JPH0949782A - Defect detector for color filter - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve the detecting reliability of a white defect or hole defect and to enhance the detecting sensitivity by providing a light source, a color CCD area sensor and an image processor. SOLUTION: A light source 2 is disposed at the rear side of a color filter 1 to illuminate a light 3a toward the rear surface of the filter, and the light 3b passed through the filter 1 is received by a color CCD area sensor 5 via an imaging lens 4. The imaging signal 6 by the sensor 5 is input to an image processor 7, and image processed. An R component extraction image is obtained from the signal 6 of the sensor 6, and binarized to a binary threshold value TR. When white points exist on the entire R, G and B transmitted parts of the filter 1, the white point in the R transmitted part cannot be extracted, but the R component can be extracted by the white point in the G and B transmitted parts. In this case, the threshold value TR is set so as to exceed the spectral characteristics and the intensity level of the G transmitted part to become suitable binary threshold value, and only G and B white points can be extracted.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a defect detection device applied in a color filter inspection device for liquid crystal display, for example.

[0002]

2. Description of the Related Art Conventionally, as a device for detecting a hole defect or a white spot defect of a black matrix of a color filter, light is emitted from a light source to the back side of the color filter which is an object to be inspected and the transmitted light of the color filter is detected. There is a method in which a monochrome CCD area sensor or CCD line sensor arranged on the front side of the color filter receives light, and the received light image is subjected to image processing to extract defects.

[0003]

However, the above-mentioned prior art has the following problems. (1) When a monochrome CCD area sensor is used, with respect to white spot defects, white spot defects are extracted by improving the contrast between the light transmitted through the normal part of the color filter and the monochrome light transmitted through the defective parts. , R for the threshold value for white spot defect detection, that is, for binarization related to image processing.
(Red), G (green), B (blue) It is difficult to individually optimize according to the color filter spectral characteristics of each color filter, and therefore, there is a limit to increase the reliability and detection sensitivity of white spot defect detection. .

This also applies to the detection of hole defects in the black matrix.
The light receiving sensitivity of the monochrome CCD area sensor is different for each of G and B. Therefore, the threshold value of the binarization related to the image processing for extracting the hole defect is different for each of R, G, and B. There is a limit to increase detection reliability and detection sensitivity.

(2) Further, when the CCD line sensor is used, the light receiving sensitivity is low, and therefore the inspection speed must be reduced.

In view of the above problems, therefore, it is an object of the present invention to provide a color filter defect detection device which can improve the reliability and detection sensitivity of defect detection and improve the detection speed without using a CCD line sensor. And

[0007]

The present invention, which achieves the above objects, has the following features. (1) A light source installed on one side of a color filter that is an object to be inspected, a color CCD area sensor that receives light transmitted through the color filter by the light source, and an R component extracted image of an image pickup signal from the sensor And an image processing apparatus that obtains an image product of the binarized image of B component extracted image and the binarized image of the B component extracted image. (2) A light source installed on one side of a color filter which is an object to be inspected, a color CCD area sensor for receiving the transmitted light of the color filter by the light source, and R, G, and An image processing device for obtaining the image sum of each binarized image of each of the B or R, B component extracted images and then obtaining the image product of this image sum and the reference black matrix inversion image.

A color CCD area sensor extracts R, B, or R, G, B component light and sets an appropriate binarization threshold value for the spectral characteristic of each filter. By obtaining the image product for the binarized image, and then obtaining the image product with the reference black matrix after obtaining the image sum, it is possible to increase the sensitivity and reliability of the white spot defect and the sensitivity of the hole defect. And can increase reliability.

[0009]

BEST MODE FOR CARRYING OUT THE INVENTION Here, embodiments of the present invention will be described. FIG. 1 shows the overall configuration. A light source 2 is arranged on the back surface side of a color filter 1 as an object to be inspected, light 3a is emitted toward the back surface of the filter, and light 3b transmitted through this filter 1 is shown. Is received by the color CCD area sensor 5 via the imaging lens 4. This color C
The image pickup signal 6 from the CD area sensor 5 is sent to the image processing device 7
Is input to and image processing is performed. In this case, the color filter 1 is driven and positioned by an X, Y table (not shown) in the arrow directions c, d shown in FIG. 1 or in the direction perpendicular to the drawing.

