JP4613662B2 - Edge defect detection method, edge defect detection apparatus, edge defect detection program, recording medium - Google Patents

Description

  The present invention relates to an edge defect detection method, an edge defect detection apparatus, an edge defect detection program, and a recording medium. For example, the present invention relates to a method for detecting a defect generated at an edge of a display pixel.

Conventionally, an image display device such as an organic EL panel is known (for example, Patent Document 1).
An organic EL panel displays an image on a display surface using a plurality of organic EL elements arranged in a matrix as display pixels.
The organic EL element has, for example, an organic laminated film having a light emitting layer containing a fluorescent organic compound of red, green, or blue, and an anode electrode and a cathode electrode that sandwich the organic laminated film. These organic EL elements are active elements that self-emit with a luminance corresponding to the amount of supplied current, and are formed along a plurality of scanning lines formed along the rows of these organic EL elements and the columns of these organic EL elements. Control is performed by current control circuits arranged in the vicinity of the intersections of the plurality of data lines.
Here, in the manufacture of an organic EL panel, a vapor deposition method may be used for a low molecular type and an ink jet method may be used for a high molecular type, but in any case, the periphery of each display pixel is often surrounded by a black matrix.

In such an image display device, display defect inspection is performed to inspect whether display defects have occurred in the manufacturing process.
For example, in the above-mentioned organic EL panel, a defect that shines in a dot shape may occur outside the display pixels where a black matrix should be originally applied to become black, which is called an edge defect (for example, FIG. 5). reference).
In the display defect inspection for such edge defects, conventionally, an inspector visually inspects the organic EL panel that is lit in a predetermined pattern to inspect the organic EL panel.

JP 2004-191752 A

However, if the inspector visually inspects the edge defect for each organic EL panel, it takes a very long time to inspect each piece of the organic EL panel, and the production amount of the organic EL panel is enormous. There is a limit to visually inspecting all organic EL panels.
Further, when visual inspection is performed by a plurality of inspectors, there is a problem in that it is difficult to keep the reliability of the product constant due to variations in the judgment of edge defects among the inspectors.

  An object of the present invention is to provide an edge defect detection method, an edge defect detection device, an edge defect detection program, and a recording medium on which this program is recorded, which can solve the conventional problems and can automatically detect edge defects appropriately. .

The edge defect detection method of the present invention is an edge defect detection method for detecting an edge defect occurring in a display pixel of a display panel, and has an imaging pixel finer than a display pixel of a display image of the display panel to be inspected. an imaging step of imaging the display image by the imaging means, a luminance value calculating step of calculating a luminance value L _i of each of the imaging pixels based on the imaging data captured by the imaging step, imaging the display pixels The brightness of the selected imaging pixel of interest is selected when scanning is performed to sequentially select the imaging pixels in the imaging data obtained along the predetermined direction and the base end side in the scanning direction is set to the top and the tip end side is set to the bottom. the value L between the luminance values L _i-1 of the imaging _{pixel i} located just above the interest imaging pixels of Toko is the luminance value L _i of the noted image pickup pixel satisfies predetermined conditions luminance value of the imaging pixel just above the Of difference from L _i-1 A difference value calculating step of calculating an absolute value as a difference value D _i , a maximum difference value searching step of searching for the maximum difference value D _max by examining an imaging pixel adjacent to the imaging pixel of _interest , and the maximum difference value search and emphasizing the difference value calculation step of the searched maximum of the difference value D _max to calculate the enhancement difference value _D'i is added to the difference value D _i of the noted image pickup pixel at step, the emphasis difference of each imaging pixel A defective portion detection step of comparing the value D ′ _i with a predetermined luminance difference threshold value D ′ _s and detecting an imaging pixel in which the enhancement difference value D ′ _i is greater than or equal to the luminance difference threshold value D ′ _s as a defective portion; It is characterized by providing.

In such a configuration, first, a display image to be inspected is imaged by an imaging means (imaging process). The imaging pixels of the imaging means are finer than the display pixels constituting the display image, and imaging data obtained by imaging the display image in units of imaging pixels obtained by further subdividing the display pixels of the display image is obtained.
Subsequently, in the luminance value calculation step, the luminance value L _i is obtained for each imaging element of the imaging data.
Next, for each display pixel that is the processing area for edge defect detection, the imaging pixel is scanned from top to bottom for each column, and the difference that is the absolute value of the luminance difference from the imaging pixel immediately above it for each imaging pixel A value D _i (= L _i−1 −L _i ) is calculated (difference value calculation step).
However, in the difference value calculation step, it is checked whether or not the relationship of the luminance value L with the imaging pixel immediately above it for each imaging pixel matches a predetermined condition, and the imaging pixel only when this predetermined condition is met. The difference value D _i is calculated for. Here, for an image sensor that does not satisfy the predetermined condition, the difference value D _i is set to 0 as an example.

  The display pixel means an area of the display pixel including an edge portion where the edge defect occurs in order to detect the edge defect.

When the difference value is obtained for each imaging pixel, the maximum difference value D is searched from the imaging pixels adjacent to the imaging pixel of interest.
Then, the searched difference value D _max is added to the difference value D _i of the imaging pixel of interest to calculate the enhancement difference value D ′ _i (= D _i + D _max ).
In the defect detection step, the enhancement difference value D ′ _i obtained for each imaging pixel is compared with a predetermined luminance difference threshold value D ′ _s, and the imaging pixel having the enhancement difference value D ′ _i equal to or greater than the luminance difference threshold value D ′ _s. Is detected. Thereby, an edge defect is detected.

The display image is composed of a plurality of display pixels arranged in a matrix, and each display pixel is surrounded by a black matrix. The edge defect is a defect that occurs as a defect that shines in a dot shape at the edge portion of the display pixel.
Therefore, when you look from the top to the bottom of the area where the edge defect that shines like dots in the part that should originally be black as the black matrix is seen in order from the top, there is a black part of the black matrix, and there is an edge defect part that suddenly becomes brighter , There will also be black parts of the black matrix.
When such an edge defect is automatically detected based on imaging data, the edge defect can be detected based on a change in luminance value when the imaging pixel is viewed from above.

