortion coefficient matrix M is obtainediThen based on the distortion coefficient matrix MiCarrying out distortion correction on the corresponding ith liquid crystal screen image; for the corrected ith liquid crystal screen image, extracting the coordinate position of the angular point in the checkerboard image and the standard liquidCarrying out feature point matching on the crystal screen calibration image to obtain feature point matching pairs, and calculating a homography matrix H corresponding to the camera according to the feature point matching pairsi。
4. The method for detecting the defect of the large-size liquid crystal display panel according to claim 1, wherein the method for stitching the images of the liquid crystal display panel in the step S5 comprises the following steps: setting an image mask of each liquid crystal screen in a complete liquid crystal screen image for each liquid crystal screen subarea, and then carrying out p 'on N liquid crystal screen images'iAnd extracting the subarea images, and splicing the N extracted subarea images to obtain a complete liquid crystal screen image.
5. The method for detecting the defect of the large-size liquid crystal display panel according to claim 1, wherein the method for detecting the edge of the liquid crystal display panel area in the step S6 comprises: the method comprises the steps of setting candidate areas of liquid crystal screen edges in four directions in a complete liquid crystal screen image in advance, scanning in the candidate areas by adopting scanning lines parallel to the image edges, obtaining maximum gray level jump points through scanning of the scanning lines, and fitting the maximum gray level jump points as points to be fitted corresponding to the liquid crystal screen edges to obtain the liquid crystal screen edges.
6. The method for detecting the defects of the large-size liquid crystal display panel as claimed in claim 5, wherein the fitting of the edges of the liquid crystal display panel adopts a least square method.
7. The method for detecting defects of a large-size liquid crystal display panel according to claim 1, wherein the method for suppressing the background texture in the step S7 is as follows: carrying out Fourier transform on the liquid crystal screen area image to obtain a spectrogram G; and filtering the spectrogram G by using a frequency domain Gaussian low-pass filter to obtain a filtered spectrogram G ', and performing inverse Fourier transform on the spectrogram G' to obtain a liquid crystal screen area image P.
8. The method for detecting the defect of the large-size liquid crystal display panel as claimed in claim 1, wherein the image enhancement in the step S8 is performed by: and performing Gaussian difference on the liquid crystal screen region image P by adopting two Gaussian kernels with the same size and different standard deviations, and taking the obtained difference image as a liquid crystal screen region image P'.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810549451.9A CN108760767B (en) | 2018-05-31 | 2018-05-31 | Large-size liquid crystal display defect detection method based on machine vision |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810549451.9A CN108760767B (en) | 2018-05-31 | 2018-05-31 | Large-size liquid crystal display defect detection method based on machine vision |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108760767A CN108760767A (en) | 2018-11-06 |
| CN108760767B true CN108760767B (en) | 2020-07-17 |
Family
ID=64001275
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810549451.9A Active CN108760767B (en) | 2018-05-31 | 2018-05-31 | Large-size liquid crystal display defect detection method based on machine vision |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108760767B (en) |
Families Citing this family (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109239901B (en) * | 2018-11-07 | 2021-08-27 | 凌云光技术股份有限公司 | Method and device for quickly calibrating focusing surface and positioning focusing of microscopic imaging system |
| CN109376792B (en) * | 2018-11-07 | 2022-02-18 | 河北工业大学 | Photovoltaic cell appearance defect classification method based on multi-channel residual error neural network |
