JP2000081368A - Lcd panel picture quality inspection method and device and image-reading method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reduce the difference in the pixel element of light and shade image data and to accurately inspect picture quality by picking up the image of each of a plurality of regions where the display screen of a liquid crystal display(LCD) panel is divided separately by a plurality of image pick-up parts. SOLUTION: The display screen of an LCD panel 1 to be inspected is divided into a plurality of regions, and the images of the plurality of regions are picked up separately by a plurality of image pick-up parts 11 that are provided corresponding to each region. Light and shade image data from each of the image pick-up parts 11 being obtained by picking up images is synthesized to one image data by correcting the difference in the sensitivity of each of the image pick-up parts 11, the synthesized light and shade image data is analyzed, and a display defect is detected. More specifically, the light and shade image data being obtained by picking up images at each of the image pick-up parts 11 is converted from analog into digital by an A/D converter 7, and the converted digital data is fetched into an image memory 8 via a bus line 10. Then, region data being selected according to an address from a region designation table 15 is written into an image synthesis memory 15 and an entire light and shade image data is synthesized.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a high-definition LCD.
The present invention relates to an LCD panel image quality inspection method for inspecting the display quality of a (Liquid Crystal Display) panel, an image quality inspection apparatus using the inspection method, and an image capturing method used for the inspection.

[0002]

2. Description of the Related Art In an image quality inspection of an LCD panel, a CCD is used.
(Charge Coupled Device) The brightness of a pixel of the inspection object is measured using a camera. FIG. 7 shows an example of a conventional LCD panel image quality inspection apparatus. The display surface of the LCD panel 1 to be inspected is imaged by the imaging unit 11, and the imaged data is converted into digital bright and dark image data by the A / D converter 7, and the bright and dark image data is stored in the image memory 8 to improve the image quality. Perform inspection. The image quality inspection is executed by the arithmetic control means 9 constituted by a CPU or the like.
The imaging unit 11 measures the camera unit 2, the optical lens 4 for forming an image on the camera unit 2, and the amounts of red, green, and blue light when the inspection target LCD panel 1 is a panel capable of displaying color. Disk 3 with red, green, and blue filters
And a motor 5 for rotating the disk 3. The motor 5 is controlled by motor control means 6.

The resolution of mass-produced LCD panels is SVGA (80
0 × 600 pixels) to XGA (1024 × 768 pixels), SXGA (1280
(× 1024 pixels). In the inspection of such a high-definition LCD panel, the brightness of a large number of pixels must be accurately measured. At this time, since both the arrangement of the pixels of the LCD panel and the arrangement of the image pickup devices of the CCD camera have a regular lattice-like structure, moire fringes easily occur in the bright and dark image data captured by the CCD camera. The image data in which the moiré fringe has occurred does not accurately reproduce the light and dark state of the pixel of the inspection object, and thus cannot be used for image quality inspection.

For example, Japanese Patent Application Laid-Open No. 8-
The image quality inspection apparatus described in Japanese Patent No. 29360 discloses a resolution that is twice or more the display resolution of the LCD panel to be inspected (one pixel of the LCD panel is imaged by a CCD camera with 2 × 2 = 4 pixels). The moiré fringes are prevented by capturing image data with a CCD camera having a). In this image quality inspection system, bright and dark image data of an LCD panel on which a calibration pattern is
D Take in the camera and pay attention to the pixel value (brightness) distribution of the calibration pattern in the light and dark image data.
The address of the pixel of the light and dark image data corresponding to each pixel of the LCD panel is accurately calculated with a real number. The calculated address is called a sampling address.

Next, a rectangular area corresponding to the photographing magnification is set as bright and dark image data centering on the sampling address, and the pixel values of a plurality of pixels intersecting the rectangular area are calculated as the ratio of the intersecting area. Multiply and add
This is the pixel value of the LCD panel. Through the above processing, bright and dark image data having the same number of pixels as the LCD panel to be inspected can be obtained with high accuracy, and a fine display defect can be detected. This method uses LC
As the resolution of D-panels increases, that is, the number of pixels increases, it is necessary to use CCD cameras with higher resolution. However, no CCD camera has been mass-produced that can capture, for example, an SXGA LCD panel with twice or more the same resolution as in the past. Therefore, when the resolution of the LCD panel is further increased, it becomes difficult to capture and inspect an image using a CCD camera.