In other words, when there are point-like bright spots in the black matrix as edge defects, if the difference in luminance value between the upper and lower imaging pixels is calculated by scanning the imaging pixels from the top to the bottom, the edge defect is changed from the black matrix portion. The difference in luminance value becomes very large when entering and exiting from the edge defect to the black matrix.
On the other hand, if there is no edge defect, the black matrix portion is uniformly black, so the difference in luminance value between the upper and lower imaging pixels is small, and the luminance value is uniform even in the display pixels. The difference is small.
Therefore, if there is a portion where the difference in luminance value is large, it can be detected that an edge defect exists.
Therefore, after obtaining the enhancement difference value D ′ _i based on the luminance difference between the upper and lower imaging pixels, the enhancement difference value D ′ _i is compared with a predetermined luminance difference threshold D ′ _s (defective portion detection step), An edge defect can be detected by detecting an imaging pixel having an enhanced difference value D ′ _i that is equal to or greater than the luminance difference threshold D ′ _s (defect portion detection step).

Here, by calculating the absolute value (difference value D _i ) of the luminance difference above and below the imaging pixel and simply comparing this difference value D _i with a threshold value, it is not merely an edge defect but merely a luminance variation. It reacts sensitively to existing abnormalities and electrical signal noise of the imaging means, and it is impossible to accurately detect only edge defects.
In this regard, in the present invention, the difference value D of the imaging pixel adjacent to the target imaging pixel is observed in the maximum difference value search step, and the maximum difference value D _max among these is added to the difference value D _i of the target imaging pixel. and in terms of calculating the emphasized difference value _D'i and compares this enhancement difference value _D'i and luminance difference threshold _D's.
As a result, there is a large difference value D in the position adjacent to the imaging pixel of interest, not mere brightness variations or electrical signal noise that are not edge defects, and a region with abnormal brightness is properly detected as an edge defect. can do.

Further, when the difference value D _i is calculated between the imaging pixel of interest and the imaging pixel immediately above it in the difference value calculation step, the luminance is similarly high at the boundary between the black matrix and the edge defect and at the boundary between the black matrix and the display pixel. since the change has occurred, becomes a large value in the same way as the difference value D _i. Therefore, when the difference value D _i is simply compared with the threshold value, not only the edge defect but also the normal boundary between the black matrix and the display pixel is detected, and the normal display pixel is erroneously detected excessively. It will be.

In this regard, in the present invention, the difference value D between the upper and lower imaging pixels is calculated only when the relationship between the luminance value of the target imaging pixel and the luminance value of the imaging pixel immediately above the target imaging pixel satisfies a predetermined condition. A difference value calculation step is performed. Then, the emphasis difference value D _i in this manner based on the calculated difference value D as compared to the luminance difference threshold _D's, defective imaging pixel of the enhanced difference value _D'i is luminance difference threshold _D's or more It detects as a part (defective part detection process).
Thereby, it is made not to react to the boundary of a black matrix and a display pixel, but to react only to an edge defect.

For example, in the case of scanning the upper half side of the display pixel, when the luminance value changes from a low state to a high state when shifting from the black matrix portion to the display pixel, the black matrix is changed to the display pixel. It is normal because it has migrated.
On the other hand, when an edge defect occurs in the black matrix portion, the luminance value changes from a low state to a high state after entering the edge defect from the black matrix, and then when the edge defect exits the black matrix. It is abnormal when the value changes from a high state to a low state.

As described above, since the change in the luminance value due to the normal transition from the black matrix to the display pixel is different from the change in the luminance value caused by the presence of the edge defect, the luminance between the upper and lower imaging pixels is based on this difference. By performing the difference value calculation step only when the value relationship matches a predetermined condition, a high difference value _Di can be obtained only at the edge defect portion. Then, by comparing the enhancement difference value _D'i based on the difference value D _i calculated in this manner to a predetermined brightness difference threshold _D's, it is possible to appropriately detect an edge defect.

In the present invention, the predetermined condition for performing the difference value calculating step is that the luminance value L _i-1 of the imaging pixel immediately above the imaging pixel of interest when the imaging pixel of interest is located in a region on the upper half side of the display pixel. Is higher than the luminance value L _{i of the} target imaging pixel, and when the target imaging pixel is located in the lower half of the display pixel, the luminance value L _i-1 of the imaging pixel immediately above the target imaging pixel. Is preferably lower than the luminance value L _{i of the} imaging pixel of interest.

In such a configuration, when the luminance value of the imaging pixel is viewed from the top to the bottom in order, if it is the upper half of the display pixel, the luminance value starts from a low luminance value when moving from the black matrix to the display pixel. Transitioning to a higher value is a normal change.
On the other hand, when there is an edge defect, after entering the edge defect from the black matrix, when the edge defect exits the black matrix, the brightness value shifts from a high brightness value state to a low brightness value state.
Therefore, on the upper half side of the display pixel, it is normal for the luminance value to shift from a low state to a high state when entering the display pixel from the black matrix, but abnormal when the luminance value changes from a high state to a low state. is there.
Therefore, for the upper half of the display pixel, only when the luminance value changes from a high state to a low state, the luminance value L _{i−1 of} the imaging pixel located immediately above the target imaging pixel and the luminance value L of the target pixel to calculate the difference value D _i and _i. Then, the difference value D _i is calculated only when there is an abnormality in which the luminance value changes from a high state to a low state.
As a result, it is possible to appropriately detect the boundary between the edge defect and the black matrix, while avoiding erroneous detection of a normal change that shifts from the black matrix to the display pixel.

In addition, when the luminance value L of the imaging pixel is viewed from the top to the bottom in the case of the lower half of the display pixel, the luminance value L is low from the state where the luminance value is high when exiting from the display pixel to the black matrix. Transitioning to a state is a normal change. On the other hand, when there is an edge defect, when the display pixel enters the black matrix and then enters the edge defect from the black matrix and shifts from a low luminance value L to a high luminance value L state. This is an unusual change. Therefore, in the lower half of the display pixel, it is normal for the luminance value L to shift from the high state to the low state when exiting from the display pixel to the black matrix, while the luminance value L is low to the high state. It is abnormal to change to.
Therefore, for the lower half of the display pixel, only when the luminance value L changes from a low state to a high state, the luminance value L _{i-1 of} the imaging pixel located immediately above the target imaging pixel and the luminance value of the target pixel A difference value D _i from L _i is calculated.
Then, the difference value D _i is calculated only when there is an abnormality in which the luminance value L _i changes from a low state to a high state.
As a result, it is possible to appropriately detect the boundary between the edge defect and the black matrix, while avoiding erroneous detection of a normal change that shifts from the display pixel to the black matrix.