In the image processing by the image processing device 7,
It is premised on the spectral characteristics of the color filters R, G, B shown in FIG. 2, that is, as is clear from the two types of spectral characteristics shown in FIG. 2, B and R have almost no common wavelength range,
Further, B and G or G and R have a common wavelength range with each other, and there is a portion where the spectral characteristic curves overlap.

The image processing process will be described with reference to FIG.
Shows a diagram in which an R component extracted image is obtained from the image pickup signal 6 of the color CCD sensor 5 and binarized by a binarization threshold value T _R. R, G, B transmission of the color filter 1 is shown. When white spots exist in all the parts, (a) white spots in the R transmissive part cannot be extracted as in the image pattern in the above figure, but R components can be extracted in the white spots in the G and B transmissive parts. The luminance level of the −e ′ portion is (a) a pattern as shown in the figure below. In this case, the threshold value T _R is set to be lower than the brightness level of the R transmissive portion and higher than the brightness level of the G transmissive portion in terms of spectral characteristics, so that an appropriate binarization threshold value is obtained. Only the white dots of B can be extracted.

FIG. 3B shows a diagram in which the B component extracted image is obtained from the image pickup signal 6 and binarized by the binarization threshold value T _B. FIG. G like the pattern
White points cannot be extracted in the B transmission part, but B components can be extracted in the white points in the R transmission part, so that the brightness level of the ff ′ part has a pattern as shown in the following diagram (b). In this case, 2
The threshold value T _B is set to be lower than the brightness levels of the G transparent portion and the B transparent portion. In the B component extracted image, even if the threshold value T _B is lowered, the luminance level of the R transmissive portion is almost zero due to the spectral characteristics shown in FIG.

In this way, by taking the AND of the image patterns shown in FIGS. 3A and 3B, only the white spot defect can be extracted as shown in FIG. 3C.

As described above, since the color CCD area sensor and the color image processing are used for the detection of the white spot defect, the binary values corresponding to the spectral characteristics of the R (red), G (green) and B (blue) filters are used. It is possible to optimize the setting of the threshold for activation, and the reliability of the white spot defect detection sensitivity is improved. Furthermore,
An image product operation is performed in the color image processing to obtain R, G, B
Of the black matrix, that is, regardless of the shape of the opening of the black matrix, a binarized image having only the white spot defect as the bright portion can be obtained. As a result, it is possible to quickly and easily determine the presence or absence of the white spot defect, the size, and the like, which contributes to the quality control of the color filter and the yield improvement.

Next, detection of hole defects on the black matrix will be described. Since FIG. 1 and FIG. 2 are the same, description is omitted. In FIG. 4, when all the R, G, B parts in the black matrix have hole defects, the R, B component extracted image is obtained from the image pickup signal 6 of the color CCD sensor 5 and binarization processing is performed by the threshold value. It is shown that. That is, in the R component extracted image of FIG. 4A, only the hole defects of the R transmission part and the black matrix part are set as the bright part and G and B
A binarized image pattern in which the transparent portion and the R, G, B black matrix portion are dark portions is obtained. This is set as the binarization threshold value T _R of the R component extracted image so as to be lower than the brightness level of the R transmission part and higher than the brightness level of the G transmission part. Then the middle diagram,
This is due to the brightness level pattern shown in the bottom diagram.

Regarding the B component extracted image of FIG. 4B,
As shown in the uppermost figure, only the hole defects of the G and B transmission parts and this black matrix part are the bright parts, and the R transmission part and R,
A binarized image pattern in which the G and B black matrix areas are dark areas is obtained. This is set as the binarization threshold value T _B of the B component extracted image so as to be lower than the brightness level of the G and B transmission parts. This is because the brightness level pattern becomes. Further, in the B component extracted image, the luminance level of the R transmission part becomes almost zero due to the spectral characteristic shown in FIG.