Hitherto, there has been proposed a method of solving this problem and using a low-resolution CCD camera to capture bright and dark image data of a high-resolution LCD panel. That is, as a second prior art, Japanese Patent Laid-Open No. 8-65656 discloses
It has two plane parallel plates and uses the refraction of light passing through the inclined plane parallel plate to shift the optical path of light incident on the CCD camera from the LCD panel. As a result, the area of the LCD panel (shooting area) captured by the CCD camera is slightly moved.

While controlling the inclination angle of the two plane-parallel plates so that the photographing area slightly moves in the X and Y directions, the CCD camera captures the bright and dark image data of the LCD panel. The resolution of the bright and dark image data can be improved by synthesizing the plurality of captured bright and dark image data in accordance with the movement of the shooting area. Such multi-scan technology has made it possible to capture high-resolution LCD panel light / dark image data using a low-resolution CCD camera.

[0008]

In the first and second prior arts described above, a special lens having no aberration is used in order to maximize the effective pixels of the CCD camera.
For this reason, the adjustment of the shooting magnification is mainly performed by adjusting the distance from the LCD panel to the CCD camera. Here, the shooting magnification is the ratio between the display resolution of the LCD panel and the resolution of the CCD camera. For example, one pixel of an LCD panel
When photographing with four pixels (2 × 2 pixels) of a CCD camera, the photographing magnification is twice.

If the size of the LCD panel is constant as the definition becomes higher, the area of the pixel decreases as the number of pixels increases. In order to shoot the pixels of the LCD panel at the same magnification as before, or as high as possible, the CCD camera must be close to the LCD panel. By bringing the CCD camera closer, the solid angle for shooting the LCD panel increases. LCD panels have a property called viewing angle characteristics. Assuming that the angle at which the center of the panel is viewed from directly in front is 0 degrees, the brightness and contrast of the display appear to change depending on the angle at which the panel is viewed (viewing angle). Due to the viewing angle characteristics, the pixel value corresponding to the center of the panel and the pixel value corresponding to the edge (peripheral) portion of the light and dark image data obtained by photographing the LCD panel with the CCD camera are different. That is, the pixel value is large at the center of the panel, and gradually decreases toward the periphery.

[0010] The CCD camera is brought close to the LCD panel, and the LCD
Since the solid angle for photographing the edge portion of the panel becomes large, the influence of the viewing angle characteristics becomes large. For this reason, also in the bright and dark image data, the difference in brightness between the pixel corresponding to the central portion and the pixel corresponding to the edge portion is enlarged. Even if the size of the LCD panel is increased with higher definition and the pixel size does not change, the influence of the viewing angle characteristics will increase. This is because in this case, the CCD camera is not brought close to the LCD panel, but the solid angle at which the image is captured becomes large due to the enlargement of the panel itself.

A description will be given of an increase in the influence of the viewing angle characteristics with the increase in definition and size of the LCD panel. The solid line in FIG. 6 is a horizontal line of bright and dark image data obtained by photographing an LCD panel with a display resolution of SXGA (1280 × 1024 pixels) at a magnification of 1.5 times using the second prior art multi-scan technology. Indicates the pixel value. The horizontal axis is the X address of the image data (in the horizontal scanning direction), and the vertical axis is the pixel value whose maximum value is normalized to 1. In this measurement example, the edge portion of the panel (X
It can be seen that the pixel value decreases toward the addresses (0, 1280) and that the pixel value at the edge is about 60% of the center.

The dotted line in FIG. 6 is a graph obtained by approximating the measured data by the cos function. The approximate curve of the cos function is considered to reflect the measured viewing angle characteristics of the LCD panel.
The display resolution UXGA (1
FIG. 5 shows a simulation result of a change in pixel value when measuring an LCD panel of (600 × 1200 pixels). However, the calculation was performed under a photographing condition in which the magnification was 2.0 times and multi-scan was not performed. As can be seen from the graph of FIG. 5, the pixel value of the light and dark image data sharply decreases due to the increase in the effect of the viewing angle characteristic accompanying the high definition of the LCD panel, and the pixel value of the edge portion is 15% of the central portion. About. Thus, 1
In the case of bright and dark image data obtained by photographing two LCD panels, if the pixel values of each pixel are largely different depending on the position of the pixel, the accuracy of the image quality inspection is reduced. That is, (1) In the inspection using the threshold processing, even if a defect occurring in the center of the LCD panel can be detected, the defect at the edge portion is missed.

