JP2006145228A - Unevenness defect detecting method and unevenness defect detector - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an unevenness defect detecting method capable of detecting an unevenness defects with high precision, and an unevenness defect detector. <P>SOLUTION: The unevenness defect detecting method has an unevenness component emphasizing processing process for applying an unevenness component emphasizing filter to each of the pixels of a photographed image to emphasizing unevenness components and an unevenness defect detecting process for detecting the uneven defect on the basis of the unevenness components emphasized value of each of the pixels obtained in the unevenness component emphasizing processing process. In the unevenness component emphasizing processing process, the luminance-comparing pixels separated from a target pixel by a predetermined distance and arranged to the periphery of the target pixel are rearranged, in the order of the magnitudes of the luminance value of the respective luminance comparing pixels to calculate the luminance average value of a predetermined number of pixels positioned at a central part and the unevenness component emphasizing filter for calculating the difference between the luminance average value and the luminance value of the target pixel to set the unevenness components emphasized value of the target pixel is used to emphasize the unevenness components. In the unevenness defect detecting process, the uneven defect is detected by comparing the unevenness components emphasized value of each of the pixels with a predetermined threshold to extract the pixel of the unevenness defect candidates. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a mura defect detection method and apparatus for automatically detecting mura defects with high accuracy in an inspection process of various products such as an inspection process in the manufacture of a display device such as a liquid crystal panel and its application product such as a projector.

  Surface system defects in the inspection of LCD panels such as TFT panels are difficult to detect automatically with an inspection apparatus because the shape is indefinite and the contrast is low. For this reason, the inspection is still performed by the inspector's visual inspection, and automation of the inspection is urgently required to reduce the manufacturing cost.

Such plane defects are roughly classified into four defects, namely, a spot, a streak, a nonuniformity, and a line.
Among these defects, as a method for automating the detection of the mura defect, a method of detecting the mura by performing an enhancement process using a secondary differential filter has already been proposed (see, for example, Patent Documents 1 to 3).

JP 2004-53477 A JP 2003-14580 A JP 2000-111492 A

However, in Patent Documents 1 and 3, only the vertical and horizontal components on the screen are considered in the second-order differential filter, and the oblique components are not taken into account. There was a problem that it was difficult to improve.
Further, in Patent Document 2, not only the vertical direction and the horizontal direction but also the oblique direction is considered. However, since the filter is set so as to react to a portion where the luminance change is large, it reacts also to the edge portion. As a result, a non-uniform portion is also detected at the same time, so that there is a problem that the detection accuracy of the non-uniform defect is lowered.

  The present invention has been made in view of the above-described problems, and an object thereof is to provide a mura defect detection method and apparatus capable of detecting a mura defect with high accuracy.

  The mura defect detection method of the present invention includes a mura component emphasis processing step of emphasizing a mura component by applying a mura component emphasis filter to each pixel of a captured image, and each pixel obtained in the mura component emphasis processing step. A mura defect detection step for detecting a mura defect based on the mura component enhancement value, wherein the mura component enhancement process step is separated from the target pixel selected in each pixel of the captured image by a predetermined distance and around the target pixel. When the luminance comparison pixels arranged in are rearranged in the order of the luminance value of each luminance comparison pixel, the luminance average value of a predetermined number of pixels located in the central portion is obtained, and this luminance average value and the target pixel A mura component is emphasized using a mura component enhancement filter that calculates a difference from a luminance value to obtain a mura component enhancement value of a target pixel, and the mura defect detection step uses the mura component enhancement value of each pixel as a predetermined threshold value. Compared with And detecting the extract and nonuniformity defect pixel unevenness defect candidate.

In the present invention, in the mura component enhancement processing step, the mura component enhancement filter that emphasizes the mura component based on the luminance value of the luminance comparison pixel arranged at a predetermined distance from the target pixel and around the target image is used. Since the processing is performed, for example, it is possible to surely detect unevenness spreading in a planar shape as compared with the case where the luminance comparison pixels are arranged only in the vertical and horizontal directions with respect to the target pixel.
In addition, when a non-uniformity emphasis filter that is calculated based on the difference between the average value of all the luminance comparison pixels and the luminance value of the target pixel is used, a black spot component that does not actually exist is generated around the defect (edge portion). There is a possibility that the spot portion of the lever is erroneously detected. On the other hand, the unevenness component enhancement filter of the present invention, when the luminance comparison pixels are sorted by the luminance value size, the luminance value average value of a predetermined number of pixels located in the central portion, the luminance value of the target pixel, Since the emphasis value in which the unevenness component is emphasized is calculated based on the difference between them, the unevenness component can be detected with high accuracy without detecting a spot that does not actually exist.
Note that the predetermined number of luminance comparison pixels located in the central portion when rearranged in order of luminance value is a predetermined value including a median value (median value) when a plurality of luminance comparison pixels are sorted in luminance value order. It means a number of luminance comparison pixels. Here, the predetermined number only needs to be smaller than the number of luminance comparison pixels, and when the number of luminance comparison pixels is large, it may be 5 pixels or more, but is preferably about 2 to 4 pixels.

