JPH07120400A - Detection of defect in color filter for viewfinder - Google Patents

Abstract

PURPOSE:To automatically inspect the outward-appearance defect such as contamination, flaw or the like on the whole front and back surface of a color filter by a method wherein an inspection region on the color filter is divided into a plurality of regions and video signals which have been imaged under exposure conditions corresponding to the individual regions, are image-processed. CONSTITUTION:An imaging camera 1 which is provided with an electronic shutter function and a CCD area sensor, and whose exposure time is variable, and a light source 2 in which a xenon light source has been made fiberlike, are used. Light from the light source 2 is reflected by a half mirror 6, and the whole face of a color filter 3 for a viewfinder, which is mounted on a holding plate 4 is illuminated. Then, by reflected light from the filter 3, the whole image of the filter 3 is formed and imaged on the CCD area sensor. The obtained image signal is image-processed by a device 5, and defect is detected. For example, a defective sample is placed on a holding plate 4, the exposure time which is optimum for a colored picture-element part and an outer- circumference light-shielding part is found, then a filter 3 is imaged by using the found exposure time, the obtained image signal is processed, and a surface defect is detected automatically.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for automatically inspecting a color filter for a viewfinder for appearance defects such as stains and scratches.

[0002]

2. Description of the Related Art Conventionally, as a visual inspection method for a color filter for a viewfinder, a human visual inspection method using a light projector and a microscopic inspection method for fine defects that are difficult to visually inspect have been adopted. . However, in recent years, miniaturization and mass production have been increasingly required for the viewfinder-color filter. With the progress of high definition, it is required to detect even finer defects,
As for conventional visual inspection using a floodlight or a microscope, there is a problem in differences and reproducibility depending on humans, so that an automated inspection method that is not manual is required. In addition, automated inspection methods have been required for mass production. However, in the viewfinder color filter, the entire front and back surfaces of the product are inspected, and the transmission and / or
Or, it is necessary to inspect defects with reflected illumination, and it is difficult to perform automatic visual inspection because the transmittance and / or reflectance of the colored pixel portion, the outer peripheral light shielding portion, and the outermost peripheral transparent portion are different. Was required.

[0003]

Under these circumstances, it is possible to inspect the entire front and back surfaces of a color filter for a viewfinder using transmitted illumination and / or reflected illumination.
Moreover, there has been a demand for an automatic visual inspection method capable of mass production. The present invention is intended to provide an automatic visual inspection method for a color filter for a viewfinder, and more specifically, to an automatic visual inspection method for a color filter for a viewfinder that has different transmittance and / or reflectance depending on a place (region). .

[0004]

