JP2004226272A - Method and apparatus for detecting spot defects - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a detection method and a detection apparatus of stain defects that have high detection output also for a defect having a low contract or a small size, automatically determine a threshold for detection as the stain defects, based on statistical brightness data in a detected image, and further achieve the quantitative evaluation of the detected stain defects. <P>SOLUTION: A screen to be inspected is imaged by a CCD camera 6, the difference between an image being captured by the imaging and a background image 14 that is created in advance is obtained to create an inspection image 15, the inspection image is flattened, reduced images in a plurality of stages are created from a flattened image 16, each of the reduced image is filtered for emphasizing defects, the statistical data of the brightness value of each pixel in an image 17 after filtering are calculated, a threshold is determined, based on the brightness statistical data, for judging the presence or absence of defect candidates, and the defect candidates are extracted from an image that is determined to have the defect candidates by using the threshold. <P>COPYRIGHT: (C)2004,JPO&NCIPI

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to a method and an apparatus for automatically detecting a spot defect on a screen in an inspection process of a liquid crystal panel or the like.
[0002]
[Prior art]
One of the defects that appear on a screen of a liquid crystal display device or the like is a so-called spot defect. A spot defect is a state in which a certain area of the display screen has a difference in luminance from another area, and a state in which a part that is lighter or darker than its surroundings is in a certain narrow range. However, there is no strict definition, and it is called an uneven defect or a spot defect having a small defect size. Since such unevenness defects or spot defects deteriorate image quality, they are subject to visual inspection of the display. Conventionally, a visual inspection by a person has been usual, but recently, an automatic inspection has been performed. There have already been many proposals for an automatic inspection method for unevenness defects or spot defects, for example, there is Patent Document 1.
Patent Literature 1 creates inspection area data by performing hierarchization processing with the original image data captured by imaging as the lowest order, performs texture analysis on the inspection area data, and creates two-dimensional statistical data. This is a method for detecting unevenness defects from two-dimensional statistical data.
[0003]
[Patent Document 1]
JP-A-10-10007 (Claims, paragraphs [0010] to [0021], FIGS. 1 to 6)
[0004]
[Problems to be solved by the invention]
However, in the detection method of Patent Literature 1, since the defect enhancement processing is not performed, the detection power for defects with low contrast, particularly small-sized defects, is insufficient in the micro hierarchy, and the illumination of the illumination is also low in the macro hierarchy. There is a problem that it is not possible to distinguish a change in luminance caused by the influence of a light source or an imaging lens from a defect to be inspected.
[0005]
The present invention has been made in view of the above-described problems, and has a high detection power even for a low-contrast or small-size defect, and detects a threshold for detecting a stain defect as a detection image. It is an object of the present invention to provide a method and apparatus for detecting a stain defect, which is automatically determined based on luminance statistical data in the image, and enables quantitative evaluation of the detected stain defect.
[0006]
[Means for Solving the Problems]
The method for detecting a stain defect according to the present invention includes a step of imaging a screen to be inspected,
From the image captured by the imaging, a step of taking a difference from a background image created in advance to create an inspection image,
Performing a flattening process of the inspection image;
A step of creating a plurality of reduced images from the flattened image,
Performing a filter process for defect enhancement on each of the reduced images, and calculating statistical data of the luminance value of each pixel in the image after the filter process,
A step of determining a threshold based on the luminance statistical data to determine the presence or absence of a defect candidate, and a step of extracting a defect candidate from an image determined to have a defect candidate using the threshold,
It is characterized by having.
[0007]
In the present invention, in order to detect from a low-contrast or large-stain defect to a small-stain defect, first, an image of a screen to be inspected is taken, and a difference between the screen and the background image is taken from the image. An inspection image, which is a background difference image from which a luminance change caused by other than the above, has been removed. Then, a flattening process of the inspection image is performed, and an image size reduction process of reducing the image size in a plurality of stages from the flattened image is performed. By this reduction processing, various size defects in the inspection image are also reduced, and as a result, any of the plurality of reduced images is present as a defect having a size detectable by the processing described later. Next, by performing filter processing for defect enhancement on each of the reduced images, the contrast of the stain defect is enhanced. Therefore, spot defects can be detected with high accuracy regardless of the size of the defect or the level of the contrast. That is, the detection power of the stain defect is high.