In the mura defect detection method of the present invention, it is preferable that after the mura component enhancement processing step, noise is removed by applying a median filter, and then the mura defect detection step is executed.
As the median filter, for example, a median filter may be used in which the median value (median value) of each luminance value of 9 pixels of 3 × 3 is used as the luminance value of the target pixel located at the center of 3 × 3.
By using such a median filter, noise other than the uneven component can be removed, and the uneven component can be detected with higher accuracy.

In the mura defect detection method of the present invention, before the mura component enhancement processing step, a background image difference processing step of taking a difference between an imaging step of imaging an inspection object and a background image created in advance from the captured image A display area extracting step for extracting a display area from the image, and a reduced image creating step for creating a reduced image with respect to the extracted image, and the unevenness component enhancement processing step is performed by the reduced image creating step. It is preferable to apply a mura component enhancement filter to the reduced image created in step (1).
Here, the background image is prepared in order to remove the influence of unevenness other than the defect generated in the inspection object, and for example, an average of about 20 non-defective sample images is used. .
In the present invention, the background image difference processing step and the display area extraction step can eliminate the influence other than the inspection target, for example, the influence of non-inspection such as illumination or a lens used for imaging other than the inspection target. And uneven defects can be detected effectively. In addition, since a reduced image can be created in the reduced image creation process, the relative unevenness defect size with respect to the unevenness component enhancement filter can be changed, and even without preparing multiple types of filters according to the unevenness defect size, Uneven defects of various sizes can be detected.

In the mura defect detection method of the present invention, the mura defect detection step extracts a mura defect candidate pixel by comparing a mura component enhancement value of each pixel obtained in the mura component enhancement processing step with a predetermined threshold. An extraction processing step, a blob processing step for performing blob processing for obtaining the characteristic value of the mura defect candidate, and a defect rank classification for classifying the mura defect rank based on the characteristic value of the mura defect candidate obtained by the blob processing step. It is preferable to provide a process.
Since the pixel of the mura defect candidate is extracted by comparing the mura component enhancement value with a predetermined threshold, the defect candidate can be easily determined in a short time. Further, since the mura defect is evaluated based on the extracted mura defect candidate information, the defect can be objectively evaluated and ranked, so that a defective product can be easily determined.

In the mura defect detection method of the present invention, the mura component enhancement processing step is obtained by applying a plurality of types of mura component enhancement filters having different distances from the target pixel to the luminance comparison pixel to each pixel, and obtaining each filter. Of the unevenness component enhancement values, it is preferable that the maximum absolute value is the unevenness component enhancement value of the target pixel.
The uneven component enhancement filter can set the size of the detectable uneven defect by appropriately setting the distance between the target pixel and the luminance comparison pixel, and therefore can easily detect the uneven defect of the size to be detected. In addition, since the maximum absolute value of the mura component enhancement value is extracted, the mura defect of each size can be reliably detected when the enhancement results of the mura component enhancement filters are combined. For this reason, it is not necessary to individually evaluate the emphasis result of each mura component emphasis filter, and only the synthesis result needs to be evaluated, so that mura defects of various sizes can be easily detected and evaluated.

In the mura defect detection method of the present invention, before the mura component enhancement processing step, a reduced image creation step of creating a reduced image of a plurality of stages of the captured image is performed, and the mura component enhancement processing step is a reduced image creation step. Applying the mura component enhancement filter to each reduced image created, and among the mura component enhancement values of each pixel in each reduced image obtained by each filter, the mura component of the pixel having the maximum absolute value The emphasis value is preferable.
If a plurality of stages of reduced images are created, the same unevenness defect size can be made different. For this reason, even if one kind of unevenness component enhancement filter is used, unevenness defects of various sizes can be detected by applying to a reduced image of a plurality of stages.

  The mura defect detection apparatus according to the present invention includes a mura component enhancement processing unit that applies a mura component enhancement filter to each pixel of a captured image to emphasize the mura component, and each pixel obtained by the mura component enhancement processing unit. Mura defect detecting means for detecting a mura defect based on the mura component enhancement value, and the mura component enhancement processing means is arranged at a predetermined distance from the target pixel selected in the captured image and around the target pixel. When the luminance comparison pixels are rearranged in the order of the luminance value of each luminance comparison pixel, a luminance average value of a predetermined number of pixels located in the central portion is obtained, and the luminance average value and the luminance value of the target pixel are calculated. The mura component is emphasized using a mura component emphasis filter that calculates the difference between the mura component and the mura component enhancement filter of the target pixel, and the mura defect detection means compares the mura component enhancement value of each pixel with a predetermined threshold value. Missing Extracts pixels of the candidate and detecting the mura defect.

  In the present invention, when the luminance comparison pixels arranged at a predetermined distance from the target pixel and around the target image are sorted by the size of the luminance value in the unevenness component enhancement processing means, the predetermined value of the median value portion is determined. Since the processing is performed using the mura component enhancement filter that obtains the mura component enhancement value based on the difference between the average value of the number of luminance values and the luminance value of the target pixel, it is possible to reliably detect unevenness spreading in a planar shape. Therefore, it is possible to detect a mura defect with high accuracy without erroneously detecting a black spot component that does not actually exist around the defect (edge portion).