| CN109813537A (en) * | 2019-03-22 | 2019-05-28 | 昆山钧沃光电有限公司 | A kind of display screen detection device and method |
| CN110533686B (en) * | 2019-10-08 | 2022-03-04 | 凌云光技术股份有限公司 | Method and system for judging whether line frequency of linear array camera is matched with object motion speed |
| CN111025701B (en) * | 2019-12-30 | 2022-06-07 | 凌云光技术股份有限公司 | Curved surface liquid crystal screen detection method |
| CN111369480B (en) * | 2020-01-20 | 2023-10-24 | 凌云光技术股份有限公司 | Method and device for processing periodic texture |
| CN111432198A (en) * | 2020-03-30 | 2020-07-17 | 中国人民解放军陆军装甲兵学院 | Perspective transformation-based projection type three-dimensional display system correction method |
| CN111551559A (en) * | 2020-05-13 | 2020-08-18 | 深圳市全洲自动化设备有限公司 | LCD (liquid Crystal display) liquid crystal screen defect detection method based on multi-view vision system |
| CN112683916A (en) * | 2020-12-17 | 2021-04-20 | 华能新能源股份有限公司云南分公司 | Method and device for identifying missing or mounting error of small hardware fittings of current collecting line tower |
| CN112649985B (en) * | 2020-12-25 | 2022-02-22 | 武汉华星光电技术有限公司 | Display panel and display device |
| CN113916892B (en) * | 2021-09-28 | 2022-10-28 | 南京理工大学 | Multi-view vision-based brake disc gluing defect detection device and method |
| CN113570605B (en) * | 2021-09-28 | 2021-12-31 | 深圳市绘晶科技有限公司 | Defect detection method and system based on liquid crystal display panel |
| CN113608378B (en) * | 2021-10-08 | 2021-12-10 | 深圳市绘晶科技有限公司 | Full-automatic defect detection method and system based on LCD (liquid crystal display) process |
| CN114757913A (en) * | 2022-04-15 | 2022-07-15 | 电子科技大学 | Display screen defect detection method |
| CN115082424B (en) * | 2022-07-19 | 2022-11-08 | 苏州鼎纳自动化技术有限公司 | 3D detection method of liquid crystal display screen |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5446536A (en) * | 1992-05-06 | 1995-08-29 | Nippon Sheet Glass Co., Ltd. | System of detecting optical distortion of a light-transmitting plate-like member |
| CN103344648A (en) * | 2013-07-10 | 2013-10-09 | 鞍钢股份有限公司 | Method and system for detecting surface quality of steel sheet based on graphics processing unit (GPU) |
| CN104568986A (en) * | 2015-01-26 | 2015-04-29 | 中国科学院半导体研究所 | Method for automatically detecting printing defects of remote controller panel based on SURF (Speed-Up Robust Feature) algorithm |
| CN105044126A (en) * | 2015-07-22 | 2015-11-11 | 瑞安市质量技术监督检测院 | Visual detection system for large-width continuous surface defects |
| CN106442560A (en) * | 2016-08-23 | 2017-02-22 | 汕头大学 | Positioning measurement and defect detection method of display screen |
| CN107179322A (en) * | 2017-06-15 | 2017-09-19 | 长安大学 | A kind of bridge bottom crack detection method based on binocular vision |
| CN107248159A (en) * | 2017-08-04 | 2017-10-13 | 河海大学常州校区 | A kind of metal works defect inspection method based on binocular vision |
| CN107292879A (en) * | 2017-07-17 | 2017-10-24 | 电子科技大学 | A kind of sheet metal surface method for detecting abnormality based on graphical analysis |
| CN107389687A (en) * | 2017-06-27 | 2017-11-24 | 北京航空航天大学 | A kind of electronic component appearance images harvester and its acquisition method |
| CN107424118A (en) * | 2017-03-28 | 2017-12-01 | 天津大学 | Based on the spherical panorama mosaic method for improving Lens Distortion Correction |
| CN107515227A (en) * | 2017-07-27 | 2017-12-26 | 武汉精测电子技术股份有限公司 | Based on the lossless method for compressing image of defect and defects of display panel detecting system |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8867817B1 (en) * | 2012-10-29 | 2014-10-21 | Amazon Technologies, Inc. | Display analysis using scanned images |
-
2018
- 2018-05-31 CN CN201810549451.9A patent/CN108760767B/en active Active
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5446536A (en) * | 1992-05-06 | 1995-08-29 | Nippon Sheet Glass Co., Ltd. | System of detecting optical distortion of a light-transmitting plate-like member |
| CN103344648A (en) * | 2013-07-10 | 2013-10-09 | 鞍钢股份有限公司 | Method and system for detecting surface quality of steel sheet based on graphics processing unit (GPU) |