Furthermore, the threshold for judging the presence or absence of a stain defect and extracting a defect candidate is automatically determined based on the luminance statistical data by calculating the statistical data of the luminance value of each pixel in the detected image. Is done. Since the presence or absence of a defect candidate is determined from the threshold value and the statistical data, it can be easily determined in a short time whether each detected image is a non-defective non-defective product.
Further, a defect candidate is extracted from the image determined as having a defect candidate by using the threshold value, and a blob process for obtaining a characteristic value of the defect candidate is performed. A predetermined process is performed based on the characteristic value and the statistical data. Since the evaluation value of the defect candidate is calculated by the equation, the degree of the defect candidate can be quantitatively evaluated.
[0008]
Further, in the present invention, when an image is captured by imaging, it is preferable that the data of the captured image has 12 bits of 4096 gradations or more.
By using such high-resolution image data, the accuracy of the luminance statistical data is increased, so that the defect detection accuracy can be further improved. Further, the accuracy of the evaluation value of the defect candidate described later is also improved.
[0009]
The background image is preferably an average of a plurality of images captured by the same optical system and the same imaging system.
The background image is created to remove a luminance change caused by a part other than the inspection target and extract a defect to be inspected. Therefore, the background image is an image of only a change in luminance caused by a part other than the inspection object, and the same optical system and the same imaging system capture a plurality of display screens with as few defects as possible, and average the plurality of captured images. By weakening, the component of the defect part of the inspection object which is randomly present in the image is weakened, and only the luminance change which is always present at the same position in the image and which is caused by a non-inspection object such as a screen, illumination, or lens characteristics. The remaining background image is obtained.
[0010]
In the filter processing, a top hat filter or a well filter is used.
Since the stain defect includes a light defect (white stain defect) and a dark defect (black stain defect), a top hat filter or a well filter is used to detect both of these defects. The top hat filter enhances bright defects in the positive direction, and the well filter has the effect of enhancing dark defects in the positive direction. In particular, in the present invention, both the white spot defect and the black spot defect are detected by using the same top hat filter for both the light defect and the dark defect, and adding the half value 2048 of 4096 as an offset value simultaneously with the filtering process. I can do it.
[0011]
In the present invention, the step of creating an inspection image includes a step of extracting a display area from an image including a screen to be inspected, and applying a geometric deformation to the display area to make it a rectangle.
Since the inspection image is created based on an image obtained by capturing the screen of the inspection target, if the inspection target screen is an image projected on the screen by, for example, a projector, the image includes an edge portion of the screen. Or the display area may be oblique to the screen. Therefore, when an inspection image is created, a display area is extracted from an image including a screen to be inspected, and the display area is subjected to geometric deformation to form a rectangle, thereby distorting the image as described above. Deformation and the like can be corrected.
[0012]
As described above, the threshold value of the defect candidate is determined for each of the light defect and the dark defect so that both the white spot defect and the black spot defect can be detected.
The method further includes a step of obtaining characteristic values of the defect candidates for the extracted defect candidates using blob processing, and calculating an evaluation value based on the characteristic values and luminance statistical data.
Therefore, defect candidates can be objectively and quantitatively evaluated, and accurate evaluation can be performed.
The evaluation value of the defect candidate is calculated for each of the bright defect and the dark defect, as in the case of the threshold value.
[0013]
The threshold is determined using the average value and the standard deviation of the luminance statistical data. For example, a white spot threshold for extracting a light defect can be obtained from a calculation formula of average luminance data + a1 × standard deviation, and a black stain threshold for extracting a dark defect can be obtained from average luminance data−a2 × standard deviation. A1 and a2 in the calculation formula are certain fixed constants.
The evaluation value of the defect candidate is calculated using the average value and the standard deviation of the luminance statistical data and the maximum value and the minimum value of the defect candidate obtained by the blob processing.
Therefore, the stain defect can be objectively evaluated for each product (including a part; the same applies hereinafter), and the rank of the stain defect and the grade of the product can be obtained. These statistical data can be used for quality control.