In the mura defect detection apparatus of the present invention, an imaging unit that images an inspection object, a background image difference processing unit that takes a difference between a background image created in advance from the captured image, and a display area extraction from the image Display area extraction means for performing reduction image creation means for creating a reduced image, and the unevenness component enhancement processing means applies a unevenness component enhancement filter to the reduced image created by the reduced image creation means. preferable.
In the present invention, the background image difference processing means and the display area extracting means can eliminate the influence other than the inspection object, for example, the influence of the unevenness caused by things other than the inspection object such as the illumination and the lens used for imaging. And uneven defects can be detected effectively. In addition, since a reduced image can be created in the reduced image creation process, the relative unevenness defect size with respect to the unevenness component enhancement filter can be changed, and even without preparing multiple types of filters according to the unevenness defect size, Uneven defects of various sizes can be detected.

FIG. 1 is a block diagram showing a configuration of a mura defect detecting apparatus according to an embodiment of the present invention.
In FIG. 1, reference numeral 1 denotes a liquid crystal panel (liquid crystal light valve) to be inspected, and reference numeral 2 denotes a projector which is an image projection device, which can set the liquid crystal panel 1 from the outside. Reference numeral 3 denotes a pattern generator that is a pattern generation device that outputs various patterns to the liquid crystal panel 1. Reference numeral 4 denotes a screen. Reference numeral 5 denotes a CCD camera that is an image pickup unit that captures an image projected on the screen 4. It has a CCD with a resolution of. Reference numeral 6 denotes a computer device which is an image processing means for controlling the pattern generator 3 and the CCD camera 5 to detect uneven defects in the liquid crystal panel 1, and 7 is a display device connected to the computer device 6.

  The computer device 6 includes an image input unit 60, a background image difference processing unit 61, a display area extraction unit 62, a reduced image creation unit 63, a mura component enhancement processing unit 64, a noise removal unit 65, and a mura defect extraction. The processing unit 66, the blob processing unit 67, and the defect rank classification processing unit 68 are configured.

Image data of the captured image captured by the CCD camera 5 is input to the image input means 60 of the computer device 6. The captured image is stored in a storage means (not shown). Accordingly, an imaging process (image input process) is performed in which the inspection target is imaged using the CCD camera 5 by the image input means 60.
The background image difference processing means 61 carries out a background image difference processing step for obtaining a background difference image in which a difference between an input image and a background image created in advance is removed and defective luminance changes caused by things other than the inspection target are removed. To do. The display area extracting means 62 performs a display area extracting step of extracting only the image portion of the part to be inspected from the background difference image.

  The reduced image creating unit 63 performs a reduced image creating step of creating a reduced image from the image extracted by the display area extracting unit 62. For example, the reduced image creating means 63 performs a process of reducing the captured image to 1/8 size in order to detect a relatively large unevenness defect in the captured image. As will be described later, the unevenness emphasis filter used by the unevenness component emphasis processing means 64 has a certain degree of unevenness that can be emphasized. For this reason, even if the same unevenness component enhancement filter is used, a relatively large unevenness defect is detected by reducing the size of the unevenness component by reducing the size of the image itself as compared with the case where the image is not reduced. be able to. When detecting a small unevenness defect, the reduced image creating step by the reduced image creating means 63 does not have to be performed.

The mura component enhancement processing means 64 performs a mura component enhancement process step of applying a mura component enhancement filter (mura defect detection filter) to an image to emphasize and detect mura defects.
The noise removing unit 65 performs a noise removing step of removing noise by applying a median filter to the result obtained by the unevenness component enhancement processing unit 64.

  The mura defect extraction processing unit 66 extracts a mura defect candidate by comparing the result processed by the noise removing unit 65 with a predetermined threshold. Note that the mura defect includes a white mura defect having a high luminance value with respect to other pixel portions and a black mura defect having a low luminance value. For this reason, as the threshold value, a white unevenness defect threshold value and a black unevenness defect threshold value are set, a white unevenness defect candidate is extracted by comparing with the white unevenness defect threshold value, and compared with the black unevenness defect threshold value. A defect candidate is extracted.

The blob processing means 67 performs a blob processing step for obtaining the area of the region extracted as the defect candidate and the average luminance.
The defect rank classification processing unit 68 evaluates the rank of the mura defect based on the area and average luminance obtained by the blob processing unit 67, and classifies which defect rank the current inspection object corresponds to. carry out.
Therefore, in this embodiment, the mura defect detection means 66, the blob processing means 67, and the defect rank classification processing means 68 constitute a mura defect detection means.

Next, the operation of the mura defect detection apparatus according to the embodiment of the present invention will be described.
FIG. 2 is a flowchart for explaining the operation of the mura defect detection apparatus of this embodiment. The operation shown in FIG. 2 is realized by a program executed on the computer device 6.
First, the liquid crystal panel 1 to be inspected is set in the projector 2, the pattern generator 3 is controlled by the computer device 6 to display a specific brightness pattern on the liquid crystal panel 1, and this is projected onto the screen 4 by the projector 2. To do. Then, the image projected on the screen 4 is photographed by the CCD camera 5, the image of the photographed data is output to the computer device 6, and the mura defect detection processing is performed by the computer device 6, thereby detecting the mura defect of the liquid crystal panel 1. The result is displayed on the display device 7.