(2) When a defect having the same brightness exists in the center portion and the edge portion, even if both of them can be detected, the defect is evaluated as a defect having different brightness. As a solution to this problem, a correction process of bright and dark image data can be considered. That is, a decrease in the pixel value due to the viewing angle characteristic is measured in advance, and each pixel is multiplied by a correction coefficient so as to have a constant level. However, when a pixel value having a difference of about 85%, such as a simulation result, is amplified to a constant level, there is a problem that a noise component of bright and dark image data is also enhanced at the same time. For example, bright and dark image data captured by a CCD camera contains about 5% random noise due to the dark current of the CCD. If the difference between the pixel values due to the viewing angle characteristic is 10 to 20%, the random noise is only about 6 to 7% even if the coefficient is multiplied so that the difference is 0. However, if the image is corrected so that the difference of 85% becomes 0, the random noise increases to 33%. That is, the emphasized noise component is erroneously detected.

SUMMARY OF THE INVENTION An object of the present invention is to provide an image quality inspection for performing an accurate image quality inspection by inspecting a high-definition LCD panel by minimizing the influence of viewing angle characteristics and minimizing the difference between pixel values of bright and dark image data. A method, an image quality inspection apparatus using the method, and a method for capturing pixel data are proposed.

[0015]

According to a first aspect of the present invention, a light / dark display of an LCD panel to be inspected is picked up by an image pickup unit, the light / dark image data is taken into an image memory, and the light / dark image data taken into the image memory is taken. Analyze and inspect
In an LCD panel inspection method for detecting a display defect of an LCD panel, a display surface of an inspection target LCD panel is divided into a plurality of regions, and the plurality of regions are formed by a plurality of imaging units provided corresponding to each region. The light and dark image data from each of the imaging units obtained by imaging separately are corrected into a single image data by correcting the difference in sensitivity of each of the imaging units, and the synthesized light and dark image data is analyzed. The present invention proposes an LCD panel inspection method for detecting display defects.

According to a second aspect of the present invention, in addition to the difference in sensitivity between the imaging units, the difference in viewing angle characteristics between the imaging units is corrected and combined into one image data, and the combined bright and dark image data is combined. The present invention proposes an LCD panel inspection method for analyzing display and detecting display defects. According to the LCD panel inspection method of the present invention, the display surface of one LCD panel is divided into a plurality of regions, and each of the divided plurality of regions is imaged by an imaging unit provided for each region, Since a method of obtaining bright and dark image data is employed, each imaging unit may target one-sixth of the number of pixels constituting one LCD panel as an imaging target.

Therefore, there is an advantage that the image quality of the LCD panel can be accurately inspected by using an image pickup section having a resolution lower than the display resolution of the LCD panel to be inspected. Further, the image capturing method according to the present invention, the display of the LCD panel is normally white, and the bright and dark image data of the panel to be inspected photographed by the camera unit of the first imaging unit is written into the image memory.
From the area designation table, an area designation address corresponding to the light and dark image data of the first imaging unit is read, data in the range of the area designation address is read from the light and dark image data on the image memory, and the average value is calculated. (1)
Shooting of an LCD panel by a plurality of first to N-th imaging units, writing of captured light / dark image data to an image memory, reading of an area designation address corresponding to each imaging unit, and area from the image memory Repeat reading data from the specified address and calculating average value G (n), (n = 1,2, ..., N) G (n), (n = 1,2, ..., N ) Is calculated as G, and G
(n), (n = 1,2, ..., N) to obtain the sensitivity coefficient g (n), (n = 1,2, ..., N) of each imaging unit. Is stored in the correction coefficient table.

A test pattern is displayed on the LCD panel, and light / dark image data of the panel to be inspected photographed by the camera unit of the first image pickup unit is written into an image memory. Is read out, data in the range of the area designation address is read out from the light and dark image data on the image memory, the maximum value is obtained as V, and the sensitivity coefficient g of the first imaging unit is obtained from the correction coefficient table. (1) is read, and for each pixel of the read brightness data, the maximum value V is divided by the pixel value, multiplied by a sensitivity coefficient g (1), and used as a correction coefficient for the sensitivity and viewing angle characteristics of the pixel. Writing the read light and dark image data in a position corresponding to the imaging unit of the image synthesizing memory,
Photographing an LCD panel displaying a test pattern by a plurality of first to N-th imaging units, writing captured bright and dark image data to an image memory, and reading an area designation address corresponding to each imaging unit; Reading data from the area designation address from the image memory and sensitivity coefficients g (n), (n = 1, 2,..., N) corresponding to the respective imaging units
Reading, calculating the correction coefficient for the sensitivity and viewing angle characteristics of each pixel, and writing to the image synthesis memory are repeated.
For all the pixels of the light and dark image data corresponding to the entire panel, the correction coefficients of the sensitivity and the viewing angle characteristics are obtained, and the coefficients of all the pixels are stored in the correction coefficient table. By reading out and multiplying the coefficients corresponding to the pixels, the brightness and darkness image data of the LCD panel can be obtained in which the unevenness of the sensitivity between the imaging units and the viewing angle characteristics of the LCD panel are eliminated.

