JPH11201867A - Address calibration method - Google Patents

Abstract

(57) [Summary] To provide a more accurate address calibration method. A predetermined display pattern is displayed on an LCD panel (step S10), and pixels of the display pattern image of the LCD panel formed on an imaging surface of a CCD camera are respectively arranged at predetermined intervals in an X direction and a Y direction. An image is captured by the n × m shots having a shift (step S11), a plurality of captured images captured by the n × m shots are synthesized, and each of the display pixels of the flat panel in the synthesized image is synthesized. A corresponding address is obtained (step S12), and the address is converted into an address corresponding to an image taken in a display test (step S1).
3).

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a CCD (charge c)
An image of a test image displayed on a flat panel (LCD (liquid crystal display) panel or the like) to be inspected is imaged using an oupled device (camera) camera, and a display test of the panel (deletion or display of display pixels) The present invention relates to an address calibration method performed in a panel test system that performs non-uniformity detection.

[0002]

2. Description of the Related Art FIG. 7 is a block diagram showing a general configuration of an LCD panel test apparatus. In the figure, an LCD panel 101 to be tested displays an image based on a drive signal from a drive unit 102. The drive unit 102 is configured to output a drive signal based on a signal from the control unit 108, and can display a desired pattern on the LCD panel 101. On this LCD panel 101, CC
A D camera 103 is arranged, and the CCD camera 1
03 captures an image displayed on the LCD panel 101.

The image capturing section 104 temporarily captures image data captured by the CCD camera 103, and transfers the captured image data to the data processing section 105 under the control of the control section 108.

The data processing unit 105 includes a presampling processing unit 106 and a loss detecting unit 107. The presampling processing unit 106 converts the image data (image data captured by the CCD camera 103) from the image capturing unit 104 into an image corresponding to the display pixels of the LCD panel 101, and performs sampling processing. The loss detection unit 107
Based on the result of the sampling processing, detection of display pixel loss or display unevenness of the LCD panel 101 is performed.

In the above-described LCD panel test apparatus, image data picked up by the CCD camera 103 is converted into L data.
In order to convert an image corresponding to the display pixel of the CD panel 101, the corresponding address of the LCD panel display pixel on the CCD image is specified, and the correspondence between the LCD panel display pixel and the CCD pixel is determined based on the address. Address calibration needs to be performed. Hereinafter, an example of an address calibration method performed in the LCD panel test apparatus will be described.

In address calibration, LC
The position (address) of each pixel of the D panel 101 in the image captured by the CCD camera 103 is determined. For example, as shown in FIG. 8, a calibration pattern including 25 bright points dividing the panel into 16 is displayed on the LCD panel 101, and the pattern is imaged by the CCD camera 103. The captured images of these bright spots are captured as images as shown in FIG. The sampling address on the image relating to each of the luminescent spots thus captured is obtained as follows.

First, as shown in FIG. 10, a pixel having the highest output value is searched for among the pixels in the vicinity of a bright point of a captured image, and the pixel (CCD pixel) is determined as a pixel (Xn ). Attention is paid to this pixel (Xn) and its left and right pixels (Xn-1) and (Xn + 1).
As shown in the figure, a point where the areas of the two shaded portions are equal is defined as the address of the center of the bright spot in the X direction. The same processing is performed for the pixel (Xn) and the pixels above and below the pixel (Xn), and the address of the center of the luminescent spot in the Y direction is obtained. Such an operation is performed for all the bright points, and the center address (X, Y) of each bright point is obtained. Then, for each area divided into 16 areas,
Based on the relative positional relationship between the center addresses (X, Y) of the four corners of the bright point on the captured image and the addresses of the bright points on the LCD panel, LC included in the area is determined.
An address (sampling address) corresponding to a pixel on the D panel is obtained as a real number. On the basis of the sampling address obtained in this way, a correspondence between the test image captured by the CCD camera and the display pixel of the LCD panel has been established. Here, the center address of each luminescent spot on the captured image is obtained. However, as shown in FIG. It may be displayed at a predetermined location, and the center address of the display pattern may be obtained.

Recently, as the number of pixels in an LCD panel has been increased, an image capturing technique using a pixel shift method described below has been proposed in order to perform a more accurate LCD panel test. .