  As an example, for an imaging pixel that does not satisfy the predetermined condition, the difference value D is set to 0.

In the present invention, in the defective portion detection step, an imaging pixel having a luminance value L _{i that} is equal to or greater than a predetermined luminance threshold value L _s and the enhancement difference value D ′ _i is equal to or greater than a predetermined luminance difference threshold value D ′ _s. A defect candidate pixel search step for searching as a defect candidate pixel, and the enhancement difference value D ′ for the imaging pixel adjacent to the defect candidate pixel searched in the defect candidate pixel search step is equal to or greater than a predetermined luminance difference threshold D ′ _s It is preferable to include a defective pixel detection step of detecting the defective candidate pixel as a defective pixel when there is a certain imaging pixel.

In such a configuration, the luminance value L and the enhancement difference value D ′ are calculated for the imaging pixel. First, the luminance value L _i is greater than or equal to the luminance threshold value L _s , and the intensity difference value D ′ _i is the luminance difference. An imaging pixel that is equal to or greater than the threshold value D' _s is searched for as a defect candidate pixel (defect candidate pixel search step).
If there is a defective candidate pixel, the imaging pixel adjacent to the defective candidate pixel is viewed to see if there is an imaging pixel whose enhancement difference value D ′ is greater than or equal to the luminance difference threshold D ′ _s .
Then, when an enhancement difference value D ′ greater than or equal to the luminance difference threshold D ′ _s exists in the adjacent imaging pixel, the defective candidate pixel is determined as a defective pixel.
In the case where there is a single imaging pixel in which the luminance value L and the enhancement difference value D ′ are each equal to or greater than a predetermined threshold value (luminance threshold value L _s , luminance difference threshold value D ′ _s ), the defect type is not an edge defect, Since it may be noise during imaging, it may be preferable not to detect the edge defect. Therefore, in the present invention, only when the emphasis difference values of adjacent imaging pixels D'even exceeds the luminance difference threshold _D's, By detected as a defective pixel, the luminance difference threshold in adjacent positions _D's Only when the above-described enhancement difference value D ′ is present and there is an abnormality in luminance can be appropriately detected as an edge defect.

In the present invention, the defect detection step includes a threshold setting step for setting the luminance threshold L _s and the luminance difference threshold D ′ _s, and the threshold setting step calculates an average luminance value M of each display pixel. An average luminance value calculating step, a luminance threshold value setting step for setting, as the luminance threshold value L _s , a value obtained by multiplying the average luminance value M calculated in the average luminance value calculation step by a first predetermined coefficient α; And a luminance difference setting step of setting, as the luminance difference threshold D ′ _s , a value obtained by multiplying the average luminance value M calculated in the average luminance value calculating step by a second predetermined coefficient β.

In such a configuration, when setting the luminance threshold value L _s and the luminance difference threshold value D ′ _s , first, an average luminance value M within the processing area of the display pixel is calculated.
Then, to calculate the brightness threshold L _s and luminance difference threshold _D's by multiplying each by a predetermined coefficient to the average brightness value M.
According to such a configuration, an appropriate threshold value can be set for each display pixel. For example, even when there is luminance unevenness such as a difference in luminance between the left and right of the display image, an appropriate threshold value is used for each display pixel. Edge defects can be detected.

The edge defect detection device of the present invention is an edge defect detection device that detects edge defects that occur in display pixels of a display panel, and has finer imaging pixels than the display pixels of the display image of the display panel to be inspected. imaging means for imaging the above display image, a luminance value calculating means for calculating a luminance value L _i of each of the imaging pixels based on the imaging data imaged by the imaging means, acquired by imaging the display pixels When the scanning of the imaging pixels in the obtained imaging data is performed while paying attention to the imaging pixels in order along a predetermined direction, the luminance of the imaging pixel of interest is when the base end side in the scanning direction is set to the top and the tip side is set to the bottom. the value L between the luminance values L _i-1 of the imaging _{pixel i} located just above the interest imaging pixels of Toko is the luminance value L _i of the noted image pickup pixel satisfies predetermined conditions luminance value of the imaging pixel just above the the absolute value of the difference between L _i-1 Difference value calculating means for calculating a frequency value D _i, and the maximum difference value search means for searching the maximum of the difference value D _max examine the imaging pixels adjacent to the focused imaging pixel, in the maximum differential value search circuit and emphasizing the difference value calculating means for calculating the enhancement difference value _D'i searched maximum of the difference value D _max is added to the difference value D _i of the noted image pickup pixels, wherein the emphasis difference value of each imaging pixel D' a defect portion detecting means for comparing _i with a predetermined luminance difference threshold value D ′ _s and detecting an imaging pixel in which the enhancement difference value D ′ _i is equal to or greater than the luminance difference threshold value D ′ _s as a defective portion. Features.

The edge defect detection program of the present invention is an edge defect detection program for detecting an edge defect occurring in a display pixel of a display panel, and has an image pickup pixel finer than a display pixel of a display image of the display panel to be inspected. incorporate computer to the display image to the edge defect detection apparatus having an imaging means for imaging, the computer calculates the brightness value L _i of each of the imaging pixels based on the imaging data imaged by the imaging means Luminance value calculating means and scanning that focuses on the imaging pixels in order along a predetermined direction with respect to the imaging pixels in the imaging data obtained by imaging the display pixels, with the base end side in the scanning direction as the upper side when the distal end side and lower, between the luminance values L _i-1 of the imaging pixel located immediately above the interest imaging pixel of the luminance value L _i Toko of interest imaging pixel satisfies a predetermined condition A difference value calculating means for calculating the absolute value of the difference between the luminance value L _i of the target image-capturing pixels of interest in case the brightness value L _i-1 of the imaging pixel just above As the difference value, adjacent to the focused imaging pixels and the maximum difference value search means for searching the maximum of the difference value D _max examine the imaging pixel, the difference value D _i of the noted imaging pixel maximum the difference value D _max, which is searched by the maximum difference value search circuit the enhancement difference value compared to the emphasis difference value calculating means for calculating a highlight difference value _D'i, the said enhancement difference value _D'i of each imaging pixel with a predetermined brightness difference threshold _D's added to D' It is characterized by functioning as a defective portion detecting means for detecting an imaging pixel in which _i is equal to or greater than the luminance difference threshold value D′ _s as a defective portion.