According to the image quality inspection method and the image capturing method of the present invention, since a plurality of imaging units are provided and the LCD panel to be inspected is divided into a plurality of imaging regions for imaging, the solid angle of the imaging unit can be reduced. The influence of the viewing angle characteristic on the captured light / dark image data can be reduced. As mentioned above,
Deviation of pixel value of light and dark image data due to viewing angle characteristics is 10 to
If it can be reduced to about 20%, correction by multiplication of coefficients is possible.

As shown in FIG. 5, the display resolution UXGA (160
When an LCD panel (0x1200 pixels) is photographed with a single CCD camera, the maximum deviation between bright and dark pixels is 85%. But,
The difference between bright and dark image data is only 22% at the maximum in the viewing angle range for shooting a panel with a display resolution of SVGA (800 × 600 pixels). If the UXGA panel is divided into four equal areas, each area is 800 x 600 pixels and therefore has the same resolution as the SVGA panel. Therefore, if the four CCD cameras that are used to shoot the SVGA panel are arranged so that each corresponds to the area obtained by dividing the UXGA panel into four equal parts, and the UXGA panel is shot, the viewing angle will be SVGA. Is the same as the viewing angle when shooting the panel.

The difference between the pixel values in the light and dark image data photographed by each CCD camera is 22%, which is the same as the photographing of the SVGA panel. As in this example, a high-definition, large LCD
Even in the case of a panel, if a plurality of image pickup units are provided and the photographing region of the panel is divided so that the viewing angle becomes small, it is possible to obtain bright and dark image data capable of correcting a difference in pixel value. .

According to the image capturing method of the present invention, the influence of the viewing angle characteristic is corrected for each of the bright and dark image data captured by the plurality of imaging units. To correct the viewing angle characteristics, an LCD panel displaying a test pattern used for image quality inspection is photographed, and the bright and dark image data is used. The data of the area of the LCD panel (the outside of the panel may be captured) is cut out from the captured light and dark image data, and the maximum value V
For The maximum value V is data captured at a viewing angle near 0 degrees. The maximum value V is divided by each pixel value of the cut-out area data to obtain a correction coefficient v (i, j), (i = 1,
2, ..., I, j = 1,2, ..., J). For each light and dark image data of each image pickup unit, region data is cut out and a correction coefficient is calculated. The influence of the viewing angle characteristic is eliminated by multiplying each pixel of the area data cut out from the captured light and dark image data by the correction coefficient v (i, j). The area data of the light and dark image data captured by the n-th imaging unit is represented by I (i, j,
n), (i = 1, 2,... I, j = 1, 2,..., J), the correction is performed as follows. Iv (i, j, n) = I (i, j, n) × v (i, j), (i = 1,2, ..., I, j = 1,2, ..., J) . (1) Here, Iv (i, j, n) is the pixel value of the data after correcting the viewing angle characteristics.

However, in the case of photographing with a plurality of image pickup sections, there is a problem that the average brightness of the picked-up bright and dark image data differs because the cameras of the respective image pickup sections have different sensitivities. For this reason, in the image capturing method of the present invention, the sensitivity of each imaging unit is corrected and made uniform. To correct the sensitivity of each imaging unit, an LCD panel whose display is normally white is photographed, and the light and dark image data is used.

This is because in a normally white state
This is because, since no voltage is applied to the liquid crystal of the LCD panel, only the light of the backlight is transmitted, and a defect due to the operation of the liquid crystal is not reflected on the bright and dark image data. The average value of the pixel values of the area data obtained by photographing the LCD panel of the light and dark image data photographed in this manner represents the sensitivity of each image pickup unit. Therefore, the average value G (n), (n = 1, 2,... N) of the pixel values for each area is calculated from the light and dark image data of each imaging unit. G (n) represents the sensitivity of each imaging unit.