FIG. 13 is a diagram for explaining a pixel shifting method. One CCD has four CCD pixels (2 × 2) for one pixel.
(A pixel enlargement factor of 4) corresponds to pixel shift. This pixel shifting method is C
As for the image of the LCD picture element formed on the imaging surface of the CD camera, the first state without pixel shift (shot a in FIG. 13)
State), a second state in which a predetermined amount of pixel shift has occurred in the X direction from the first state (the state of shot b in FIG. 13),
A third state in which a predetermined amount of pixel shift has occurred in the Y direction from the second state (the state of shot c in FIG. 13), and a predetermined amount of pixel shift in the X direction from the third state (that is, 4) which has a predetermined amount of pixel shift in the Y direction from the state of 1)
(The state of shot d in FIG. 13), and the four captured images captured by each shot are combined. In the example shown in FIG. 13, the same resolution as in the case of capturing an image by 16 (4 × 4) CCD pixels can be obtained. By adopting such a pixel shifting method, it is possible to capture an image at a high resolution with a smaller number of CCD pixels.

[0010]

The LCD panel is composed of SVGA having 800 × 600 pixels and 1024 × pixels.
768 XGA, and the number of pixels is 1240 × 102
4, the number of pixels has increased to SXGA, and with this, LC
In the test of the D panel, it is necessary to improve the accuracy of the address calibration. However, in order to increase the accuracy of address calibration without using the above-described pixel shifting method, it is necessary to increase the CCD pixel density and capture images at high resolution. Is expensive. For this reason, the conventional method not using the pixel shift method has a limit in improving the accuracy of the address calibration.

On the other hand, when the above-described pixel shifting method is adopted, an image can be captured at a high resolution, so that address calibration can be performed with high accuracy. However, conventionally, address calibration is performed under the same conditions as image capturing when performing an LCD panel test. Therefore, there are the following limitations in improving the accuracy of address calibration.

In the pixel shift method, if the number of times of image capture (the number of shots) by pixel shift is increased, an image can be captured at a higher resolution, but on the other hand, an image is captured a plurality of times while shifting pixels. Since it is necessary to capture, it takes time to capture the image. Therefore, if the number of shots is increased in order to increase the accuracy of the address calibration, there is a problem that the test time becomes longer. As described above, in the conventional method, the number of shots at the time of address calibration is limited due to the limitation of the test time. Therefore, as in the case where the above-described pixel shifting method is not used, the accuracy of the address calibration is improved. Had limitations.

An object of the present invention is to provide an address calibration method capable of solving the above-mentioned problem and improving the accuracy of address calibration.

[0014]

According to the present invention, a test image displayed on a flat panel to be inspected is picked up using a CCD camera, and the flat panel is used based on the picked-up image. An address calibration method performed in a test system for performing a display test of the flat panel, wherein a step of displaying a calibration pattern on the flat panel, and an image of the flat panel formed on an imaging surface of the CCD camera. For the display image of the calibration pattern, X
Nx which caused pixel shift in the direction and the Y direction, respectively.
capturing an image by m shots, combining a plurality of captured images captured by the n × m shots, and determining a corresponding address of each display pixel of the flat panel in the combined image, Converting an address in the composite image into an address corresponding to an image captured at the time of the display test.

(Operation) In the case of using the pixel shifting method, although the resolution increases as the number of shots increases, it takes time to capture an image. Conventionally, address calibration has been performed under image capturing conditions when a panel display test is performed, so that the image capturing time is limited. On the other hand, in the present invention, an image is captured at a high resolution using image shifting, and an address obtained from the composite image is converted into an address corresponding to an image captured at the time of a display test. As described above, there is no time restriction on the image capture in the address calibration. Therefore, it is possible to capture images at high resolution by increasing the number of shots. As described above, according to the present invention, it is possible to capture an image at a high resolution, and calibration is performed based on the high-resolution image data thus captured. Is required with high precision.

[0016]

Next, an embodiment of the present invention will be described with reference to the drawings.

FIG. 2 shows a configuration of an LCD panel image quality inspection apparatus provided with a pixel shifting mechanism to which the address calibration method of the present invention can be applied. This LCD panel image quality inspection apparatus uses, for example, a pixel enlargement ratio (CCD pixel / LCD).
Even if only a resolution of 4 or less can be obtained, an image can be captured at a resolution of 4 or more (CCD pixels 2 × 2) for one LCD pixel by adopting the pixel shifting method. It is configured as follows. Hereinafter, this LC
Each configuration of the D panel image quality inspection apparatus will be described.