A defect detection method for a color filter for a viewfinder according to the present invention is obtained by imaging the entire surface of a color filter for a viewfinder using transmitted illumination light and / or reflected illumination light. This is a method of automatically inspecting appearance defects such as dirt and scratches by performing image processing on the signal, and the inspection area of the color filter is divided into a plurality of areas according to the difference in transmittance and / or reflectance, and defect inspection of each area is performed. The image signal level of the image pickup means is adjusted according to the exposure condition corresponding to each area, the entire inspection area is imaged, and the defect is detected by the image signal obtained by the image pickup. The entire inspection area is inspected. Normally, since the image is taken under the exposure condition corresponding to each area, the entire inspection area is imaged under a plurality of different exposure conditions, and the defect of each area is detected from the video signal obtained under each condition, and The entire inspection area is inspected. It is necessary to detect appearance defects such as stains and scratches on the viewfinder color filter on the entire front and back surfaces. Normally, the viewfinder color filter has a different transmittance and / or reflectance depending on the location (part). Since the difference is different, if the signals obtained by imaging are image-processed under the same conditions on the entire front and back surfaces, the video signal level obtained by imaging will have a large difference depending on the location (region), and appropriate defect detection will be performed. I can't. Usually, the colored portion surface of the color filter for the viewfinder has low reflectance, the outer peripheral light shielding portion has high reflectance, the colored portion and outer light shielding portion have low light transmittance, and the outermost peripheral transparent portion has high light transmittance. Therefore, the area (area) on the front and back of the color filter for the viewfinder.
The image signal level of the image pickup means is adjusted so that the image signal level becomes an appropriate level for defect detection by the photoelectric conversion function of the image pickup exposure time variable of the image pickup means. Whether to adjust the exposure time, or to adjust the illumination light intensity of the transmitted light and / or the reflected light from the color filter for the viewfinder,
Alternatively, adjust both by using both. The entire inspection area is exposed and imaged corresponding to each area, but from the video signal corresponding to the specific area, which is obtained under the exposure conditions, a signal of a level suitable for detecting defects in the specific area is extracted. However, the signal components in other regions are processed so that they can be removed or eliminated as necessary. For example, when capturing an image for defect detection, the intensity of reflected illumination light is increased in a colored portion having a low reflectance, the video signal level is increased, and the intensity of reflected illumination is reduced in an outer peripheral light-shielding portion having a high reflectance, Lower the video signal level,
The video signal level is adjusted to capture an image. Also, when imaging for defect detection, the transmitted light intensity is increased in the low transmittance colored portion and the outer peripheral light-shielding portion to increase the video signal level, and the transmitted light intensity is increased in the outermost transparent portion of high transmittance. Is small,
The video signal level is lowered and the video signal level is adjusted to capture an image. Further, for example, when imaging for defect detection, the image exposure level is increased and the photoelectric conversion is performed for the colored portion having a low reflectance to increase the image signal level, and the peripheral light shielding portion having a high reflectance is provided. In (1), the imaging exposure time is shortened and photoelectric conversion is performed to lower the video signal level, and the video signal level is adjusted for imaging. Then, at the time of image pickup for defect detection, the image exposure level is increased in the colored portion and the outer peripheral light-shielding portion having a low transmittance, and the photoelectric conversion is performed to increase the image signal level, and the outermost periphery having a high transmittance is obtained. The exposure time is shortened and photoelectric conversion is performed on the transparent portion to lower the video signal level and adjust the video signal level for imaging.

[0005]

The defect detecting method of the color filter for the viewfinder of the present invention is obtained by dividing the inspection area of the color filter into a plurality of areas, and detecting the defect of each area by imaging under the exposure condition corresponding to each area. Since the video signal is processed by image processing, the exposure condition can be adjusted and selected according to the light transmittance and / or light reflection of the area so that the defect can be detected easily. The defect detection in that area is ensured. Then, the entire inspection area is imaged under the exposure condition corresponding to each area, the defect of each inspection area is detected from the video signal obtained under each condition, and the defect of the entire inspection area is also detected. It is possible to detect defects on the entire surface of the color filter. The video signal level can be adjusted relatively easily by adjusting the photoelectric exposure function of the image pickup means by changing the image pickup exposure time, adjusting the illumination light intensity of transmitted light and / or reflected light, or adjusting both in combination. It is possible to automate the appearance defect inspection such as stains and scratches on the entire front and back surfaces of a color filter for a viewfinder having different transmittances and / or reflectances depending on places (regions).

