JP2595906B2 - Pattern inspection equipment - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a pattern inspection apparatus, and more particularly, to a pattern inspection apparatus in a product inspection process of a liquid crystal image display device.
The present invention relates to a pattern inspection apparatus used for detecting a black spot caused by dust interposed between a CD panel device and a backlight.

[0002]

2. Description of the Related Art In a conventional pattern inspection apparatus of this type, an LCD panel product to be inspected is arranged in a dark place,
This is done by turning on the backlight and observing the presence or absence of black spots by the operator. When automating this pattern inspection device, it is difficult to detect black spots due to uneven illumination of the backlight. There is. As an example of a conventional pattern inspection apparatus, for example,
In an inspection method for a color liquid crystal image display device proposed in Japanese Patent Application Laid-Open No. 993, there is a method capable of detecting a point defect or “unevenness” on a display screen.
No method has been proposed for a method for stably detecting a black point from among the above.

Another conventional example is disclosed in Japanese Unexamined Patent Publication No.
No. 2094 proposes a pattern inspection apparatus.
FIGS. 4A, 4B, 4C, 4D, 4E, and 4F explain the inspection method of the pattern inspection apparatus proposed in the above-mentioned Japanese Patent Application Laid-Open No. 4-52094. FIG. In this conventional example, in order to stably detect a display defect, by taking the difference between two images, "unevenness" caused by a pack light or the like taken in when capturing an image on the LCD is removed. ing. FIG.
(A) and (b) show an all-off pattern and an all-on pattern, respectively, and a defect n is displayed on each pattern. As an inspection method corresponding to these defects, these two patterns are compared, and a difference image between the two patterns is detected. As shown in the comparison pattern in FIG. 4C, the position of the defect n in the all light-off pattern is detected as the “0” level when the image has a low gray level and is binarized. On the other hand, FIG.
In the reference pattern of (d), patterns to be lit are formed in the effective area where the defect inspection is performed, and these comparison patterns and the reference pattern are compared with each other.
As a result, only valid defects n are detected as shown in FIG.

[0004]

In the conventional pattern inspection apparatus described above, in the case of Japanese Patent Application Laid-Open No. Hei 3-146993, the black spots existing in the "unevenness" on the LCD display screen can be stably removed. There is no method for detecting the black spot, and there is a drawback that a black spot which is difficult to separate from the background due to the illumination “unevenness” of the backlight cannot be effectively detected.

Also, in the proposal of Japanese Patent Application Laid-Open No. 4-52094, the difference between two images is calculated to obtain
Although it is possible to remove "unevenness" caused by a backlight or the like taken in when capturing a CD, the backlight irradiates a black point caused by dust or foreign matter existing between the LCD and the backlight. Under the same condition of light, since the level of the black point portion is the same, there is a disadvantage that the black point disappears by taking the difference between the two images, making it impossible to detect the black point. Furthermore, since there is the above-mentioned "unevenness" in one image, there is a disadvantage that it is impossible to separate only the black points by simple binarization processing.

[0006]

Means for Solving the Problems To solve the above problems,
For example, the pattern inspection apparatus of the present invention is a pattern detection apparatus that detects a black spot caused by dust or foreign matter interposed between an LCD panel and a backlight. A backlight for irradiating the LCD panel with light, image data generating means for receiving and transmitting the irradiating light by the inspection object, generating and outputting digital image data, and the image data generating means The image data output from the image data is defined by a predetermined position in a small area of the image data and the brightness of the image data.
Correlation coefficient conversion means for converting and outputting correlation coefficient data indicating a correlation with the degree level, and binarization means for binarizing the correlation coefficient data to generate and output a binary image; with reference to the binarized image, the position of the black spot portion interposed between the backlight and the predetermined LCD panel, in which and a defect position detecting means which detects and outputs as the defect position information.