[0014]
A stain defect detection device according to the present invention uses the stain defect detection method according to any one of claims 1 to 12.
More specifically, the inspection apparatus includes an imaging unit and an inspection apparatus main body that performs image processing. The inspection apparatus main body is generally configured by a computer. By incorporating an inspection program for performing each of the above-described processes into this computer, a stain defect can be automatically inspected.
[0015]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, embodiments of the present invention will be described with reference to the drawings.
FIG. 1 is a configuration diagram of a stain defect detection device showing an embodiment of the present invention.
In this embodiment, for example, the screen 10 to be inspected is a screen of a liquid crystal panel (also referred to as a liquid crystal light valve) 2 using a TFT element by the projector 1. When performing an inspection, the image 4 is projected on the screen 3 by the projector 1. The image 4 is drawn by giving a predetermined pattern to the liquid crystal panel 2 by the pattern generator 5. For example, an image 4 is picked up by a CCD camera 6 as an image pickup means, and the image signal is converted from an analog signal to a digital signal by an A / D converter (not shown) and is taken into a computer 7 which is a main body of the inspection apparatus. At this time, the image data is represented by, for example, 12-bit data with black being “0” and white being “4095” for each pixel by the A / D converter with a luminance value of 4096 gradations. Further, the computer 7 processes the image data of the image 4 fetched into the image memory by a method described later, thereby detecting a spot defect for each light / dark defect. In order to detect defects, filter images for defect enhancement using multiple reduced sizes and filter processing for defect enhancement, and then perform statistical processing on luminance information in the detected images, and extract defect candidates based on the statistical data. Is determined, defect candidates are extracted, and an evaluation value for quantitatively evaluating the extracted defect candidates is calculated. These inspection results are displayed on the display device 8.
[0016]
FIG. 2 is a flowchart used in the stain defect detection processing, and FIG. 3 is a schematic diagram of an image after image processing at each stage from an input image to a stain defect detection image. This spot defect detection processing is automatically performed according to the inspection program incorporated in the computer 7 or the image processing apparatus.
The processing procedure will be described according to the flowchart of FIG.
[0017]
(1) Display area extraction processing (step S1)
Extracting a display area means extracting only a screen portion to be inspected from an input image captured by imaging. For example, FIG. 3A shows an input image at the time of imaging. As shown in this figure, the input image 11 at the time of imaging includes the edge portion 31 of the screen 3 or the display area image 12 corresponding to the screen portion is not exactly rectangular but is distorted obliquely with respect to the screen 3. May go out. This is due to the fact that the screen 3 and the CCD camera 6 of the imaging means are not strictly parallel to each other, or that distortion occurs due to the projection lens characteristics of the projector 1, the lens characteristics of the CCD camera 6, and the like. In addition, even when an image of a liquid crystal panel or the like is directly captured instead of indirectly captured as described above, distortion or deformation of the input image may occur. Of course, when the imaging unit or the screen of the inspection target is accurately set and imaged so that the edge portion 31 does not enter as described above, for example, the field of view is set so that the entire screen portion of the detection target falls within the field of view of the imaging unit. When set correctly, the display area extraction processing and the correction processing described below can be omitted, and the image captured by imaging directly becomes the original image.
[0018]
When the input image 11 is distorted as shown in FIG. 3A, only the display area image 12 of the screen portion is extracted as shown in FIG. A display area correction image 13 that is deformed and corrected to an accurate rectangle is created. Here, the display area correction image 13 is an original image to be actually inspected.
In the image correction processing by the geometric deformation, the coordinates of the four corners of the display area image 12 that does not include the edge portion 31 of the screen 3 are detected by pattern matching processing, and the coordinates match the coordinates of the four corners of the rectangle. This is performed by converting the coordinates to.
By setting the size of the rectangle set at this time to, for example, 1200 × 1000 pixels, the image size of the original image 13 becomes 1200 × 1000 pixels. This size is not particularly fixed, and may be set to a size equal to or close to the pixel size of the CCD camera.