  The recording medium of the present invention is characterized in that the edge defect detection program is recorded so as to be readable by a computer.

According to such a configuration, the same effects as those of the above-described invention can be achieved.
If a program is configured to cause a computer having a CPU (central processing unit) and a memory (storage device) to function as each of the above-described units, the parameters in each unit can be easily changed. Then, the program may be recorded on a recording medium, and the recording medium may be directly inserted into the computer to install the program in the computer. A reading device that reads information on the recording medium is externally attached to the computer, and the reading device The program may be installed on the computer.
The program may be supplied and installed on a computer via a communication line such as the Internet, a LAN cable, a telephone line, or wirelessly.

DESCRIPTION OF EMBODIMENTS Hereinafter, embodiments of the present invention will be illustrated and described with reference to reference numerals attached to respective elements in the drawings.
(First embodiment)
A first embodiment according to the edge defect detection method of the present invention will be described.
FIG. 1 is a diagram illustrating an overall configuration of an edge defect detection apparatus 100 that performs an edge defect detection method.
The edge defect detection apparatus 100 supplies a test signal to the organic EL panel 400 on which the organic EL panel 400 to be inspected is mounted, and drives the organic EL panel 400 to display the display surface 410. An OLED drive unit 120 for displaying an image on the screen, a CCD camera 130 as an imaging means for imaging the display surface 410 of the organic EL panel 400, and edge defects in the image are detected by image processing based on imaging data from the CCD camera 130 An arithmetic processing unit 200, a display unit 140 for displaying a detection result by the arithmetic processing unit 200, and an input unit 150 connected to the arithmetic processing unit 200.

  The CCD camera 130 has an image sensor finer than the display pixel 420 of the organic EL panel 400, and can capture one display pixel with a plurality of image pickup pixels. For example, one display pixel is imaged with 20 × 20 imaging pixels. The CCD camera 130 captures an image with a gradation value (255 gradations) that is more accurate than the gradation of the organic EL panel 400. The image data captured by the CCD camera 130 is sent to the arithmetic processing unit 200.

FIG. 2 is a diagram illustrating a configuration of the arithmetic processing unit 200.
The arithmetic processing unit 200 includes an imaging data storage unit 210, a luminance value calculation unit 220, a processing area extraction unit 230, a processing area selection unit 240, a luminance enhancement unit 250, and a defect portion detection unit 300. Yes. The luminance enhancement unit 250 includes a difference value calculation unit 251, a maximum difference value search unit 252, and an enhancement difference value calculation unit 253. The defect portion detection unit 300 includes a threshold setting unit 310, a defect candidate pixel, and the like. A search unit 320; an adjacent pixel search unit 330; and a flag setting unit 340. The threshold setting unit 310 includes an average luminance value calculation unit 311, a luminance threshold setting unit 312, a luminance difference threshold setting unit 313, It has.
For example, by installing and executing a predetermined edge defect detection program in a computer having a CPU (central processing unit) and a memory (storage device), the computer can function as the arithmetic processing unit 200.
The operation of the arithmetic processing unit 200 will be described with reference to the flowchart of FIG.

(Edge defect detection method)
FIG. 3 is a flowchart showing the procedure of the edge defect detection method.
In detecting an edge defect occurring in a display pixel of the organic EL panel 400, first, a display image of the organic EL panel 400 is picked up by the CCD camera 130 in an image pickup step ST100. At this time, an inspection drive signal is applied from the OLED drive unit 120 to the organic EL panel 400 so that each display pixel of the organic EL panel 400 emits light with a predetermined luminance.

FIG. 4 is a diagram illustrating a configuration of the display surface 410 of the organic EL panel 400. As shown in FIG. 4, R, G, and B display pixels 420 are arranged in order on the display surface 410 of the organic EL panel 400, and each display pixel 420 is surrounded by a black matrix 430. FIG. 5 is an example of image data captured by the CCD camera 130. FIG. 6 is a diagram in which only the R display pixels 420 are extracted from the imaging data and rearranged.
When the display pixels 420 are arranged on the display surface 410 of the organic EL panel 400, the image pickup device of the CCD camera 130 is finer than the display pixels 420. Therefore, as shown in FIG. Are imaged by being divided by a plurality of imaging pixels 131.
In the imaging data, a luminance value is obtained for each imaging pixel 131.

When the imaging range of the CCD camera 130 is smaller than the display surface 410 of the organic EL panel 400, the display surface 410 of the organic EL panel 400 is divided into a plurality of parts and captured by the CCD camera 130, and then each part is captured. The image data may be connected in correspondence with the display image.
The captured image data is stored in the image data storage unit 210.

Here, as shown in FIG. 6, each display pixel 420 constituting the display image is an oval type having a longitudinal direction in the vertical direction in FIG. 6, and each display pixel 420 is surrounded by a black matrix 430. Yes. Then, an edge defect 440 in which a defect shining in a dot shape is seen at a position slightly deviated from the oval display pixel with respect to the oval display pixel 420 if normal.
FIG. 6 is an example in which an edge defect 440 is generated mainly at the upper left of the display pixel 420. The edge defect is a defect generated at an edge portion of the display pixel 420, such as slightly above and slightly below the display pixel 420. It is.
In the following processing, such an edge defect 440 is detected based on the imaging data.

Next, in ST110, the luminance value calculation unit 220 performs a luminance value calculation step of calculating the luminance value L for each imaging pixel 131 based on the imaging data. Brightness value calculating means 220, created (see FIG. 7) the sequence space 221 of the same sequence as the imaging pixel 131 shown in FIG. 5, will place the luminance values L _i to each pixel in the sequence space.
Here, FIG. 8 is a numerical example of the luminance value L of each imaging pixel 131 calculated in this way.

  The array space 221 is created by the luminance value calculation means 220, and the cells of the array space 221 are different from the imaging pixels 131 of the imaging data captured by the CCD camera 130, but in the following description. The cells in the array space 221 will also be referred to as imaging pixels.