Next, the average value G of G (n) is calculated. G is the average sensitivity of all imaging units. G is divided by G (n), and correction coefficients g (n), (n = 1, 2,...,
N). For each imaging unit, the sensitivity of each imaging unit is made uniform by multiplying each pixel of the area data cut out from the captured light and dark image data by a correction coefficient g (n). The area data of the bright and dark image data captured by the n-th imaging unit is represented by I
If (i, j, n), (i = 1,2,..., I, j = 1,2,..., J), the correction is performed as follows. Ig (i, j, n) = I (i, j, n) × g (n), (i = 1,2, ..., I, j = 1,2, ..., J). (2) Here, Ig (i, j, n) is the pixel value of the corrected data.

When the correction processing is performed a plurality of times on the captured light and dark image data, the processing time increases and the throughput decreases. In particular, the correction processing for all the captured images lowers the throughput. Therefore, it is better that the number of correction processes is small. As is clear from the equations (1) and (2), the correction of the viewing angle characteristic and the correction of the sensitivity of the imaging unit are not too much multiplication of the area data I (i, j, n) of the bright and dark image data by the coefficient. This calculation can be performed by one process as in the following equation. Ic (i, j, n) = I (i, j, n) × v (i, j) × g (n), (i = 1,2, ..., I, J = 1,2, .., J, n = 1, 2,..., N). (3) Here, Ic (i, j, n) is the pixel value of the corrected data.

According to the image capturing method of the present invention, the correction coefficient v (i, j) of the viewing angle characteristic is calculated by dividing the maximum value V by each pixel value of the area data and simultaneously correcting the camera sensitivity.
Multiplies g (n). That is, the product v (i, j) × g (n) of the correction coefficient of the viewing angle characteristic and the correction coefficient of the camera sensitivity is obtained. Assuming that the product is c (i, j, n), the correction processing for the area data I (i, j, n) of the light and dark image data is as follows. Ic (i, j, n) = I (i, j, n) × c (i, j, n), (i = 1,2, ..., I, j = 1,2, ..., J, n = 1,2, ..., N). (4) As a result, two correction processes for the light and dark image data can be performed once, so that a decrease in throughput due to the correction process can be reduced. When there are a plurality of test patterns used for image quality inspection, that is, when a plurality of inspections are performed by changing display gradations and colors, bright and dark image data is photographed for each test pattern display, and a correction coefficient is obtained. In addition, a correction coefficient is obtained in advance by using a non-defective LCD panel, and the correction coefficient is stored in a correction coefficient table.
It is also used to correct captured light and dark image data. To specify the area to be cut out from the captured light and dark image data, the address is set in advance by observing the light and dark image data of each imaging unit on a display device such as a CRT. Alternatively, a calibration pattern is displayed at a position where the display area of the LCD panel is equally divided into N areas, and the address is obtained by using a method described in JP-A-8-29360 by using light and dark image data obtained by photographing the calibration pattern. calculate. The preset address is stored in the area designation table and used to capture the light and dark image data of the LCD panel.

[0028]

FIG. 1 shows an embodiment of an LCD panel image quality inspection apparatus to which an LCD image quality inspection method according to the present invention is applied. The present invention is characterized in that the display screen of the LCD panel 1 to be inspected is divided into a plurality of regions, and each of the divided regions is photographed by a different imaging unit 11.

In the example shown in FIG. 1, a case is shown in which the display surface of one LCD panel 1 to be inspected is picked up by four image pickup units 11. Therefore, in this case, each imaging unit 11
Means that an image of 1/4 of the entire area of the LCD panel to be inspected may be taken. Each imaging unit 11 includes a camera unit 2, a rotating plate 3 equipped with a color filter, an optical lens 4,
A case is shown in which the motor is configured to rotate the rotating plate 3.

The rotating plate 3 is provided with optical filters of red, green, blue and transparent. When the LCD panel 1 to be inspected is capable of color display, the optical filters red, green and blue are provided on the incident side of the camera unit 2.
A case is shown in which any one of green and blue is interposed so that the amount of red, green and blue light can be measured separately. When measuring the amount of white light, a transparent filter is attached to the camera unit 2.
To the front of the These switching operations are performed by rotating the rotating plate 3 by the motor 5. The motor 5 of each imaging unit 11 is controlled by a motor control unit 6.