[0006]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A first embodiment of a color filter defect detecting method for a viewfinder according to the present invention will be described below with reference to the drawings. FIG. 1 is a schematic view of an apparatus for detecting defects in a color filter for a viewfinder according to the first embodiment. The color filter for the viewfinder is a CCD whose exposure time is variable.
It is the figure which showed the apparatus which uses the area sensor and performs the automatic inspection by reflecting and illuminating the surface of the sample. In the figure, 1 is an imaging camera, 2 is a light source, 3 is a color filter for a viewfinder, 4 is a holding plate, 5 is an image processing device, 6 is a half mirror, and the imaging camera 1 is a CCD area sensor having an electronic shutter function. (C3077-50 manufactured by Hamamatsu Photonics KK), and the exposure time is variable. As the light source, a fiber-shaped xenon light source was used.
The light from the light source 2 is reflected by the half mirror 6, and the holding plate 4
The entire surface of the sample viewfinder color filter 3 placed on the above is illuminated. The entire surface of the image of the color filter 3 is imaged by the lens on the CCD area sensor of the imaging camera 1 by the reflected light from the sample viewfinder color filter 3. Then, the video signal obtained by the image pickup is subjected to image processing to detect a defect.
Here, an area sensor of 512 pixels × 480 pixels is used, and 40 μm is provided in the central portion on the substrate of the color filter for the viewfinder of 15 mm × 13 mm shown in FIG.
A color filter for a viewfinder in which R, G, and B colors were repeatedly arranged at a pitch of m × 20 μm was used as a sample, and the entire surface of the sample was photographed through a lens. First, prior to the defect detection of the viewfinder color filter, since the viewfinder color filter has a different reflectance between the colored pixel portion and the outer peripheral light-shielding portion, a predetermined defect sample is placed on the holding plate 4 and the colored pixel is placed. Department,
The exposure time of the electronic shutter is adjusted so that the appropriate defect can be detected at each position of the outer peripheral light-shielding portion, and the appropriate exposure time at each position is 1/60 sec or 1/1200 s.
ec was obtained. Then, the color filter 3 for the viewfinder, which is a sample, is placed on the holding plate 4 by the imaging camera 1, and the entire surface of the color filter 3 for the viewfinder is exposed at an exposure time 1 / Images were taken for 60 sec and an exposure time of 1/1200 sec that is appropriate for detecting defects in the outer peripheral light-shielding portion, and image signals obtained by imaging at each exposure time were subjected to image processing, and defects were detected for each image processing. . A defect due to reflected illumination of the colored part is detected from the image pickup at an exposure time of 1/60 sec suitable for detecting a defect of the colored pixel part, and an exposure time of 1/1200 sec suitable for detecting a defect of the outer peripheral light-shielding part.
A defect in the light-shielding portion on the outer periphery was detected from the image pickup at.

Next, a second embodiment will be described. Second Embodiment
Defect Detection for Color Filter for Viewfinder in Japan
Fig. 2 is a schematic view of the device when it is taken out.
Using a CCD filter and a Rahl filter,
A device that illuminates from the back side and performs automatic inspection by the transmitted light of the sample.
It is the figure which showed the arrangement. In the figure, 11 is an imaging camera, 12 is light
Source, 13 is a color filter for viewfinder, 1
Reference numeral 4 is a transparent holding plate, and 15 is an image processing apparatus.
The entire surface of 3 is imaged on the CCD area sensor by the lens.
The light source 12 for illuminating the sample is
The illumination light source intensity is variable, and it is a color for the viewfinder.
-The entire surface of the filter 13 is illuminated by transmission. Incidentally, CC
The same D area sensor and light source as in Example 1 are used.
It was The light from the light source 12 is placed on the transparent holding plate 14.
Of the sample-viewfinder color filter 13
Illuminate the entire surface from the back. Sample viewfinder
Color filter by the transmitted light from the color filter 13 for
The entire surface of the image of the Luther 13 is C of the imaging camera 11 by the lens
An image is formed and imaged on the CD area sensor. And
Image signals obtained by image processing are image processed and defects are detected.
To be done. First, the color filter for the viewfinder
Prior to detecting defects in the
Luther has a high transmittance, such as a colored pixel portion with a low transmittance and an outer peripheral light-shielding portion.
In the outermost transparent part of the transmittance, the transmittance of the illumination light varies greatly.
Therefore, a predetermined defect sample is mounted on the transparent holding plate 14.
The colored pixel portion, the outer peripheral light-shielding portion, and the outermost transparent portion.
In order to detect defects properly at each place, the light source 13
Adjust the illumination light intensity of the
Intensity brightness 2 million cd / m²50,000 cd / m ²Got
Next, the imaging camera 11 is used to
On the transparent holding plate 14
On the color filter 13 for the viewfinder.
Colored pixels with low transmittance when transilluminating the entire surface from the back
Brightness of 2 millionc, which is suitable for detecting defects in light-shielding parts and outer light-shielding parts
d / m², And the illumination brightness suitable for detecting defects in the outermost transparent part
50,000 cd / m²At different lighting intensities.
Image processing of the video signals obtained by
Defects were detected for each image processing. Defects in the above colored pixel section
Illumination brightness suitable for detection 2 million cd / m²From imaging at
Indicates that a defect due to transmitted illumination in the colored area or the light-shielding area on the outer periphery is detected.
Therefore, the illumination brightness is 50,000 cd, which is suitable for detecting defects in the outermost transparent area.
/ M²A defect in the outermost transparent part was detected from the
It was