Next, in ST120, a processing area extraction unit 230 performs a processing area extraction process for extracting a processing area.
The edge defect 440 is a defect that occurs in the edge portion of the display pixel 420. Therefore, when detecting the edge defect 440, it is not necessary to detect the edge defect in the entire display image, and the defect detection is mainly performed in the oval display pixel 420 portion. What is necessary is just to process. Therefore, in the processing area extraction step (ST120), only the imaging pixels 131 having the luminance value L equal to or higher than the predetermined threshold are extracted and marked as the processing target area 231. Then, the area of each oval display pixel 420 is extracted as each processing area 231.
Note that the threshold for extracting the processing area 231 is set to a sufficiently low value to include the display pixel 420 and the edge portion of the display pixel 420 in the processing area 231, for example, about 30. .

  Then, when the edge defect detection processing is sequentially performed on each processing area 231, a processing area selection step for selecting one processing area 231 is performed by the processing area selection means 240 in ST 130.

Edge defect detection processing is specifically performed on the processing area 231 selected in the processing area selection step of ST130.
First, in ST140, the luminance enhancement unit 250 performs a luminance enhancement step of calculating an enhancement difference value based on the luminance difference between the upper and lower imaging pixels 131.
The brightness enhancement step ST140 will be described with reference to the flowchart of FIG.

In the brightness enhancement step (ST140), first, a difference value calculation unit 251 performs a difference value calculation step of calculating a brightness difference between the upper and lower imaging pixels 131 in ST141.
At this time, as shown in FIG. 10, the processing area 231 is scanned along the vertical direction (predetermined direction) for each column. Then, the absolute value (| L _i-1 −L _i |) of the difference between the luminance value L _i of the target imaging pixel 131 of interest and the luminance value L _i−1 of the imaging pixel 131 immediately above it is calculated. Further, a value obtained by dividing the absolute value of this difference (| L _i-1 −L _i |) by L _i−1 and then multiplying by 255 is normalized to be a difference value D _i .

However, this difference value calculation step ST141 is performed only when the following conditions are met.
That is, when the interest imaging pixels 131 are positioned on one half side of the upper region of the display pixel 420, the luminance value L _i of the luminance values L _i-1 of the imaging pixels 131 immediately above the interest imaging pixel 131 is focused imaging pixel 131 The difference value D _i is calculated only when it is higher.
Further, when the attention capturing pixels 131 are located on the lower half side area of the display pixel 420, the luminance value L _i of the luminance values L _i-1 of the imaging pixels 131 immediately above the interest imaging pixel 131 is focused imaging pixel 131 The difference value D _i is calculated only when it is lower.

In performing the difference value calculation step ST141, it is determined whether the imaging pixel 131 of interest is above or below the processing area 231. For example, when the coordinates of 000 to 037 are set in the vertical direction of the processing area 231 as in the numerical example of FIG. 8, if the coordinates of the imaging pixel 131 of interest is 018 or more, it is positioned in the upper half of the processing area 231. If the coordinate of the imaging pixel 131 of interest is 019 or less, it is determined that it is located in the lower half of the processing area 231.
When the target imaging pixel 131 is on the upper half side of the processing area 231, the luminance value L _i of the target imaging pixel 131 and the luminance value L _i-1 of the imaging pixel 131 immediately above the target imaging pixel 131 are shown in FIG. When the brightness value L _i-1 of the imaging pixel 131 directly above is higher than the brightness value L _i of the imaging pixel 131, as confirmed in the luminance value array space shown in FIG. A difference value D _i of luminance values from the imaging pixel 131 positioned immediately above is calculated.

When the target imaging pixel 131 is on the lower half side of the processing area 231, the luminance value L _{i of the} target imaging pixel 131 and the luminance value L _i− of the imaging pixel 131 positioned immediately above the target imaging pixel 131 are _{displayed. 1} is confirmed in the luminance value array space shown in FIG. 7 (FIG. 8), and if the luminance value L _i-1 of the imaging pixel 131 directly above is lower than the luminance value L _i of the imaging pixel 131 of interest, A difference value D _i of luminance values between the imaging pixel 131 and the imaging pixel 131 located immediately above the imaging pixel 131 is calculated.

  When performing the difference value calculation step (ST141), if the target imaging pixel 131 does not meet the predetermined condition, the difference between the target imaging pixel 131 and the immediately preceding imaging pixel 131 is not calculated. The value is 0.

The difference value D is calculated for the imaging pixel 131 satisfying the predetermined condition in the processing area 231 by (Equation 1). Then, an array space having the same array as the imaging pixels 131 in FIG. 5 is prepared, and the calculated difference values D _i are arranged in this array space (see FIG. 11).
FIG. 12 is a numerical example of the difference value D _i of each imaging pixel 131 calculated in this way.

Here, the difference value calculation step ST141, the relationship between the luminance values L _i-1 of the imaging pixel 131 located directly above the brightness value L _i between the target image-capturing pixels 131 of interest imaging pixel 131 only when a predetermined condition is satisfied The calculation of the difference value D _i is based on the following reason.
That is, taking the case where the upper half side of the display pixel 420 is scanned as an example, when the luminance value changes from a low state to a high state when shifting from the black matrix 430 portion to the display pixel 420, the black matrix 430. Since the display pixel 420 is shifted to the display pixel 420, it is normal.
On the other hand, when the edge defect 440 is generated in the black matrix 430, the luminance value is changed from the low state to the high state after entering the edge defect 440 from the black matrix 430, and then the black defect 440 is further transferred from the edge defect 440 to the black defect. When the luminance value changes from a high state to a low state when entering the matrix 430, it is abnormal.
As described above, the change in the luminance value due to the normal transition from the black matrix 430 to the display pixel 420 is different from the change in the luminance value caused by the presence of the edge defect 440. Therefore, if the difference value is calculated only when the predetermined condition is satisfied, a high value is obtained only in a dark valley-like part (reference numeral 441 in FIG. 6) existing between the edge defect 440 part and the display pixel 420. This is because the enhanced difference value D ′ _i can be obtained.

When the difference value D _i has been calculated for each imaging pixel 131, next, in ST 142, a maximum difference value search step for searching for the maximum difference value D _max among the imaging pixels 131 adjacent to the imaging pixel 131 of interest is performed. This is performed by the maximum difference value search means 252.
In the maximum difference value search step ST142, as shown in FIG. 13, the maximum difference value _Dmax is searched among the surrounding eight pixels p to w adjacent to the imaging pixel i at the center (see FIG. 14). . The enhancement difference value calculation step for calculating an emphasis difference value _D'i by adding the maximum difference value D _max found in this maximum difference value search step (ST142) the difference value D _i of the target image-captured pixel i (ST143) is performed by the emphasis difference value calculation means 253.