Here, the operation of detecting the mura defect by the computer device 6 will be described based on the flowchart of FIG.
First, an image projected on the screen 4 is photographed by the CCD camera 5, and an image of the photographed data is taken into the image input means 60 of the computer device 6, and an image input process (imaging process) is performed (step S1). . At this time, the photographing data is taken into the computer device 6 as 4096 gradation (12 bits) digital data by an A / D converter (not shown).
Next, the background image difference processing means 61 performs a background image difference processing step for removing defective luminance changes caused by things other than the liquid crystal panel such as illumination and lenses from the captured image data (step) S2).

  In this background image difference processing step, the background image shown in FIG. 3C is created by subtracting the background image shown in FIG. 3B from the input image (projected image) shown in FIG. The background image is an image of luminance change of the optical system excluding the liquid crystal panel 1. Since the luminance change on the defect due to the projection lamp and the projection lens occurs in both the input image and the background image, if the background image is subtracted from the input image, the background difference image is not a liquid crystal panel such as a projection lamp or a projection lens. The luminance change component on the defect caused by the object is removed.

Subsequently, the display area extracting means 62 performs a display area extracting process for extracting only the screen portion of the part to be inspected (step S3).
In the display area extraction processing step, the coordinates of the four corners of the inspected part image (background difference image) shown in FIG. 4A are subjected to pattern matching processing (four small regions of several tens of pixels by several tens of pixels near the four corners of the image data). The four corner reference images prepared in advance are subjected to pattern matching processing and the coordinates of the four corners are specified), and the display is performed by affine transformation so that the positional relationship of the coordinates of the four corners is rectangular. Extract area. As a result, as shown in FIG. 4B, the peripheral edge on the screen 4 is removed, and only the screen portion having an accurate rectangle is extracted.

Next, the reduced image creating means 63 performs a reduced image creating process for reducing the inspection target image that has undergone the display area extraction process (step S4).
In the reduced image creation processing step, an image reduced from the inspection target image to a predetermined size, for example, 1/8 size, is created. Specifically, when a reduced image of 1/8 size is created, a reduced image of 1/2 size is created by setting an average value of four pixels of the inspection target image to 1 pixel, and this 1/2 image A reduced image of 1/4 size is created by setting the average value of four pixels of the reduced image of 1 to 1 pixel, and 1/8 by setting the average value of 4 pixels of the reduced image of 1/4 to 1 pixel. Create a reduced-size image.

Therefore, if the flattened original image is 1200 × 1000 1.2 million pixels, 1 /
The reduced image of 2 has 300,000 pixels of 600 × 500, the reduced image of 1/4 has 75,000 pixels of 300 × 250, and the reduced image of 1/8 has 18750 pixels of 150 × 125. .
In this way, a reduced image of a predetermined size is created because a later-described mura component enhancement filter is designed to emphasize a mura defect of a predetermined size, and by reducing the image size, This is to make it possible to emphasize relatively large unevenness that cannot be emphasized by a later-described unevenness component emphasis filter by reducing the image size.

Next, the mura component enhancement processing means 64 performs a mura component enhancement processing step of enhancing the mura component on the created reduced image (step S5). This uneven component enhancement processing step emphasizes only the uneven defect in the image because it is difficult to detect a minute level of white uneven defect / black uneven defect.
In the case of detecting only a mura defect of a predetermined size, it is only necessary to apply one type of mura component enhancement filter to the generated reduced image. A plurality of types of mura component enhancement filters having different sizes of mura defects to be detected may be applied to detect mura defects of a predetermined size.

As shown in FIG. 5, the unevenness component emphasizing filter has a central pixel of a set detection area as a target pixel, is arranged in a substantially circular shape around the target pixel, and is arranged a predetermined length away from the target pixel. A luminance comparison pixel is provided.
Then, each luminance comparison pixel is sorted by its luminance value, and an average value of a predetermined number of pixels in the median value (median value) portion when sorted in order of the luminance value is obtained, and the luminance value of the target pixel The difference from the average value is calculated to calculate the unevenness component enhancement result value.
For example, in the unevenness component enhancement filter shown in FIG. 6, when the luminance value of the target pixel is O and the luminance values of the luminance comparison pixels are S1 to S32, the luminance values of S1 to S32 are sorted in the order of their sizes. If the result is d1 to d32, the median value (median value) is d16 and d17 because there are an even number. When two pixels are selected as a predetermined number of pixels in the central portion of the luminance comparison pixels S1 to S32 sorted in the order of the luminance values d1 to d32, the pixels having the luminance values d16 and S17 are selected, and four pixels Is selected, pixels having luminance values d15 to d18 are selected, and an average luminance value of these selected pixels may be obtained.

In the unevenness component enhancement filter, the number of luminance comparison pixels extracted in the central portion may be smaller than the total number of luminance comparison pixels, but is usually sufficiently small relative to the number of luminance comparison pixels. A value of about 2 to 4 pixels is preferable. However, it is not limited to this, and may be 5 pixels or more.
In the unevenness component enhancement filter shown in FIG. 6, the distance from the target pixel to the upper and lower, left and right luminance comparison pixels is six pixels, and the distance between the target pixel and the other luminance comparison pixels is also approximately six pixels. It is. The unevenness component enhancement filter of the present embodiment is suitable for extracting unevenness defects having an area that is slightly smaller than the area surrounded by the luminance comparison pixels. In other words, when the unevenness area to be measured is larger than the area in the area surrounded by the luminance comparison pixels, the unevenness component enhancement filter emphasizes only the contour portion and cannot emphasize the entire unevenness portion. . Accordingly, the size of the non-uniformity defect that can be detected varies depending on the distance between the target pixel and the luminance comparison pixel. For this reason, in the following description, in each mura component emphasis filter, a pixel whose distance between the target pixel and the luminance comparison pixel is n is called a distance n pixel filter. For example, the unevenness component enhancement filter shown in FIG. 6 is a 6-pixel distance filter.