In the first and second embodiments described above, a method of excluding areas other than the predetermined area and extracting only defects in the predetermined area, that is, a method of extracting defects only in the area of interest will be described below. For simplicity, the description will be limited to the case where the colored pixel portion region is inspected with transmitted illumination. An image of the entire surface of the sample is picked up under an exposure condition suitable for detecting a defect in the colored pixel portion region by transmitted illumination, and the image signal thus picked up is emphasized. Here, the image signal is differentiated in the x direction. This emphasis processing emphasizes the change of the signal in the x direction. When the colored pixel area is inspected with transmitted illumination, the entire surface of the sample is imaged, and when the obtained image signal is emphasized, coloring occurs together with the defect in the colored pixel area as shown in FIG. Boundaries other than the colored pixel area, such as the boundaries of the pixel area, the outer light-shielding area, and the outermost transparent area, are also extracted as defects, but only the area of interest (colored pixel area) is extracted, and therefore, as shown in FIG. Then, the binary image and the masking image of the area to be detected are ANDed. In this masking image, for example, the area to be detected in the memory has all the bits 1 and the other area is 0.
By ANDing with the binarized value image, the binarized value only in the region of interest (colored pixel portion region) can be left and other regions can be excluded. Next, a case of creating a masking image of the colored portion area will be described. As shown in FIG. 5 (1), first, the entire surface of the sample without defects is imaged, and the image signal obtained by imaging corresponding to the brightness level of each region of the sample is binarized to a predetermined slice level, The colored pixel area image a is extracted as in 5 (2). In order to reduce the image a in FIG. 5 (2), the image a is shifted to the left and right (up and down when the emphasis processing is the differential processing in the y direction), and ANDed as shown in FIG. To create. Since the masking image is created smaller than the actual colored pixel portion area, even if the boundary line due to the signal change of the colored portion and the outer peripheral light shielding portion is extracted as shown in FIG. 3, as shown in FIG.
Only the defects in the colored pixel area are extracted by ANDing with the masking image. Here, the shift amount in the processing of FIG. 6 will be described. The image signal obtained by imaging is as shown in FIG. 7 in the x direction. When the element filter shown in FIG. 8 is used for the image signal of FIG. 7 and filtering is performed, the result is as shown in FIG. So in this case,
Since two pixels on the left side and three pixels on the right side appear as boundary lines in the inspection area and must be excluded, the shift amount of the image a is 2 pixels on the right side and 3 pixels on the left side. The image obtained by shifting the image a and taking the AND is the masking image corresponding to the colored pixel portion area, but the exposure condition and the slice level are adjusted and imaged for other areas of the sample. By binarizing and shifting, a masking image corresponding to each area can be created.