Bright and dark image data obtained by photographing in each image pickup section 11 is converted into a digital signal by the A / D converter 7, and the converted digital bright and dark image data is processed by arithmetic control means (CPU).
Are taken into the image memory 8 via the bus line 10.
When calculating the sensitivity correction coefficient of each imaging unit 11, the operation is performed in the following procedure. First, the display on the LCD panel 1 is set to normally white. The first imaging unit 11 photographs the LCD panel. The light / dark image data converted to digital by the A / D converter 7 is temporarily stored in the image memory 8. In accordance with the address read from the area designation table 15, light / dark image data of the area where the LCD panel 1 is photographed is selected from the data in the memory 8.

An average value of the selected area data is obtained. For each of the imaging units 11, the acquisition of bright and dark image data and the calculation of the average value of the area data are sequentially performed. After calculating the average value of the area data for all the imaging units 11, the sensitivity correction coefficient of each imaging unit 11 is calculated from the average value. The calculated sensitivity correction coefficient is written in the correction coefficient table 13.

Next, a test pattern is displayed on the LCD panel 1, and the first image pickup section 11 photographs the LCD panel 1.
Bright and dark image data converted into digital by the imaging unit 11 is:
It is temporarily stored in the image memory 8. In accordance with the address read from the area designation table 15, from the data in the memory 8, the light and dark image data of the area where the LCD panel is photographed is selected.

The maximum value of the selected area data is obtained.
From the correction coefficient table, a correction coefficient for the sensitivity of the first imaging unit 11 is read. A value obtained by multiplying the maximum value of the area data by the correction coefficient of the sensitivity is divided by each pixel value of the area data to obtain a correction coefficient relating to the sensitivity and the viewing angle characteristics of each pixel of the area data. The obtained correction coefficient is stored in the correction coefficient table 13.
Write to. For each of the imaging units 11, the acquisition of bright and dark image data and the calculation of correction coefficients are sequentially performed. The obtained correction coefficient is written in the correction coefficient table 13.

Light and dark image data used for the image quality inspection are taken in the following procedure. The test pattern is displayed on the LCD panel 1, and the first imaging unit 11 photographs the LCD panel 1. The light / dark image data converted to digital by the A / D converter 7 is temporarily stored in the image memory 8. According to the address read from the area designation table 15,
From the data in the memory 8, the light and dark image data of the area where the LCD panel is photographed is selected.

The selected area data is stored in the image synthesizing memory 12.
Write to. The light / dark image data is fetched sequentially for each imaging unit 11 and the area data is written into the image synthesizing memory 12 to synthesize the light / dark image data of the entire LCD panel 1. The image correcting means 14 reads the light and dark image data in the image synthesizing memory 12 and the correction coefficient in the correction coefficient table, and obtains image data obtained by multiplying the light and dark image data by the correction coefficient. The image correction means 14 is composed of, for example, a memory for temporarily storing data and a multiplier. Various image quality inspections are performed using the image data.

The arithmetic and control unit performs control of the image pickup unit 11, calculation of the average or maximum value of the area data, calculation of the correction coefficient, and image quality inspection using the corrected light and dark image data. An example of the processing flow of the image capturing method of the present invention is shown in FIGS. In step S1, the display on the LCD panel 1 is set to normally white. In step S2, the LCD
A register n indicating the number of the image pickup unit 11 that executes photographing of the panel 1 is initialized to 1. In step S3, the bright / dark image data captured by the n-th imaging unit 11 is written to the image memory 8. Let the written data be tmp (x, y). In step S4, from the area specification table, the area specification addresses (xs (n), ys (n)) and (xe (n), ye (n)) corresponding to the n-th imaging unit 11
Is read.

The area is defined by the start address (xs (n), ys (n)) and the end address (xe (n), ye (n)) of the range to be cut out for the written two-dimensional data. It is specified.
In step S5, the data of the designated area is read from the image memory 8, and the average value G (n) is obtained. Step S6
Then, 1 is added to a register n representing the number of the imaging unit 11.

In step S7, the register n and the imaging unit 11
Are compared, and if n is equal to or smaller than N, the process returns to step S3, where the image data is fetched again and the average value of the designated area is calculated. If n is greater than N, step S
Proceeding to 8, the average value of G (n) is calculated. In step S9,
A correction coefficient g (n) for equalizing the sensitivity of the camera unit 2 provided in each imaging unit 11 is calculated. In step S10, the correction coefficient g (n) is written into the correction coefficient table 13.