As shown in FIG. 2, the LCD panel image quality inspection apparatus has a CCD camera 19 for taking in image brightness data above the LCD panel 10 to be inspected.
Are arranged, and an X-direction plane parallel plate 1 is interposed between them.
1, a plane parallel plate 12 in the Y direction, an optical lens unit 13, and a rotating plate 14 provided with four types of optical filters of R, G, B, and colorless and transparent. The rotating plate 14
Is for inspecting a color-compatible LCD panel, and is not necessary for a monochrome-compatible LCD panel.

The plane parallel plate 11 in the X direction, the plane parallel plate 12 in the Y direction, and the rotary plate 15 are connected to motors 15, 16 and 17, respectively, and their rotation is controlled based on a drive signal from a motor control unit 18. . The driving by the motor control unit 18 is controlled by the CPU control unit 23. These X direction plane parallel plate 11 and Y direction plane parallel plate 12
As shown in FIG. 3, when the both sides of the glass plate are polished in parallel to the plane and parallel to the LCD panel 10, the image light of the LCD panel 10 goes straight through the plane parallel glass, If you have a slope to
The image light of the LCD panel 10 is refracted by the plane parallel glass,
A shift of a distance d occurs on the imaging surface of the CCD camera 19.

The output of the CCD camera 19 is supplied to the A / D converter 2
0 is input to the image memory 21, and the contents of the image memory 21 are taken into the image synthesis memory 22 under the control of the CPU control unit 23.

In this image quality inspection apparatus, the resolution can be improved by utilizing the principle of the pixel shift caused by the plane parallel plates 11 and 12 as shown in FIG. Specifically, the X-direction plane parallel plate 11 and the Y-direction plane parallel plate 12
Is tilted with respect to the LCD panel 10 to shift the image (pixel shift) to improve the resolution as follows.

First, as shown in FIG. 4A, both the X-direction plane parallel plate 11 and the Y-direction plane parallel plate 12 are LC
An image is fetched in a state parallel to the D panel 10 (image memory 21) and stored in a predetermined position of a 4 × 4 matrix of the image synthesis memory 22. Subsequently, as shown in FIG. 4B, the X-direction plane parallel plate 11 is tilted to shift the image in the X direction to take in the image, and store it at a predetermined position in the 4 × 4 matrix. Further, FIG.
As shown in (c), the plane parallel plate 12 in the Y direction is
Shift the image in the direction and capture the image, 4x earlier
4 at a predetermined position. Finally, FIG.
As shown in (d), the plane parallel plate 11 in the X direction is returned to a state parallel to the LCD panel 10, the image is shifted only in the Y direction, and the image is captured. Store.

By shifting the image as described above, 2 × 2 image data can be converted to 4 × 4 image data, and image data equivalent to a resolution corresponding to 16 CCD pixels per LCD pixel can be obtained. can get. In this manner, even if the resolution of which the pixel enlargement ratio is less than 4 can be obtained, the image data corresponding to the resolution of the pixel enlargement ratio of 4 or more can be obtained.

Next, the address calibration method of the present invention will be described with reference to the flowchart of FIG.

First, a calibration pattern is displayed on the LCD panel 10 (step S10). Subsequently, the LCD panel 10 on which the calibration pattern is displayed is picked up by the CCD camera 19, and the pattern image displayed on the LCD panel 10 formed on the imaging surface of the CCD camera in the X direction at predetermined intervals. Then, an image is captured by an n × m shot in which a pixel shift occurs in the Y direction (step S11). At this time, the image data is captured at a resolution higher than the resolution of the image capturing performed during the panel test, using the above-described pixel shift. For example, if the image capture at the time of the display test is performed by capturing an image of 4 (2 × 2) shots using pixel shift (see FIG. 4), the image capture in the calibration is performed by 6 (2 × 3 or 3 × 2) shots, 9 (3 ×
3) The number of shots taken during the display test, such as shots, is larger than the number of shots, and the shots are taken at high resolution. In the case where the resolution is increased by increasing the number of shots for image capture in this way, it takes time to capture the image. In the present invention, however, calibration is performed regardless of image capture conditions during a test, as described later. Therefore, there is no time restriction on image capture in calibration.

Hereinafter, as an example, a description will be given of an image capture by 9 (3 × 3) shots using pixel shift.