The image data generating means receives the transmitted light, performs photoelectric conversion, generates and outputs analog image data corresponding to the inspection object, and converts the analog image data into a digital image data. An A / D conversion circuit that converts the data into data and outputs the data, and an image memory circuit that inputs and temporarily stores the digital image data may be provided. A smoothing circuit that smoothes digital image data output from the memory circuit, generates and outputs smoothed image data, and converts the smoothed image data into the smoothed image data.
Between predetermined position in small area and brightness level
And a correlation coefficient conversion circuit that converts the data into correlation coefficient data indicating the following and outputs the result.

[0008]

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

FIG. 1 is a block diagram showing one embodiment of the present invention. As shown in FIG.
For the DUT 1 disposed in the inside, a backlight 2 also disposed in the dark place 12 corresponding to the DUT 1 and a driving signal 101 for lighting the backlight.
Backlight lighting circuit 3 that outputs
Receives the irradiating light 102 output from the backlight 2, inputs the transmitted light 103 output from the test object 1, performs photoelectric conversion, and converts the transmitted light 103 into an analog signal 104 corresponding to image information.
And a photoelectric conversion scanner 4 for outputting the analog signal 1
A / D conversion circuit 5 that inputs and converts the digital signal 105 into a digital signal 105, outputs the digital signal 105, an image memory circuit 6 that inputs and stores the digital signal 105, and receives image data 106 output from the image memory circuit 6. A smoothing circuit 7 for outputting a smoothed image signal 107, a correlation coefficient converting circuit 8 for inputting the smoothed image data 107 and outputting a correlation coefficient data 108, Binarization circuit 9 for binarizing and outputting binary image data 109
And the binary image data 109 are input and the defect position information 11 is input.
It comprises a position detection circuit 10 for detecting and outputting 0, and a defect position information memory circuit 11 for inputting and storing the defect position information 110.

FIGS. 2 (a), (b), (c), (d),
FIGS. 3E and 3F and FIGS. 3A and 3B show the correlation coefficient circuit 8 and the binarization circuit 9 in this embodiment.
It is a figure for explaining operation of. FIG. 2 (d), (e)
And (f) show an example of the positional relationship between the black spots 201 and 202 on the image display screen. In these figures, the coordinate systems of the black spots 201 and 202 are (x ₁ , y ₁ ), respectively. And (x ₂ , y ₁ ). FIG. 2A shows the smoothing circuit 7.
FIG. 14 shows smoothed image data 107 output from
Corresponding to FIG. 2 (d), the distribution of luminance levels (smoothed image data 107) in the line of y = y ₁ are shown along the x-axis. FIG. 2B shows the correlation coefficient data 108 output from the correlation coefficient conversion circuit 8.
Corresponding to (e), the distribution of the luminance level correlation coefficient data 108 on the line of y = y ₁ is also shown along the x-axis. FIG. 2C shows a binary image 109 output from the binarization circuit 9 via the binarization level “L” shown in FIG. 2B, and FIG. ), The distribution of the binary image 109 on the line of y = y ₁ is also shown along the x-axis. Also, FIG.
FIG. 3B is a diagram illustrating an arrangement relationship of pixels in a window size, and FIG. 3B is a diagram illustrating a correlation coefficient relationship expression for calculating linearity.

In FIG. 1, in a correlation coefficient conversion circuit 8, smoothed image data 1 output from a smoothing circuit 7 is output.
07 (see FIG. 2A) is converted into correlation coefficient data 108 (see FIG. 2B) in the correlation coefficient conversion circuit 8 and input to the binarization circuit 9. In this case, in the correlation coefficient conversion circuit 8, FIGS. 2A and 3A
As shown in FIG. 3B, a window is set, and the correlation coefficient of the luminance with respect to the position x is set as a linearity in the window, and the window is sequentially moved one pixel at a time to obtain the phase relationship. Numerical data 108 is obtained. In the binarization circuit 9, the correlation coefficient data 108 is binarized to generate and output a binary image 109. As shown in FIG. If there is almost no noise in
The change in the brightness level of “unevenness” on the display screen occurs over a wide range, and the black spot portions 201 and 202 change.
Since the change in the brightness level occurs within a narrow range, the linearity of the brightness level is used to emphasize the black spots 201 and 202 in FIG.
This is performed as shown in FIG. More specifically, since the correlation coefficient is used in the calculation of the linearity, the linearity increases in the “uneven” portion on the display face, which changes gradually, and the black spots 201 and 202 are increased. Has a low linearity. Conversely, in the case of image data with a lot of noise, the linearity in the range including the black spots 201 and 202 is high, and the linearity in the surroundings is low, so that it is possible to differentiate from noise. Further, when a differential value is used as in the related art, it becomes difficult to distinguish between the black spots 201 and 202 and noise because the differential value is used as a reference for the difference from the surroundings.