Here, the process of the unevenness component emphasis process of the present embodiment will be described in comparison with a comparative example.
The results of applying the mura component enhancement filter of the present embodiment to the pseudo mura defect image shown in FIG. 7 are shown in FIGS. 8 to 15, and the results of applying the mura component enhancement filter of the comparative example are shown in FIGS.
The pseudo mura defect shown in FIG. 7 is a pseudo mura defect created in a pseudo manner, and the brightness of the background when the luminance of each pixel is set to 12-bit gradation in the image data of 1280 × 960 pixel size. No. 512 and white unevenness defect brightness of 1024. Further, the diameter of the white uneven defect portion is the length of 151 pixels.
Then, the image shown in FIG. 7 was reduced to 1/8 size, and the following unevenness component enhancement filters were applied to the reduced image. Accordingly, the diameter of the white uneven defect portion is 151/8 = about 19 pixels.

The unevenness component enhancement filter of the present embodiment stores the difference between the average value of the two pixels at the center of the comparison target pixel and the luminance value of the target pixel as the unevenness enhancement value of each target pixel. Three types of filters, a distance 3 pixel filter, a distance 6 pixel filter, and a distance 12 pixel filter, are prepared, and the results of the filters are synthesized and output.
On the other hand, the unevenness component enhancement filter of the comparative example stores the difference between the average value of all pixels of the comparison target pixel and the luminance value of the target pixel as the unevenness enhancement value of each target pixel. Then, four types of filters, a distance 3 pixel filter, a distance 6 pixel filter, a distance 12 pixel filter, and a distance 18 pixel filter, are prepared, and the results of the filters are synthesized and output. The reason why four types of filters are prepared in the comparative example is that only three types of filters have many erroneous detection portions in their combined outputs, and cannot be compared with the mura component enhancement filter of this embodiment.

FIGS. 8 and 9 show the enhancement results when a non-uniformity component enhancement filter (distance 3-pixel filter) having a distance of 3 pixels between the target pixel and the luminance comparison pixel is applied. FIG. 9 shows the enhancement result value of each pixel in the horizontal direction passing through the mura defect center point.
As shown in FIG. 9, since the distance three-pixel filter is smaller than the diameter of the white uneven defect, only the outline portion of the uneven defect is emphasized, but the target pixel and the luminance comparison pixel are both white in the central portion of the uneven defect. Since it exists in a nonuniformity part and a brightness | luminance difference does not arise, it will become the same emphasis result as the exterior of a nonuniformity defect, and cannot be emphasized.

FIGS. 10 and 11 show the enhancement results when a non-uniformity component enhancement filter (distance 6 pixel filter) having a distance of 6 pixels between the target pixel and the luminance comparison pixel is applied. FIG. 11 shows the enhancement result value of each pixel in the horizontal direction passing through the mura defect center point.
As shown in FIG. 11, since the distance 6-pixel filter is also smaller than the diameter of the white unevenness defect, the emphasized area is large compared to the distance 3-pixel filter, but only the contour portion of the unevenness defect is emphasized. Is the same as in the case of the distance 3 pixel filter.

12 and 13 show the enhancement results when a non-uniformity component enhancement filter (distance 12 pixel filter) having a distance of 12 pixels between the target pixel and the luminance comparison pixel is applied. FIG. 13 shows the enhancement result value of each pixel in the horizontal direction passing through the mura defect center point.
As shown in FIG. 13, the 12-pixel distance filter has a diameter of 24 pixels, which is slightly larger than the diameter of the white unevenness defect, so that the entire area of the unevenness defect can be emphasized.

14 and 15 show the result of combining the enhancement results of the filters of distances 3, 6, and 12 shown in FIGS. In the filter defect synthesis process, the enhancement result values of the pixels at the same position are compared in the enhancement results of the filters, and the value having the maximum absolute value among them is used as the synthesis result.
When the distance of the mura component enhancement filter is very large compared to the size of the mura defect, if the luminance comparison pixel is located in another mura region, the mura component cannot be detected correctly due to the mura component. For this reason, it is preferable to use a mura component enhancement filter that is slightly larger than the size of the mura defect to be detected, for example, a size that is larger by about 1 to 3 pixels.
Therefore, if the enhancement results of the filters of the distances 3, 6, and 12 pixels are synthesized, the unevenness component corresponding to each size of the distances 3, 6, and 12 can be enhanced and detected.

On the other hand, a case where a mura component enhancement filter that takes the difference between the average value of all the luminance comparison pixels and the luminance value of the target pixel is applied to the same pseudo mura defect image will be considered.
16 and 17 show the results of enhancement using the distance 3 pixel filter, FIGS. 18 and 19 show the results of enhancement using the distance 6 pixel filter, FIGS. 20 and 21 show the results of enhancement using the distance 12 pixel filter, and FIGS. Results of enhancement by the 18-pixel filter, FIGS. 24 and 25 show synthesis results of these filters.
It can be seen from any of the results that the black spot component that does not exist around the defect is output and the uneven component cannot be detected correctly.
On the other hand, as shown in FIGS. 8 to 15, the unevenness component enhancement filter according to the present embodiment, which takes the difference between the average value of a predetermined number of pixels in the central portion of the luminance comparison pixels and the luminance value of the target pixel, is applied. In this case, it is understood that the black spot component that does not exist around the defect is not output, and only the spot component can be emphasized with high accuracy.