| CN104568986A (en) * | 2015-01-26 | 2015-04-29 | 中国科学院半导体研究所 | Method for automatically detecting printing defects of remote controller panel based on SURF (Speed-Up Robust Feature) algorithm |
| CN105044126A (en) * | 2015-07-22 | 2015-11-11 | 瑞安市质量技术监督检测院 | Visual detection system for large-width continuous surface defects |
| CN106442560A (en) * | 2016-08-23 | 2017-02-22 | 汕头大学 | Positioning measurement and defect detection method of display screen |
| CN107424118A (en) * | 2017-03-28 | 2017-12-01 | 天津大学 | Based on the spherical panorama mosaic method for improving Lens Distortion Correction |
| CN107179322A (en) * | 2017-06-15 | 2017-09-19 | 长安大学 | A kind of bridge bottom crack detection method based on binocular vision |
| CN107389687A (en) * | 2017-06-27 | 2017-11-24 | 北京航空航天大学 | A kind of electronic component appearance images harvester and its acquisition method |
| CN107292879A (en) * | 2017-07-17 | 2017-10-24 | 电子科技大学 | A kind of sheet metal surface method for detecting abnormality based on graphical analysis |
| CN107515227A (en) * | 2017-07-27 | 2017-12-26 | 武汉精测电子技术股份有限公司 | Based on the lossless method for compressing image of defect and defects of display panel detecting system |
| CN107248159A (en) * | 2017-08-04 | 2017-10-13 | 河海大学常州校区 | A kind of metal works defect inspection method based on binocular vision |
Non-Patent Citations (5)
| Title |
|---|
| TFT-LCD Mura defects automatic inspection system using linear;S-L Chen et al.;《Proc. IMechE》;20080704;第222卷;第1489-1501页 * |
| 基于动态阈值分割的汽车轮毂缺陷检测技术研究;张俊生;《中国优秀硕士学位论文全文数据库 信息科技辑》;20120815;第3.2.2节 * |
| 基于机器视觉的OLED缺陷检测系统设计;李桂锋;《中国优秀硕士学位论文全文数据库 信息科技辑》;20111215;第2.4节 * |
| 无人机图像处理关键技术的研究与实现;温尔雅;《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》;20180215;第三、四章 * |
| 液晶显示屏缺陷自动检测系统的研究;张戈;《中国优秀硕士学位论文全文数据库 信息科技辑》;20081115;第2.3节 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108760767A (en) | 2018-11-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108760767B (en) | Large-size liquid crystal display defect detection method based on machine vision | |
| CN111563889B (en) | Liquid crystal screen Mura defect detection method based on computer vision | |
| TWI435162B (en) | Low complexity of the panoramic image and video bonding method | |
| US7019713B2 (en) | Methods and measurement engine for aligning multi-projector display systems | |
| WO2020042153A1 (en) | Correction method, apparatus, and system for full-screen arbitrary splicing | |
| US9810602B2 (en) | Method and system for measuring lens distortion | |
| EP2571261A1 (en) | Method and device for multi-camera image calibration | |
| CN111025701B (en) | Curved surface liquid crystal screen detection method | |
| CN111862224B (en) | Method and device for determining external parameters between camera and laser radar | |
| CN107024485A (en) | The defect inspection method and device of camber display screen | |
| CN109495729B (en) | Projection picture correction method and system | |
| JP2005331929A5 (en) | ||
| CN108716890A (en) | A kind of high-precision size detecting method based on machine vision | |
| TWI621827B (en) | Method of length measurement for 2d photography | |
| CN110060625B (en) | LED display screen acquisition vignetting compensation method | |
| JP2009229197A (en) | Linear defect detecting method and device | |
| CN108508022B (en) | Multi-camera splicing imaging detection method | |
| WO2019144482A1 (en) | Mura repair method for display panel | |
| CN109493293A (en) | A kind of image processing method and device, display equipment | |
| TWI758630B (en) | Image zooming method and device | |
| CN111462105A (en) | Method for measuring sawtooth degree of edge pixel | |
| CN109961393A (en) | Subpixel registration and splicing based on interpolation and iteration optimization algorithms | |
| CN106504200B (en) | Image illumination compensation method and system based on hue offset estimation and point-by-point hue mapping | |
| CN112085803B (en) | Multi-lens multi-detector spliced camera color consistency processing method | |
| CN114331835A (en) | Panoramic image splicing method and device based on optimal mapping matrix |
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 |