Next, in the binarization circuit 9 to which the correlation coefficient data 108 is input, the correlation coefficient data 108 is binarized and, as shown in FIG.
Only the binary image 109 corresponding to the positions of 1 and 202 is output in a state where it remains. The binary image 109 is input to the position detection circuit 10, where the black dots 201 and 20 are output from the position detection circuit 10.
2 is output as defect position information 110 and is input to and stored in the defect position information memory circuit 11. Therefore, in the present embodiment, it is possible to emphasize the black spot portion without being affected by the noise level and irrespective of the magnitude of the differential value, as compared with the case where the conventional differential value is used. It is possible, and a black spot on the display screen can be stably detected.

[0013]

As described above, the present invention is applied to a pattern inspection apparatus for detecting a black spot caused by dust between an LCD panel device and a backlight in a product inspection process. The correlation coefficient data of the image data obtained through the transmitted light from the inspection object is obtained, and the correlation coefficient data is binarized through a predetermined level, so that the correlation coefficient data is not affected by noise. There is an effect that the position of the black spot can be reliably detected.

[Brief description of the drawings]

FIG. 1 is a block diagram showing one embodiment of the present invention.

FIG. 2 is an explanatory diagram of an operation of a correlation coefficient circuit and a binarization circuit in the present embodiment.

FIG. 3 is a diagram illustrating calculation of linearity and a window size.

FIG. 4 is a schematic diagram showing a display change method of a conventional example.

[Explanation of symbols]

 REFERENCE SIGNS LIST 1 inspection object 2 backlight 3 backlight lighting circuit 4 photoelectric conversion scanner 5 A / D conversion circuit 6 image memory circuit 7 smoothing circuit 8 correlation coefficient conversion circuit 9 binarization circuit 10 position detection circuit 11 defect position information memory Circuit 12 dark place

Claims (3)

(57) [Claims] 1. A pattern detecting apparatus for detecting a black spot caused by dust or foreign matter interposed between an LCD (liquid crystal display) panel and a backlight, wherein the LCD as an inspection object arranged in a predetermined dark place
A backlight that generates light for irradiating the panel; an image data generating unit that captures an image by receiving the transmitted light of the irradiated light through the inspection object, generates and outputs digital image data; and outputs the image data by the image data generating unit. The image data to be processed is defined by a predetermined position and brightness within a small area in the image data.
A correlation coefficient conversion unit that converts the correlation coefficient data into a correlation coefficient data indicating a correlation with a level and outputs the correlation coefficient data; a binarization unit that binarizes the correlation coefficient data to generate and output a binary image; Defect position detecting means for detecting and outputting, as defect position information, a position of a black spot interposed between a predetermined LCD panel and the backlight with reference to the binarized image. Pattern inspection device. 2. A photoelectric conversion scanner, wherein said image data generating means receives and transmits said transmitted light, generates and outputs analog image data corresponding to said inspection object, and converts said analog image data into a digital image. 2. The pattern inspection apparatus according to claim 1, further comprising an A / D conversion circuit that converts the data into data and outputs the data, and an image memory circuit that inputs and temporarily stores the digital image data. 3. The image processing apparatus according to claim 2, wherein the correlation coefficient conversion unit smoothes digital image data output from the image memory circuit,
A smoothing circuit for generating and outputting smoothed image data, and converting the smoothed image data into small data in the smoothed image data.
The pattern inspection apparatus according to claim 1, further comprising: a correlation coefficient conversion circuit that converts and outputs correlation coefficient data indicating a correlation between a predetermined position in the area and the luminance level .