[0019]
(2) Background image difference processing (step S2)
The background image difference process is a process of subtracting a previously created background image 14 as shown in FIG. 3C from the original image 13 created as described above. The background image 14 is obtained by, for example, imaging about 20 liquid crystal panels with few defects with the apparatus configuration of the present invention (the configuration including the same optical system and the same imaging system) and averaging the images. It is created in the same manner as the original image 13. The background image 14 is stored in the memory of the computer 7. By subtracting the background image 14 from the original image 13, it is possible to remove a change in brightness caused by the illumination of the screens and imaging units other than the inspection target, the projection lens characteristics of the projector 1, and the like included in the original image 13. Can be. In this difference processing, 2048 (4096 × １ ／) is added as an offset value so that the luminance data does not become a negative value.
As a result of the background image difference processing, a background difference image as shown in FIG. If so, the background difference image (inspection image) 15 contains a spot defect 20 having a different size or contrast or an uneven defect 21 having a relatively large size.
[0020]
(3) Flattening process (Step S3)
The background difference image (inspection image) 15 has a large luminance change that occurs in the entire display area, that is, a change in brightness such that there is a bright or dark place over a wide range that is not so problematic for humans. As described above, there is a possibility that the unevenness defect 21 having a relatively large size may be included. In order to exclude the unevenness defect 21 from the detection target, the background difference image, that is, the inspection image 15 is flattened. By this flattening process, a large change in brightness such as illuminance unevenness and an uneven defect having a large size are removed. Therefore, only a stain defect having a relatively small size and a stain defect having a different contrast remain in FIG. The flattened image 16 as shown in FIG.
The flattening process is a process using a smoothing filter or a process for smoothing by a morphological process or the like.
[0021]
(4) Image size reduction processing (steps S4 to S8)
Although the size of the spot defect is small, there are various sizes among them, and the contrast of the spot defect is also various such as high and low.
Since a top hat filter described later can enhance only a spot defect having a predetermined size, the image size is reduced in several steps from the flattened image 16 in order to correspond to a spot defect having various sizes existing in the flattened image 16. Is performed. Here, （(600 × 500 pixels), ４ (300 × 250 pixels), ８ (150 × 125 pixels), 1/16 ( Five reduced images reduced to five stages of 75 × 62 pixels) and 1/32 (38 × 31 pixels) are created.
In the method of reducing the image size, as shown in FIG. 4, the average value of the luminance data of four pixels of the original image is allocated to one pixel of the new image to reduce the size to half the size. By repeating this method, the image size can be reduced by half.
By gradually reducing the image size of the flattened image 16 in this manner, some of the reduced images have a size that can be enhanced by the top hat filter.
[0022]
(5) Top hat filter processing (steps S9 to S13)
Next, a contrast of a spot defect is enhanced using a Tophat filter as a spatial filter for each reduced image. The top-hat filter is a filter for enhancing the contrast, and is composed of, for example, 7 × 7 pixels as shown in FIG. 5, and performs a convolution operation on the target pixel with this filter configuration value to obtain a background in the pixel. If there is a bright spot (here, a white spot defect) having a size of about 3 × 3 pixels as compared with, this is emphasized.
However, the top hat filter is effective for detecting a light defect (white spot) because the result of enhancing a bright point has a positive value, but the result of enhancing a dark point (here, a black spot defect) is a minus value. Since the image format of image processing usually takes only a positive value, the result is replaced with 0, and there is a difficulty in detecting a dark defect (black spot). Therefore, by performing the convolution operation in the top hat filter process and adding the value of 2048 as the offset value, the emphasized portion of the scotoma which is originally a negative value is made positive, and as a result, only the top hat filter process is performed. Enables bright and dark point enhancement. Of course, the spatial filter is not limited to the top-hat filter, and the result of emphasizing the scotoma has a positive value (the sign is opposite to that of the top-hat filter) (see FIG. 6). May be used. Further, the filter size, weighting, and the like of these spatial filters are not limited to those shown in the drawings.
[0023]
FIG. 3F schematically shows a detected image 17 when the above-described top hat filter processing is performed on the reduced image (however, the image size is illustrated with the same size for easy understanding). . Actually it is a reduced size).
By simultaneously performing the top hat filter processing and the offset processing as described above, a result in which white spots and black spots are emphasized can be obtained at one time, and the detection of the spot defect 20 becomes easy.