Further, FIGS. 26 to 30 show 3 of the distance 6 pixel filter, the distance 24 pixel filter, and the distance 48 pixel filter of the present embodiment for the uneven defect image in which the white uneven defect and the black uneven defect are simulated. An example in which various types of uneven component enhancement filters are applied will be described.
That is, FIG. 26 is a mura defect image in which white mura defects and black mura defects are created in a pseudo manner. In this process, in order to detect a relatively large mura defect, as a pre-processing, the mura defect image is reduced to １ ／ size by the reduced image creating means 63 in advance, and the reduced image is processed as a target. A filter in which the distance between the pixel and the luminance comparison image was set to 6, 24, and 48 pixels was applied.
27 shows the result of applying the distance 6 pixel filter to the 1/8 reduced image. FIG. 28 shows the result of applying the distance 24 pixel filter to the 1/8 reduced image. FIG. 30 shows a result of combining three images obtained by applying the distance 6, 24, and 48 pixel filters into one sheet as a result of applying the distance 48 pixel filter to the / 8 reduced image. As described above, according to the unevenness component enhancement processing step (S5) using the unevenness component enhancement filter of this embodiment, white unevenness and black unevenness can be emphasized and detected.

  Next, the noise removing unit 65 performs a noise removal processing step of applying a median filter to the result of the unevenness component enhancement processing step to remove noise other than the unevenness component (S6). As the median filter, for example, a median filter may be used in which the median value (median value) of each luminance value of 9 pixels of 3 × 3 is used as the luminance value of the target pixel located at the center of 3 × 3.

Next, the mura defect extraction processing unit 66 sets a threshold value for cutting out the white mura defect and a threshold value for cutting out the black mura defect for the mura defect enhanced image from which the noise has been removed, and cuts out each mura defect candidate area. A nonuniformity defect extraction process is performed (S7). Here, each threshold value may be set to an optimum value according to the situation of the image. For example, the average value of the unevenness emphasis value (luminance value) of the unevenness defect emphasis image (composite image) and its standard deviation are obtained, and the threshold value is set by the following equation.
White spot defect threshold wslevel = avr + α ・ σ + β ₁
Black unevenness defect threshold bslevel = avr−α ・ σ + β ₂
Here, avr is an average value of the unevenness enhancement value of the composite image, σ is a standard deviation of the unevenness enhancement value of the composite image, and α, β ₁ , β ₂ are arbitrarily determined depending on the situation of the image to be inspected. The

  Here, the result of binarizing the composite image using the white unevenness defect threshold and the black unevenness defect threshold calculated with α set to 1 is shown in FIGS. FIG. 31 shows white spot defect candidates cut out with the white spot defect threshold, and FIG. 32 shows white spot defect candidates cut out with the black spot defect threshold. In both FIGS. 31 and 32, the region displayed in white is the region of the mura defect candidate that has been cut out.

Next, the blob processing means 67 performs a blob processing step for obtaining the area and average brightness of the region cut out as the defect candidate for the white uneven defect candidate image and the black uneven defect candidate image (S8).
Next, the defect rank classification processing unit 68 performs a defect rank classification step of classifying the rank of the mura defect in the image based on the calculated area and average luminance of the defect candidate region (S9).
The defect rank obtained by the defect rank classification processing means 68 is displayed on the display device 7, and the inspector can easily grasp the defect rank of the liquid crystal panel 1 to be inspected.
Therefore, in this embodiment, the mura defect detection process is configured by the mura defect extraction process S7, the blob process S8, and the defect rank classification process S9.

According to this embodiment, there are the following effects.
(1) The mura component enhancement processing unit 64 uses a mura component enhancement filter that emphasizes the mura component based on the luminance value of the luminance comparison pixels arranged at a predetermined distance from the target pixel and around the target image. Therefore, for example, as compared with the case where the luminance comparison pixels are arranged only in the upper, lower, left and right directions with respect to the target pixel, it is possible to reliably detect unevenness spreading in a planar shape.
In addition, when the luminance comparison pixels are sorted by the magnitude of the luminance value, the unevenness component enhancement filter calculates the difference between the luminance value average value of a predetermined number of pixels located in the median value portion and the luminance value of the target pixel. Since the emphasis value in which the unevenness component is emphasized is calculated, a black spot component that does not exist around the defect occurs, as in the case of calculating the difference between the average value of all the luminance comparison pixels and the luminance value of the target pixel. The spot portion is not erroneously detected, and the uneven component can be detected with high accuracy.

(2) Further, since the unevenness component enhancement filter can be set appropriately according to the distance between the target pixel and the luminance comparison pixel, the size of the unevenness defect that can be detected by the filter can be set. Can be easily detected. In addition, even when a plurality of unevenness component enhancement filters according to the size of each unevenness defect are applied, it is possible to easily detect unevenness defects of each size by evaluating only the combined result of the enhancement results.