In this way, by obtaining the image sum of FIGS. 4A and 4B, the transmissive portion and the hole defect shown in FIG. 4C can be obtained as the bright portion.

Next, although it becomes necessary to distinguish between the transparent portion and the hole defect, the image processing shown in FIG. 5 is performed. In FIG. 5, a reference pattern of a black matrix having no hole defect is shown in FIG. 5A, and a dark portion of the reference pattern in FIG. Get the reverse image pattern. Then, the AND (image product) of the reverse image pattern of FIG. 4C and the image pattern (d) of FIG. 4C is calculated to obtain the hole defect image shown in FIG.

In FIGS. 4 and 5, the R component extracted image and the B component are shown.
The component extraction image is used, but when the R, G, and B component extraction images are used, the threshold value T for the R component extraction image is used.
_R is required, and a threshold value T _B for the B component extracted image
In addition to the above, a threshold value T _G for the G component extracted image shown in FIG. 6 is required, and this threshold value T _G is lower than the brightness level of the G transmission part and higher than the brightness levels of the R and B transmission parts. Requires setting. In this case, the threshold value T _B of the B component extracted image shown in FIG. 4B is set to be lower than the brightness level of the B transmission part and higher than the brightness level of the G transmission part.

[0020] Incidentally, R component and G in the above description, distinctions between the B component, R component, G component, B component, respectively segmentation have been described per, R, also divided between the G component and the B component T _R , T _B can be extracted by setting the level.
However, since the common wavelength range of R and G is narrower than the common wavelength range of G and B due to the spectral characteristics, it becomes difficult to set the level of the threshold value T _R.

As described above, in detecting hole defects in the black matrix, since the binarization threshold value is set by color image processing according to the spectral characteristics of the R, G, and B filters on the hole defects, the hole defect detection sensitivity. And reliability is increased. Further, the image sum, the image product, and the like are calculated by color image processing to obtain a binarized image in which only the hole defect is the bright part regardless of the transmissive part. As a result, it becomes easy to determine the presence or absence of hole defects and the quality of the size, which contributes to the quality control of the color filters and the improvement of the yield.

[0022]

As described above, according to the present invention, the detection reliability of white defects and hole defects can be improved and the detection sensitivity can be increased, and the detection speed can be improved without using a CCD line sensor. It is possible to improve.

[Brief description of drawings]

FIG. 1 is a schematic configuration diagram of a defect detection device.

FIG. 2 is a spectral characteristic diagram of a color filter.

FIG. 3 is an explanatory diagram of an image processing process of a white spot defect.

FIG. 4 is an explanatory diagram of the first half of an image processing process of a hole defect.

FIG. 5 is an explanatory diagram of the latter half of the image processing process of hole defects.

FIG. 6 is an explanatory diagram of a modified example of the image processing process of hole defects.

[Explanation of symbols]

1 Color Filter 2 Light Source 5 Color CCD Area Sensor 6 Imaging Signal 7 Image Processing Device T _R , T _B , T _G Threshold

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁶ Identification code Internal reference number FI technical display location // G02B 5/20 101 G06F 15/62 400

Claims (2)

[Claims] 1. A light source installed on one side of a color filter, which is an object to be inspected, a color CCD area sensor for receiving transmitted light of the color filter by the light source, and an R component of an image pickup signal from the sensor. An image processing apparatus that obtains an image product of a binarized image of an extracted image and a binarized image of a B component extracted image. 2. A light source installed on one side of a color filter which is an object to be inspected, a color CCD area sensor for receiving the transmitted light of the color filter by the light source, and R, among the image pickup signals from the sensor, G, B or R,
B: A defect detecting device for a color filter, comprising: an image processing device that obtains an image sum of each binarized image of each component extracted image and then obtains an image product of this image sum and a reference black matrix inversion image.