In the above-described first and second embodiments, the image signal obtained by image pickup enters the image processing apparatus and is subjected to an emphasis process or the like to detect a defect. Hereinafter, an example of the image processing will be described. Will be described with reference to FIG. First,
Images are picked up by the image pickup means a plurality of times at the same location of the sample to obtain image signals, and the obtained image signals for a plurality of times are integrated and averaged. This is because the light transmittance distribution as shown in (a) of FIG. 10 is obtained for the portion including the defective portion, but the video signal from the image pickup device is as shown in (b) of FIG. It corresponds to a signal that includes gradual signal changes (shading) due to uneven sensitivity, random noise generated in the video signal circuit, and local changes in the signal due to dust adhering to the optical system. Then, at the same location, the image pickup means picks up an image a plurality of times to obtain an image signal, and the obtained image signals for a plurality of times are integrated and averaged. When N is set, the ratio of random noise is 1 / N ^1/2 . Where A1 and B
Reference numerals 1, A2 and B2 indicate defective portions, and C indicates a local change of the signal due to dust adhering to the optical system. Then, by slightly shifting the image pickup position and similarly performing integration and averaging, a signal (d) obtained by slightly displacing the signal (c) can be obtained. When the difference signal is obtained, it becomes as shown in (e), and the signal component that does not change due to the gradual signal change (shading) or the movement of the imaging position is removed. Here, only the signal component due to the change in light transmittance and the reduced random noise component remain, but when the signal (e) is further subjected to subtraction or differentiation of the neighborhood average value, It can be seen that the gradual signal change (shading) component is also removed, and as a result, the obtained signal (f) is locally changed only in the defect portion and is emphasized. Regarding binarization by slice processing, the signal (f) is processed at a predetermined slice level SL, and only the defective portion is judged and binarized. It should be noted that the type of defect (white defect, black defect) can be identified by the order of signal inversion in the defect portion in FIG.
In addition, here, the white defect refers to a pinhole or a pattern defect in the colored pattern portion, and the black defect refers to an extra colored portion or a light-shielded portion other than a predetermined pattern in the colored pattern portion, which causes stains or stains. Scratches on the glass part are also detected as black defects.

[0010]

With the above-described structure, the method of detecting defects in the color filter for a viewfinder according to the present invention enables reliable automatic inspection of appearance defects such as dirt and scratches on the front and back of the color filter for a viewfinder. I am trying.

[Brief description of drawings]

FIG. 1 is a schematic view of an apparatus according to a first embodiment of the present invention.

FIG. 2 is a schematic diagram of an apparatus according to a second embodiment of the present invention.

FIG. 3 is an emphasis processing diagram.

FIG. 4 is a masking diagram of a binarized image.

[Figure 5] Masking image creation process diagram

FIG. 6 is a shift processing diagram for creating a masking image.

FIG. 7: x-direction image signal diagram

FIG. 8: Element filter diagram

FIG. 9 is an image signal diagram after filtering.

FIG. 10 is a diagram showing an example of image processing.

FIG. 11 is a schematic view of a color filter for a viewfinder.

[Explanation of symbols]

1, 11 Imaging camera 2, 12 Light source 3, 13 Viewfinder color filter 4 Holding plate 5, 15 Image processing device 6 Half mirror 14 Transparent holding plate

Claims (4)

[Claims] 1. A signal obtained by imaging the entire surface of a color filter for a viewfinder, which has different transmittance and / or reflectance depending on a location, with an imaging means using the transmitted illumination light and / or the reflected illumination light. A method of automatically inspecting appearance defects by performing image processing, wherein the inspection area of the color filter is
Alternatively, the defect inspection of each area is performed according to the difference in the reflectance, and the image signal level of the image pickup means is adjusted according to the exposure condition for each area, and the entire inspection area is imaged to obtain an image obtained by imaging. A defect detecting method for a color filter for a viewfinder, wherein a defect is detected by a signal, and the entire inspection region is inspected together with defect detection for each region. 2. The defect detection method for a color filter for a viewfinder according to claim 1, wherein the adjustment of the video signal level of the image pickup means according to the exposure condition is performed during the image pickup exposure time of the electronic shutter of the image pickup means. 3. The color filter for a viewfinder according to claim 1, wherein the image signal level of the image pickup means is adjusted by adjusting the illumination light intensity of the transmitted light and / or the reflected light according to the exposure condition. Defect detection method. 4. The adjustment of the video signal level of the image pickup means according to the exposure conditions, the adjustment by the image pickup exposure time of the electronic shutter of the image pickup means of the image pickup means, and the illumination light intensity of the transmitted light and / or the reflected light according to claim 1. A method for detecting defects in a color filter for a viewfinder, which is characterized in that adjustment and combination are performed.