(3) Both the white unevenness defect and the black unevenness defect can be emphasized by the same unevenness component emphasizing filter. Both mura defects and black mura defects can be easily detected. For this reason, various unevenness defects can be efficiently detected by a simple process.

(4) Since the rank of the extracted mura defect can be classified by the blob processing means 67 and the defect rank classification processing means 68, the objective ranking of the defects can be performed in a short time, and the inspector can detect the mura defect. The degree can be easily determined, and it can be easily determined in a short time whether the product is non-defective.

The present invention is not limited to the above embodiment.
For example, in the above-described embodiment, in order to cope with various mura defect sizes, a plurality of types of mura component enhancement filters having different distances between the target pixel and the luminance comparison pixel are used. It is also possible to detect unevenness components of various sizes by reducing the target image in a plurality of stages and applying predetermined unevenness component enhancement filters to the respective images having different reduced sizes.

In the embodiment, the threshold values for the white unevenness defect and the black unevenness defect are set so that both the white unevenness defect and the black unevenness defect can be detected. However, only one defect may be detected. For example, when it is sufficient to detect only the white unevenness defect, the processing may be performed without performing the comparison process with the black unevenness defect threshold.
In the present invention, the pre-processing of the background difference processing step S2 and the display area extraction step S3 is performed before the mura component emphasis processing step S5 by the mura component emphasis processing means 64, but these depend on the state of the photographed image. May not be implemented. Similarly, the reduced image creating step S4 by the reduced image creating means 63 may not be performed depending on the size of the detected unevenness. Further, the noise removal processing step S6 by the noise removing means 65 may not be performed if the image state is good.

  In the embodiment, the mura defect extraction process S7, the blob process S8, and the defect rank classification process S9 are performed as the mura defect detection process. However, even if the mura defect is detected by other methods / procedures. Good. In short, the mura defect detection process may be any process that can determine whether or not it corresponds to a mura defect based on the mura component emphasized in the mura component enhancement processing step S5.

  In the above embodiment, when a plurality of unevenness component enhancement filters are applied, the results of the filters are combined and determined. However, whether or not there is a unevenness defect is determined for each filter result. May be. However, it is possible to more efficiently detect the mura defect if it is determined based on the synthesis result.

The detection target of the mura defect is not limited to the liquid crystal light valve using the TFT element as described above, but a liquid crystal panel, plasma display, organic EL display, DMD (direct mirror mirror) using other diode elements. It can be used for inspection of display body parts such as devices) and display devices and products using them, and it goes without saying that even if used for these, they are not excluded from the scope of the present invention. Absent.
Furthermore, the present invention is not limited to the inspection of various display devices. For example, when there are uneven scratches on the case of a household electrical appliance or the body of a car, an image with these uneven defects can be obtained by imaging them. Since the unevenness can be detected, it can be applied to unevenness inspection such as surface coating of various products.
In short, the present invention can be used to detect a luminance unevenness defect or a color unevenness defect because it can detect any unevenness having a luminance difference from the surroundings.

The block block diagram of the nonuniformity defect detection apparatus of the screen by embodiment of this invention. The flowchart for demonstrating operation | movement of the nonuniformity defect detection apparatus. Explanatory drawing which shows the background image difference process of the nonuniformity defect detection apparatus. Explanatory drawing which shows the display area extraction process of the nonuniformity defect detection apparatus. The figure which shows the example of a nonuniformity component emphasis filter. The figure which shows the example of the luminance value in a nonuniformity component emphasis filter. The figure which shows the image which provided the pseudo | simulation nonuniformity defect. The figure which shows the result of applying a distance 3 pixel filter. The graph which shows the emphasis result value in each pixel of the horizontal direction which passes along the pseudo | simulation unevenness center in FIG. The figure which shows the result of applying a distance 6 pixel filter. The graph which shows the emphasis result value in each pixel of the horizontal direction which passes along the pseudo | simulation unevenness center in FIG. The figure which shows the result of applying a distance 12 pixel filter. The graph which shows the emphasis result value in each pixel of the horizontal direction which passes along the pseudo | simulation unevenness center in FIG. The figure which shows the state which synthesize | combined the application result of distance 3, 6, and 12 pixel filter. The graph which shows the emphasis result value in each pixel of the horizontal direction which passes along the pseudo | simulation unevenness center in FIG. The figure which shows the result of applying a distance 3 pixel filter in a comparative example. The graph which shows the emphasis result value in each pixel of the horizontal direction which passes along the pseudo | simulation unevenness center in FIG. The figure which shows the result of applying a distance 6 pixel filter in a comparative example. The graph which shows the emphasis result value in each pixel of the horizontal direction which passes along the pseudo | simulation unevenness center in FIG. The figure which shows the result of applying a distance 12 pixel filter in a comparative example. The graph which shows the emphasis result value in each pixel of the horizontal direction which passes along the pseudo | simulation unevenness center in FIG. The figure which shows the result of applying a distance 18 pixel filter in a comparative example. The graph which shows the emphasis result value in each pixel of the horizontal direction which passes along the pseudo | simulation unevenness center in FIG. The figure which shows the state which synthesize | combined the application result of distance 3, 6, 12, 18 pixel filter in a comparative example. The graph which shows the emphasis result value in each pixel of the horizontal direction which passes along the pseudo | simulation unevenness center in FIG. The figure which shows the image which provided the pseudo | simulation nonuniformity defect which consists of white nonuniformity and black nonuniformity. The figure which shows the result of having applied the distance 6 pixel filter to the 1/8 reduction image. The figure which shows the result of having applied the distance 24 pixel filter to the 1/8 reduction image. The figure which shows the result of having applied the distance 48 pixel filter to the 1/8 reduction image. The figure which shows the state which synthesize | combined the application result of distance 6,24,48 pixel filter. The figure which shows the white nonuniformity result candidate cut out the synthetic | combination result by the white nonuniformity defect threshold value. The figure which shows the black nonuniformity result candidate cut out the synthetic | combination result by the black nonuniformity defect threshold value.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 ... Liquid crystal light valve, 2 ... Projector, 3 ... Pattern generator, 4 ... Screen, 5 ... CCD camera, 6 ... Computer apparatus, 7 ... Display apparatus, 60 ... Image input means, 61 ... Background image difference processing means, 62 ... Display area extracting means 63 ... Reduced image creating means 64 ... Uneven component emphasis processing means 65 ... Noise removing means 66 ... Uneven defect extraction processing means 67 ... Blob processing means 68 ... Defect rank classification processing means