When taking in 9 (3 × 3) shot images, the angles of the X-direction plane parallel plate 11 and the Y-direction plane parallel plate 12 shown in FIG. 2 with respect to the LCD panel 10 are controlled in two stages. It is done by doing. For example, (1) the Y-direction plane parallel plate 1 with respect to the LCD panel 10
2 is a parallel state, and the X direction plane parallel plate 11 is in a parallel state, a first angle, and a second angle. Board 1
2 is at a first angle, and three shots are obtained by the X-direction plane parallel plate 11 taking three states of parallel, first angle, and second angle. , Y direction plane parallel plate 1
In the state where 2 takes the second angle, 9 shots are obtained by combining each of three shots obtained by taking the three states of the X-direction plane parallel plate 11 being parallel, the first angle, and the second angle. obtain.

As described above, by controlling the plane parallel plate 11 in the X direction and the plane parallel plate 12 in the Y direction, image data of nine shots in which the LCD panel display image is shifted by a predetermined amount in the X direction and the Y direction is captured. Each is stored in a predetermined position of a predetermined matrix (9 × 9) in the image synthesis memory 22 (step S11).

After the image is captured in each shot, the CCD address in the composite image of the image data stored in the image compositing memory 22 is calculated (step S12). Then, the calculated CCD address is converted into a CCD address in the image data captured under the image capturing condition (the number of shots) at the time of a panel test performed later (step S13). In this CCD address conversion, for example, when image capture at the time of address calibration is performed with 9 shots and image capture at the time of a display test is performed with 4 shots, as shown in FIG. The CCD address (300, 300) in the composite image is converted to the CCD address (200, 200) in the composite image of 4 (2 × 2) shots.

As described above, according to the present invention, the CCD address calculated based on the image captured at high resolution is the CCD address in the captured image in the panel display test.
Since it is converted into an address, the image capturing condition (the number of shots) at the time of calibration can be set regardless of the image capturing condition (the number of shots) at the time of the panel display test. Therefore, when performing address calibration, it is possible to capture images at high resolution by increasing the number of shots, and to perform address calibration with high accuracy. For example, as shown in FIG. 6, when an image captured at the time of a panel display test is a CCD image indicated by a solid line, in the address calibration, a detailed CCD image indicated by a broken line and a solid line is used. Is obtained, and the CCD address can be calculated in detail.

[0031]

As described above, according to the present invention,
Since the correspondence between the display pixels of the LCD panel and the CCD pixels can be obtained with higher accuracy, the accuracy of the address calibration can be improved, and there is an effect that a test result with high reproducibility can be obtained.

[Brief description of the drawings]

FIG. 1 is a flowchart for explaining an address calibration method of the present invention.

FIG. 2 is a block diagram showing an embodiment of an LCD panel image quality inspection apparatus having a pixel shift mechanism to which the address calibration method of the present invention is applied.

FIG. 3 is a diagram for explaining the principle of pixel shift.

FIGS. 4A to 4D are diagrams for explaining image capture using pixel shifting.

FIG. 5 is a diagram schematically showing address conversion.

FIG. 6 is a diagram illustrating a comparison between an image captured during a panel display test and an image captured during address calibration.

FIG. 7 is a block diagram showing a general configuration of an LCD panel test apparatus.

FIG. 8 is a diagram illustrating an example of a calibration pattern.

FIG. 9 is a diagram showing an image of capturing the bright spots of the calibration pattern shown in FIG. 8;

FIG. 10 shows each CC in the captured image shown in FIG. 9;
FIG. 4 is a diagram illustrating an output of a D pixel.

FIG. 11 is a diagram for explaining calculation of a center address of a luminescent spot in the captured image shown in FIG. 9;

FIG. 12 is a diagram illustrating an example of a calibration pattern.

FIG. 13 is a diagram for explaining a pixel shifting method.

[Explanation of symbols]

 Reference Signs List 10 LCD panel 11 X direction plane parallel plate 12 Y direction plane parallel plate 13 Optical lens unit 14 Rotating plate 15, 16, 17 Motor 18 Motor control unit 19 CCD camera 20 A / D conversion unit 21 Image memory 22 Image synthesis memory 23 CPU Control unit

Claims (1)

[Claims] 1. An address calibration method which is performed in a test system in which a test image displayed on a flat panel to be inspected is captured using a CCD camera and a display test of the flat panel is performed based on the captured image. Displaying a calibration pattern on the flat panel; and displaying a calibration pattern display image of the flat panel formed on the imaging surface of the CCD camera in the X direction at predetermined intervals. Capturing an image using nxm shots in which pixel shifts have occurred in the Y and Y directions, and combining a plurality of captured images captured by the nxm shots, Determining a corresponding address of each display pixel of the composite image; The address calibration method characterized by having the steps of converting the corresponding address to the image captured at the time of the display test.