Note that, in the case of a spot defect having a size in which the defect is not emphasized in FIG. 3F, the defect is emphasized by the same method in another reduced image which has been reduced in several stages to a size that can be emphasized. Will be done.
[0024]
(6) Statistical calculation (steps S14 to S18)
Next, statistical data is calculated based on the luminance value of each pixel in the detection image 17. In the calculation of the statistical data, the average value Lave, the standard deviation σ, the maximum value Lmax, and the minimum value Lmin of the luminance value of the entire image are obtained. From these four pieces of luminance statistical data, thresholds for white spots and black spots are determined, for example, as follows.
White spot threshold: Lave + a1 × σ
Black stain threshold: Lave-a2 × σ
Here, a1 and a2 are certain fixed constants.
Therefore, the threshold value for detecting white spots and black spots can be automatically determined based on the luminance statistical data by statistically calculating the luminance data in the detected image 17. Therefore, the threshold value is not artificial, trial and error, but objective and relative.
[0025]
(7) Check 1: Judgment of presence / absence of a defect candidate (step S19)
Then, a first check is performed to determine whether there is a candidate for a white spot defect or a black spot defect in the detected image 17 based on the threshold. That is, if the maximum value Lmax of the luminance value obtained by the calculation of the statistical data exceeds the white stain threshold, it is determined that there is a white stain defect in the detected image 17, and the minimum luminance value Lmin is equal to or less than the black stain threshold. If so, it is determined that there is a black spot defect in the detected image 17. If there are no defect candidates, the product is determined to be non-defective at this stage, and the inspection ends.
[0026]
(8) Blob processing (steps S20 to S24)
When it is determined in the above check 1 that there is a defect candidate in the detected image 17, the defect candidate is extracted using a threshold, and blob processing is performed to calculate the characteristic value of the defect candidate. The blob is a “cluster” having a value in a specific range existing in the image, and here is a defect candidate that is an area equal to or more than a white stain threshold or an area equal to or less than a black stain threshold. Therefore, a defect candidate is extracted by a binarization process using the white stain threshold and the black stain threshold, and the extracted region is subjected to a blob process, which is an image processing method, to calculate the characteristic value of the defect candidate. I do.
Here, as the characteristic values of the defect candidate, the barycentric position X and Y coordinates of the area, the maximum luminance value (Lmax (n)) in the area for a white spot defect, and the area value for a black spot defect The minimum value (Lmin (n)) of the luminance is calculated.
[0027]
(9) Check 2: Calculation of evaluation value (step S25)
In the second (ie, final) check, evaluation values are calculated for the defect candidates extracted as the defect candidates in order to quantify the degree of the defect. The evaluation value is the luminance statistical data (average value Lave (i), standard deviation σ (i), i is the screen number of the reduced image) obtained for each of the reduced images (described above) and the characteristics obtained by the blob processing. Using the values (the maximum value Lmax (n), the minimum value Lmin (n), and n is a blob number), the evaluation value is calculated by the following equation.
White stain:
Ev (n) = k (i) × (Lmax (n) −Lave (i)) / σ (i)
Black stain:
Ev (n) = k (i) × (Lave (i) −Lmin (n)) / σ (i)
Where k (i) = reduced screen coefficient i = screen number Lmax (n) = maximum luminance of defect candidate n = Blob number Lmin (n) = minimum luminance of defect candidate Lave = average luminance of entire image σ = luminance of entire image By calculating the evaluation value using these formulas, the white spot defect and the black spot defect extracted as defect candidates are quantitatively evaluated by objective data together with the coordinate position and the number (Blob number). It is possible to determine not only the presence or absence of a white spot defect and a black spot defect, but also to determine a rank based on the degree of the defect. Therefore, the detection accuracy of the stain defect is high.
[0028]
For this evaluation value, the threshold value for rank classification can be set in several stages for each good product rank. As a result, it is possible to assign a rank (grade) of the spot defect to a non-defective product and to grade the product. For example, when a liquid crystal panel is used as a light valve of a projector for each of R (red), G (green), and B (blue) colors, the relative luminous efficiency is best when green (wavelength λ = 555 nm). Since it is high, the threshold value for setting the rank of the green light valve is set to the strictest value smaller than in other cases.