Claims (8)

A mura component enhancement processing step of emphasizing the mura component by applying a mura component enhancement filter to each pixel of the captured image;
A mura defect detection step of detecting a mura defect based on the mura component enhancement value of each pixel obtained in the mura component enhancement processing step;
In the unevenness component enhancement processing step, the luminance comparison pixels arranged at a predetermined distance from the target pixel selected in the captured image and around the target pixel are rearranged in the order of the luminance value of each luminance comparison pixel. In this case, a luminance average value of a predetermined number of pixels located in the central portion is obtained, and a difference between the luminance average value and the luminance value of the target pixel is calculated to use a mura component enhancement filter as a mura component enhancement value of the target pixel. Emphasizing unevenness,
The mura defect detection step includes detecting a mura defect by extracting a mura defect candidate pixel by comparing the mura component enhancement value of each pixel with a predetermined threshold value. In the mura defect detection method according to claim 1,
A mura defect detection method comprising: performing a mura defect detection step after removing noise by applying a median filter after the mura component enhancement processing step. In the mura defect detection method according to claim 1 or 2,
Before the uneven component enhancement processing step,
An imaging process for imaging the inspection object;
A background image difference processing step for taking a difference from a previously created background image from the captured image;
A display area extraction process for extracting a display area from an image;
A reduced image creation step of creating a reduced image with respect to the extracted image;
In the mura component enhancement processing step, a mura component enhancement filter is applied to the reduced image created in the reduced image creation step. In the mura defect detection method according to any one of claims 1 to 3,
The uneven defect detection step includes
A mura defect extraction processing step of extracting mura defect candidate pixels by comparing the mura component enhancement value of each pixel obtained in the mura component enhancement processing step with a predetermined threshold;
A blob processing step for performing a blob processing for obtaining a characteristic value of the mura defect candidate;
And a defect rank classification step of classifying the rank of the mura defect based on the characteristic value of the mura defect candidate obtained by the blob processing step. In the mura defect detection method according to any one of claims 1 to 4,
In the unevenness component enhancement processing step, multiple types of unevenness component enhancement filters having different distances from the target pixel to the luminance comparison pixel are applied to each pixel, and the absolute value of the unevenness component enhancement values obtained by each filter is A method for detecting a mura defect, wherein a value having a maximum value is used as a mura component emphasis value of the target pixel. In the mura defect detection method according to any one of claims 1 to 4,
Prior to the unevenness component enhancement processing step, a reduced image creation step for creating a reduced image of a plurality of stages of the captured image is performed,
The unevenness component enhancement processing step applies the unevenness component enhancement filter to each reduced image created in the reduced image creation step, and among the unevenness enhancement values of each pixel in each reduced image obtained by each filter, A method for detecting a mura defect, characterized in that a pixel having the maximum absolute value is used as a mura component emphasis value of the pixel. A non-uniform component emphasis processing means for emphasizing the non-uniform component by applying a non-uniform component emphasis filter to each pixel of the captured image;
Mura defect detection means for detecting a mura defect based on the mura component enhancement value of each pixel obtained by the mura component enhancement processing means,
The unevenness component enhancement processing means rearranges the luminance comparison pixels arranged around the target pixel at a predetermined distance from the target pixel selected in the captured image in the order of the luminance value of each luminance comparison pixel. In this case, a luminance average value of a predetermined number of pixels located in the central portion is obtained, and a difference between the luminance average value and the luminance value of the target pixel is calculated to use a mura component enhancement filter as a mura component enhancement value of the target pixel. Emphasizing unevenness,
The mura defect detection device detects a mura defect by extracting a mura defect candidate pixel by comparing the mura component enhancement value of each pixel with a predetermined threshold value. In the nonuniformity defect detection apparatus of Claim 7,
An imaging means for imaging the inspection object;
A background image difference processing means for taking a difference from a background image created in advance from the captured image;
Display area extraction means for extracting a display area from an image;
A reduced image creating means for creating a reduced image,
The mura component enhancement processing unit applies a mura component enhancement filter to the reduced image created by the reduced image creation unit.

Images