Next, at step S11, a test pattern is displayed on the LCD panel 1, and at step S12, a register n indicating the number of the image pickup section 11 for executing photographing of the LCD panel 1 is initialized to 1. In step S13, the bright / dark image data captured by the n-th imaging unit 11 is stored in the image memory 8
Write to. Let the written data be tmp (x, y). In step S14, the n-th imaging unit 11 is
(Xs (n), ys (n)) and (xe (n), y
e (n)). In step S15, data in the designated area is read from the image memory, and the maximum value V is obtained. In step S16, a correction coefficient g (n) of the camera sensitivity of each imaging unit is read from the correction coefficient table, and step S1 is executed.
7, a correction coefficient c (x, y) for correcting the camera sensitivity and the viewing angle characteristic for each pixel of the specified area data from the maximum value V, the correction coefficient g (n), and each pixel value of the specified area data. , n) is calculated. Correction coefficient c (x, y, n) determined in step S18
Is written in the correction coefficient table.

In step S 19, the designated area data is written to the image compositing memory 12. At this time, data is written corresponding to the position where the imaging unit 11 that has performed the shooting is arranged. In step S20, 1 is added to the register n indicating the number of the imaging unit 11. In step S21, the register n is compared with the number N of the imaging units 11, and if n is equal to or smaller than N, the process returns to step S13 to fetch the image data again, calculate the correction coefficient, and store the designated area data in the image synthesis memory. Is written. If n is greater than N, the process proceeds to step S22, where
Reads the bright and dark image data of the entire panel 1. Step S
At 23, the correction coefficient c (x, y, n) for each pixel of the light and dark image data is read from the correction coefficient table 13. Step S
At 24, the light and dark image data 12 is corrected by multiplying each pixel of the light and dark image data by the correction coefficient c (x, y, n). As described above, light and dark image data in which the sensitivity and the viewing angle characteristics of the camera unit 2 of the imaging unit 11 have been corrected are obtained.

[0042]

[Modification] It is not necessary to calculate the correction coefficient for each LCD panel 1. A correction coefficient is calculated using an LCD panel having no defect, stored in the correction coefficient table 13, and the correction coefficient may be used in the image quality inspection.
FIG. 4 shows an example of the processing flow of image capture in this case. The processes corresponding to the flows in FIGS. 2 and 3 are denoted by the same step numbers.

The imaging device used in the camera section 2 of the present invention is not limited to a CCD, but may be a MOS (Metal Oxide Semiconductor).
It is also conceivable to apply a shape sensor or the like.

[0044]

As described above, according to the present invention, the display surface of one inspection-target LCD panel 1 is divided into a plurality of regions, and the plurality of divided regions are arranged corresponding to each region. The image is captured by the imaging unit 11 to obtain light and dark image data,
Since we proposed an image quality inspection method for inspecting image quality using this light and dark image data, even if the LCD panel to be inspected becomes higher definition and the number of pixels will increase dramatically in the future, the number of pixels will be smaller than the number of pixels of the LCD panel to be inspected It is possible to obtain bright and dark image data with good resolution by taking an image with an image sensor composed of pixels.

After all, a highly accurate image quality inspection can be performed using an imaging unit having a low resolution, and the effect is extremely large for practical use.

[Brief description of the drawings]

FIG. 1 is a block diagram for explaining an embodiment of an LCD panel image quality inspection apparatus to which an LCD panel image quality inspection method according to the present invention is applied.

FIG. 2 is a flowchart for explaining an image capturing method according to the present invention.

FIG. 3 is a flowchart showing a continuation of FIG. 2;

FIG. 4 is a flowchart for explaining another example of the image capturing method of the present invention.

FIG. 5 is a diagram showing a simulation result of a pixel value change due to a viewing angle characteristic of an LCD panel.

FIG. 6: Display resolution SXGA captured by a conventional image quality inspection device
FIG. 3 is a diagram showing pixel values of a panel.