[0029]
This embodiment is configured as described above, and automatically removes stain defects existing on the screen of a display device such as a liquid crystal panel regardless of the size of the defect and the contrast level. And stain defects can be individually and quantitatively evaluated.
In addition, since spot defects are evaluated based on luminance statistical data, collecting and analyzing the quality data of products and parts can be used for quality control, and a method aimed at further improving quality is developed. It is also possible to build.
[0030]
The present invention is not limited to a liquid crystal panel using the above-described TFT element, and displays such as a liquid crystal panel, a plasma display, an organic EL display, and a DMD (direct mirror device) using other diode elements. It can be used for inspection of body parts and display devices and products using them, and it is needless to say that the use of such components is not excluded from the scope of the present invention.
[Brief description of the drawings]
FIG. 1 is a configuration diagram of a stain defect detection device according to an embodiment of the present invention.
FIG. 2 is a flowchart showing a stain defect detection process.
FIG. 3 is a schematic diagram of each image from an input image to a stain defect detection image.
FIG. 4 is an explanatory diagram of a method for reducing an image size.
FIG. 5 is a diagram illustrating an example of a top hat filter.
FIG. 6 is a diagram showing an example of a well filter.
[Explanation of symbols]
Reference Signs List 1 projector, 2 liquid crystal panel, 3 screen, 4 images, 5 pattern generator, 6 CCD camera, 7 computer, 8 display device, 10 inspection target screen, 11 input image, 12 display area image, 13 original image (display area correction image) ), 14 background image, 15 inspection image (background difference image), 16 flattened image, 17 detected image, 20 spot defect, 21 spot defect, 31 screen edge

Claims (13)

Imaging a screen to be inspected;
From the image captured by the imaging, a step of taking a difference from a background image created in advance to create an inspection image,
Performing a flattening process of the inspection image;
A step of creating a plurality of reduced images from the flattened image,
Performing a filtering process for defect enhancement on each of the reduced images;
Calculating statistical data of the luminance value of each pixel in the image after the filter processing,
A step of determining a threshold based on the luminance statistical data and determining the presence or absence of a defect candidate;
Extracting a defect candidate from the image determined to have a defect candidate using the threshold,
A method for detecting a stain defect, comprising: 2. A method according to claim 1, wherein the data of the captured image has 12 bits or more than 4096 gradations. 3. The method according to claim 1, wherein the background image is obtained by averaging a plurality of images captured by the same optical system and the same imaging system. 4. 4. The method according to claim 1, wherein a top hat filter or a well filter is used in the filtering. The method according to claim 1, wherein the step of creating the inspection image includes a step of extracting a display area from an image including the screen to be inspected, and applying a geometric deformation to the display area to make the display area rectangular. 4. The method for detecting a stain defect according to any one of 4. 6. The method according to claim 1, wherein the threshold value for determining the presence or absence of the defect candidate is determined for each of the bright defect and the dark defect. 7. The method according to claim 6, wherein the threshold for determining the presence or absence of the defect candidate is determined using an average value and a standard deviation of the luminance statistical data. The method according to claim 1, further comprising the step of: performing a blob process for obtaining a characteristic value of the defect candidate on the extracted defect candidate; and calculating an evaluation value based on the characteristic value of the defect candidate obtained by the blob process and the luminance statistical data. The method for detecting a stain defect according to any one of claims 1 to 7, wherein: 9. The method according to claim 8, wherein the characteristic values obtained by the blob processing are a maximum luminance and a minimum luminance of the defect candidate. The method according to claim 1, wherein the evaluation value of the defect candidate is calculated for each of a light defect and a dark defect. 11. The detection of a stain defect according to claim 8, wherein the evaluation value of the defect candidate is calculated using an average value and a standard deviation of the luminance statistical data, and a maximum luminance and a minimum luminance of the defect candidate. Method. 12. The spot defect detection method according to claim 10, wherein a good product rank of the product is classified based on the magnitude of the evaluation value of the defect candidate. An apparatus for detecting a stain defect, comprising using the method for detecting a stain defect according to claim 1.

Images