D ′ _i = D _i + D _max

Then, the same array space as that of the imaging pixels in FIG. 5 is prepared, and the calculated enhancement difference value D ′ _i is arranged in this array space (see FIG. 15). FIG. 16 is a diagram illustrating a numerical example of the enhancement difference value D ′.

In this way, the difference value D of the imaging pixel 131 adjacent to the imaging pixel 131 that is adjacent to the imaging pixel 131 is viewed in the maximum difference value search step (ST142), and the maximum differential value _Dmax among these is set to the differential value D _i of the imaging pixel. Since the emphasis difference value D ′ _i is calculated by addition, a large difference value D exists not in the edge defect 440 but in a position adjacent to the imaging pixel 131 of interest, and is not a mere luminance variation or electric signal noise. When there is an abnormality in luminance, the difference value D between the adjacent imaging pixels 131 is added to calculate a very large enhancement difference value D ′.

When the enhancement difference value D ′ has been obtained for all the imaging pixels 131 in the processing area 231, the luminance enhancement step ST 140 ends, and then the edge defect 440 is based on the luminance value L _i and the enhancement difference value D ′ _i. The defective portion detecting means 300 performs a defective portion detecting step (ST150).
The defect detection step (ST150) will be described with reference to the flowchart of FIG.

In the defective portion detection step (ST150), first, in threshold value setting means 310, a threshold value setting step for setting a threshold value for detecting edge defect 440 is performed in ST160.
In the threshold value setting step (ST160), first, the average luminance value calculating unit 311 performs an average luminance value calculating step (ST161) for calculating the average luminance value M of the processing area 231.
That is, a value (average luminance value M) obtained by averaging the luminance values L (see FIGS. 7 and 8) of the imaging pixels 131 in the processing area 231 is calculated as the average luminance value M of the processing area 231.

Subsequently, a luminance threshold value setting step (ST162) is performed by the luminance threshold value setting means 312, and the average luminance value M is multiplied by a predetermined coefficient α (first predetermined coefficient) preset in the luminance threshold value setting means 312. Thus, the brightness threshold value L _s is calculated.
For example, when the average luminance value of the processing area 231 is “123”, the luminance threshold L _s is set to “86” by multiplying by a predetermined coefficient α (= 0.7).
Then, a luminance difference threshold setting step (ST163) is performed by the luminance difference threshold setting means 313, and a predetermined coefficient β (second predetermined coefficient) preset in the luminance difference threshold setting means 313 is set to the average luminance value M. The luminance difference threshold value D ′ _s is set by multiplication.
For example, when the average luminance value of the processing area 231 is “123”, the luminance difference threshold value D′ _s is set to “61” by multiplying by a predetermined coefficient β (= 0.5).

Next, the luminance value L and the enhancement difference value D ′ of the imaging pixel 131 are compared with the respective threshold values (luminance threshold value L _s , luminance difference threshold value D ′ _s ) set in the threshold value setting step (ST160), and imaging that is a defect candidate is performed. A defective candidate pixel search unit 320 performs a defective candidate pixel search step (ST164) for searching for the pixel 131.
In the defect candidate pixel search step (ST164), the luminance value L _i (see FIGS. 7 and 8) of the imaging pixel 131 is compared with the luminance threshold L _s and the enhancement difference value D ′ _i (FIG. 15, FIG. 15) of the imaging pixel 131 is compared. 16) is compared with the luminance difference threshold value D' _s .
If the luminance value L _i is equal to or greater than the luminance threshold value L _s and the enhancement difference value D ′ _i is equal to or greater than the luminance difference threshold value D ′ _s (ST165: YES), the imaging pixel 131 is set as a defect candidate pixel. Mark.

When such a defect candidate pixel is found (ST165: YES), next, in ST166, the enhancement difference value D ′ _i is equal to or greater than the luminance difference threshold D ′ _s for 8 pixels (p to w) adjacent to the defect candidate pixel. The adjacent pixel search step (ST166) for searching whether there is any is performed by the adjacent pixel search means 330 (see FIG. 18).
If there is at least one enhancement difference value D ′ exceeding the luminance difference threshold D ′ _s in this adjacent pixel search step (ST166) (ST167: YES), a defective candidate pixel is detected as a defective portion.

The flag setting unit 340 performs a flag setting step (ST168) for setting a flag for the detected defective portion. The enhancement difference value D ′ is held for the imaging pixel 131 in the defective portion where the flag is set, and the enhancement difference value D ′ is reset to 0 for the imaging pixel 131 where the flag is not set as the defective portion.
Here, FIG. 19 is an example showing a state in which a flag (“1”) is set for the imaging pixel 131 detected as a defective portion.

  Here, the defective pixel detection process is configured by the adjacent pixel search process ST166 and the flag setting process ST168. Similarly, the neighboring pixel search means 330 and flag setting means 340 constitute defective pixel detection means.

As described above, when the brightness enhancement step (ST140) and the defect detection step (ST150) are completed for one processing area 231 and the edge defect 440 is detected for one processing area 231 (ST170: YES), The luminance enhancement step (ST140) and the defect portion detection step (ST150) are performed on all the processing areas 231 in the imaging data (ST180).
That is, in the processing area selection step (ST130), processing areas 231 to be processed are selected in order, and edge defect 440 detection processing (ST140, ST180) is sequentially performed on the selected processing area 231.

When the detection of the edge defect 440 is completed for all the processing areas 231 and the edge defect detection process is completed for the entire display image (ST180: YES), the flag (“1” is set for the imaging pixel 131 detected as the defective portion. ”) Is set, the edge defect 440 is visualized when the imaging pixel 131 with the flag set is displayed on the display unit 140 (see FIG. 20).
For example, FIG. 20 is an example in which the detected edge defect 440 is displayed.

When the edge defect detection process is completed for the entire display image (ST190: YES), the rank determination unit performs rank determination of the organic EL panel 400 based on the detected edge defect 440 (ST200).
As the rank determination, for example, the number of detected edge defects 440 is counted, and when the edge defects 440 are generated in a predetermined number or more, it is determined as a defective product and the edge defects 440 are less than the predetermined number. For example, it may be determined that the product is non-defective.