FIG. 7 is a block diagram for explaining a conventional LCD panel image quality inspection apparatus.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 LCD panel to be inspected 2 Camera part 3 Rotating plate 4 Optical lens 5 Motor 6 Motor control means 7 A / D converter 8 Image memory 9 Operation control means 10 Bus line 11 Imaging part 12 Image synthesis memory 13 Correction coefficient table 14 Image correction Means 15 Area designation table

Claims (4)

[Claims] An image pickup unit picks up a light / dark display of an LCD panel to be inspected, fetches the light / dark image data into an image memory, analyzes the light / dark image data fetched into the image memory, and displays the data on the LCD panel to be inspected. LCD to detect defects
In the panel inspection method, the display surface of the inspection target LCD panel is divided into a plurality of regions, and the plurality of regions are separately imaged by a plurality of imaging units provided corresponding to each region, and are obtained by imaging. And correcting the difference in the sensitivity of each imaging unit to synthesize one image data, and analyzing the synthesized light and dark image data to detect a display defect. LCD panel image quality inspection method. 2. The LCD panel image quality inspection method according to claim 1, wherein a difference in sensitivity of each imaging unit and a difference in viewing angle characteristics are corrected and combined into one image data, and the combined bright and dark image is combined. An image quality inspection method for an LCD panel, comprising analyzing data and detecting a display defect. 3. An LCD drive unit for displaying various test patterns on a panel to be inspected, and B. a display surface of the panel to be inspected is divided into a plurality of areas;
A plurality of image pickup units for individually picking up each area; C, an AD converter for converting the light and dark image data picked up by the image pickup unit into digital data; and D, a light and dark image data AD-converted by the AD converter. An image memory for temporarily storing; E, a combined image memory for combining the light and dark image data stored in the image memory into one image data; and F, data used for combining an image from the light and dark image data stored in the image memory. An area designation table recording addresses for designating an area; G, an image correction means for correcting the sensitivity between the image pickup units and a viewing angle characteristic of bright and dark image data photographed by each image pickup unit; A correction coefficient table in which coefficients for each pixel for correction are recorded; I, controlling the capture of bright and dark image data captured by the plurality of image capturing units; An LCD panel image quality inspection device, comprising: an arithmetic control unit that detects display defects of an inspection target panel using image data. 4. A normally white display on an LCD panel, light and dark image data of a panel to be inspected photographed by a first image pickup unit is written into an image memory, and light and dark image data of the first image pickup unit are read from an area designation table. And reads out the data in the range of the above-mentioned area designation address from the bright and dark image data on the image memory, and calculates the average value.
Let G (1) be the above, photograph the LCD panel with the first to Nth imaging units, write the captured bright and dark image data to the image memory, and specify the area designation address corresponding to each imaging unit. Reading, reading of the data in the area designation address from the image memory, and the average value G (n), (n = 1, 2,...,
N), and the average of G (n), (n = 1,2, ..., N) is calculated as G, G (n), (n = 1,2, ... , N) to divide G into the sensitivity coefficient g (n) (n = 1, 2,..., N) of each imaging unit, and this sensitivity coefficient g (n)
Is stored in the correction coefficient table, the test pattern is displayed on the LCD panel, the light and dark image data of the panel to be inspected photographed by the first image pickup unit is written into the image memory, and the light and dark of the first image pickup unit is read from the area designation table. The area designation address corresponding to the image data is read out, the data in the range of the area designation address is read out from the bright and dark image data in the image memory, the maximum value is obtained as V, and the sensitivity of the first imaging unit is obtained from the correction coefficient table. Coefficient g (1)
For each pixel of the read light and dark data, the maximum value V is divided by the pixel value and multiplied by a sensitivity coefficient g (1) to obtain a correction coefficient for the sensitivity and viewing angle characteristics of the pixel. The bright and dark image data is written in the image composing memory at a position corresponding to each image capturing unit. The photographing of the LCD panel displaying the test pattern by the first to Nth plurality of image capturing units, and the captured bright and dark image data Writing to the image memory, reading of the area designation address corresponding to each imaging unit, reading of data in the area designation address from the image memory, and the sensitivity coefficient g (n), (N = 1,2, ...
・ Repeat N), calculate the correction coefficient of the sensitivity and viewing angle characteristics of each pixel, and write it to the image synthesis memory. The sensitivity and viewing angle are obtained for all pixels of bright and dark image data corresponding to the entire LCD panel. The correction coefficients of the characteristics are obtained, the coefficients of all the pixels are stored in a correction coefficient table, and the light and dark image data in the image synthesis memory are read out from the correction coefficient table and multiplied by the coefficient corresponding to each pixel, and the sensitivity between the imaging units is obtained. An image capturing method characterized by obtaining light and dark image data of an LCD panel, which eliminates unevenness of the LCD panel and viewing angle characteristics of the LCD panel.