According to such 1st Embodiment, there can exist the following effects.
(1) In the luminance enhancement step (ST140), an enhancement difference value D ′ based on a change in the luminance value L when the imaging pixel 131 is viewed from the top to the bottom is calculated, and the enhancement difference value D ′ is used as the luminance difference threshold D. By comparing with ' _s , the edge defect 440 can be automatically detected by arithmetic processing. Therefore, since it is not necessary to rely on the visual inspection of the inspector, the determination of the detection of the edge defect 440 can be unified, and the edge defect 440 can be detected quickly for a large number of organic EL panels 400.

(2) searching the maximum difference value D _max among the imaging pixels 131 adjacent to the target image-capturing pixels 131 in the maximum differential value search step ST142, emphasizing the difference value by adding the difference value D _i of the target pixel _D'i (Enhancement difference value calculation step), there is a large difference value D at a position adjacent to the imaging pixel 131 of interest, not a mere luminance variation or electrical signal noise that is not the edge defect 440, and the luminance is gathered together. A large enhancement difference value D ′ is obtained only when there is an abnormality. Therefore, it is possible to appropriately detect the edge defect 440 in which a collective abnormality in luminance exists, not a noise or the like that exists solely as a luminance variation.

(3) located only directly above the focus imaging pixels 131 when a predetermined condition is satisfied between the luminance values L _i-1 of the imaging pixel 131 is directly above the brightness value L _i between the target image-capturing pixels 131 of interest imaging pixel 131 Since the difference value calculation step ST141 for calculating the difference value D of the luminance value L from the imaging pixel 131 to be performed is performed, not the region where the luminance value L is normally shifted at the boundary between the black matrix 430 and the display pixel 420, A large difference value D _i can be obtained only in a region where the luminance value L changes abnormally at the boundary between the edge defect 440 and the black matrix 430.
Therefore, it is possible to appropriately detect only the edge defect 440, not the point where the black matrix 430 is normally shifted to the display pixel 420.

(4) After marking the defect candidate pixel in the defect candidate pixel search step (ST164), the imaging pixel 131 adjacent to the defect candidate pixel in the adjacent pixel search step (ST166) is emphasized more than the luminance difference threshold value D′ _s. Only when the imaging pixel 131 having the difference value D ′ is present adjacently, it is detected as a defective portion and a flag is set (flag setting step ST168). Therefore, enhancement of the luminance difference threshold value D′ _s or more is also performed at the adjacent position. Only when there is a difference value D ′ and there is an abnormality in luminance can be appropriately detected as the edge defect 440.

(5) In the luminance threshold setting step ST162 and the luminance difference threshold step ST163, the average luminance value M for each processing area 231 is multiplied by a predetermined coefficient (first predetermined coefficient α, second predetermined coefficient β) to set each threshold value. Therefore, an appropriate threshold value can be set for each processing area 231. For example, even when there is a luminance unevenness such as a difference in luminance between the left and right of the display image, the edge defect 440 is set using an appropriate threshold value for each processing area 231. Can be detected.

It should be noted that the present invention is not limited to the above-described embodiment, and modifications, improvements, and the like within the scope that can achieve the object of the present invention are included in the present invention.
Although the inspection target is an organic EL panel, it is not limited to the organic EL panel, and edge defects generated in display pixels of various optical panels such as a liquid crystal panel and a plasma display panel can be detected.

In the processing area extraction step ST120, the oval display pixel including the edge portion is extracted as the processing area. However, since the edge defect is the edge portion of the display pixel, the edge defect detection processing (luminance enhancement step ST140, defect As a processing area for performing the part detection step ST150), the central part of the display pixel may be excluded.
For example, assuming that the length of the processing area in the vertical direction is 1 unit, a region of a quarter from the center to the upper side and a quarter from the center to the lower side may be excluded from the processing area. Then, since the area for performing the edge defect detection process is narrowed, the processing speed can be improved.

In the difference value calculation step, the direction in which the imaging pixels are scanned is not particularly limited. For example, in FIG. 10, the processing area may be scanned from the bottom to the top instead of from the top to the bottom. You may scan.
In addition, as the predetermined condition for performing the difference value calculation process, the case where the relationship between the luminance values of the imaging pixel of interest and the imaging pixel immediately above is viewed has been described, but not limited to this, at the boundary between the black matrix and the display pixel If the difference value can be calculated so that it can react only to the edge defect, the predetermined condition is not limited to the above embodiment. For example, it may be determined whether difference values are calculated by referring to the luminance values of a plurality of pixels above and below the imaging pixel of interest.

In the defect candidate pixel search step, an image pickup pixel whose luminance value L _i is equal to or greater than a predetermined luminance threshold value L _s and whose emphasis difference value D ′ _i is equal to or greater than a predetermined luminance difference threshold value D ′ _s is searched as a defect candidate pixel. However, only the enhancement difference value D ′ _i may be viewed, and an imaging pixel whose enhancement difference value D ′ _i is greater than or equal to the luminance difference threshold D ′ _s may be marked as a defect candidate pixel.

  In the maximum difference value search process, an example in which eight pixels adjacent to the target imaging pixel are searched is shown. However, not all eight adjacent pixels but four pixels above, below, left, and right of the target imaging pixel may be searched.

  The present invention can be used for an image display inspection of an optical panel.

The figure which shows the whole structure of an edge defect detection apparatus. The figure which shows the structure of an arithmetic processing part. The flowchart which shows the procedure of the edge defect detection method. The figure which shows the structure of the display surface of an organic electroluminescent panel. The figure which shows the example of the imaging data imaged with the CCD camera. The figure which extracted only the display pixel of R from the imaging data, and rearranged it. The figure which shows the arrangement | sequence space which arranges the luminance value for every imaging pixel. The figure which shows the numerical example of the luminance value for every imaging pixel. The figure which shows the process sequence of a brightness | luminance emphasis process. The figure which shows a mode that an imaging pixel is scanned to an up-down direction. The figure which shows the arrangement | sequence space which arranges the difference value calculated for every imaging pixel. The figure which shows the numerical example of a difference value. The figure which shows 8 pixels adjacent with respect to the imaging pixel i of interest. The figure which shows the mode of the largest difference value search process of searching for the largest difference value in the imaging pixels adjacent to the imaging pixel of interest. The figure which shows the arrangement | sequence space which arranges the emphasis difference value calculated for every imaging pixel. The figure which shows the numerical example of an emphasis difference value. The flowchart which shows the process sequence of a defect part detection process. The figure which shows a mode that the imaging pixel adjacent to a focused imaging pixel is searched in an adjacent pixel search process. The figure which shows the state which set the flag "1" to the imaging pixel detected as a defective part. The figure which shows the example which displayed the detected edge defect.

Explanation of symbols

DESCRIPTION OF SYMBOLS 100 ... Edge defect detection apparatus, 110 ... Mounting table, 120 ... OLED drive part, 130 ... CCD camera (imaging means), 131 ... Imaging pixel, 140 ... Display part, 150 ... Input means, 200 ... Arithmetic processing part, 210 ... Image data storage unit 220. Luminance value calculating means 221... Array space 221. Display pixel 230. Processing area extracting means 231. Processing area 240. Processing area selecting means 250. Difference value calculating means, 252... Maximum difference value searching means, 253... Enhanced difference value calculating means, 300... Defective portion detecting means, 310. ... brightness difference threshold value setting means, 320 ... defect candidate pixel search means, 330 ... adjacent pixel search means, 340 ... flag setting means, 400 ... Machine EL panel, 410 ... display surface 420 ... display pixels, 430 ... black matrix 440 ... edge defects.

Claims (7)

An edge defect detection method for detecting an edge defect occurring in a display pixel of a display panel,
An imaging step of imaging the display image by imaging means having imaging pixels finer than the display pixels of the display image of the display panel to be inspected;
A luminance value calculation step of calculating a luminance value L _i of each of the imaging pixels based on the imaging data captured by the imaging step,
Selected when imaging pixels in the imaging data obtained by imaging the display pixels are sequentially selected along a predetermined direction, and the base end side in the scanning direction is set to the top and the tip side is set to the bottom. the luminance value L _i of the target image-capturing pixels and immediately thereon between the luminance values L _i-1 of the imaging pixel located immediately above the interest imaging pixel of the luminance value L _i Toko of interest imaging pixel satisfies predetermined conditions have A difference value calculating step of calculating an absolute value of a difference from the luminance value L _i-1 of the imaging pixel as a difference value D _i ;
A maximum difference value search step of searching for the maximum difference value D _max by examining an imaging pixel adjacent to the imaging pixel of interest;
And emphasizing the difference value calculation step of calculating an enhancement difference value _D'i by adding the difference value D _max of the maximum that is searched by the maximum difference value search process on the difference value D _i of the noted imaging pixels,
Detecting the image pickup pixel the enhancement difference value _D'i is the luminance difference threshold _D's or the enhancement difference value _D'i is compared with a predetermined luminance difference threshold _D's of each imaging pixel as a defective portion An edge defect detection method comprising: a defect portion detection step. The edge defect detection method according to claim 1,
The predetermined condition for performing the difference value calculating step is:
When the target imaging pixel is located in the upper half area of the display pixel, the luminance value L _i-1 of the imaging pixel immediately above the target imaging pixel is higher than the luminance value L _{i of the} target imaging pixel;
When the target imaging pixel is located in the lower half of the display pixel, the luminance value L _i-1 of the imaging pixel immediately above the target imaging pixel is lower than the luminance value L _{i of the} target imaging pixel. Edge defect detection method characterized by the above. In the edge defect detection method according to claim 1 or 2,
The defect detection step includes
A defect candidate pixel search step of searching for an imaging pixel whose luminance value L _i is equal to or _larger than a predetermined luminance threshold value L _s and whose enhancement difference value D ′ _i is equal to or larger than a predetermined luminance difference threshold value D ′ _s as a defective candidate pixel. When,
If there is an imaging pixel in which the enhancement difference value D ′ is greater than or equal to a predetermined luminance difference threshold D ′ _s for an imaging pixel adjacent to the defect candidate pixel searched in the defect candidate pixel search step, the defect candidate pixel is determined to be defective. An edge defect detection method comprising: a defective pixel detection step of detecting as a pixel. In the edge defect detection method according to claim 3,
The defect detection step includes a threshold setting step for setting the luminance threshold L _s and the luminance difference threshold D ′ _s ,
The threshold setting step includes an average luminance value calculating step for calculating an average luminance value M of each display pixel;
A luminance threshold value setting step of setting, as the luminance threshold value L _s , a value obtained by multiplying the average luminance value M calculated in the average luminance value calculation step by a first predetermined coefficient α;
A luminance difference setting step of setting, as the luminance difference threshold value D′ _s , a value obtained by multiplying the average luminance value M calculated in the average luminance value calculation step by a second predetermined coefficient β. Edge defect detection method. An edge defect detection device for detecting an edge defect occurring in a display pixel of a display panel,
An imaging unit having imaging pixels finer than the display pixels of the display image of the display panel to be inspected, and imaging the display image;
A luminance value calculation means for calculating a luminance value L _i of each of the imaging pixels based on the imaging data imaged by the imaging means,
The imaging pixel in the imaging data obtained by imaging the display pixel is scanned with attention paid to the imaging pixel in order along a predetermined direction, with the base end side in the scanning direction being up and the tip side being down. Occasionally, the luminance value L _i of the noted imaging pixels when the relationship satisfies the predetermined condition between the luminance values L _i-1 of the imaging pixel located immediately above the interest imaging pixel of the luminance value L _i Toko of interest imaging pixels Difference value calculating means for calculating the absolute value of the difference from the luminance value L _i-1 of the imaging pixel immediately above as a difference value D _i ;
Maximum difference value search means for searching an image pickup pixel adjacent to the image pickup pixel of _interest and searching for the maximum difference value _Dmax ;
And emphasizing the difference value calculating means for calculating the enhancement difference value _D'i by adding the difference value D _max of the maximum that is searched by the maximum difference value search means the difference value D _i of the noted imaging pixels,
Detecting the image pickup pixel the enhancement difference value _D'i is the luminance difference threshold _D's or the enhancement difference value _D'i is compared with a predetermined luminance difference threshold _D's of each imaging pixel as a defective portion An edge defect detection device comprising: defect portion detection means. An edge defect detection program for detecting an edge defect occurring in a display pixel of a display panel,
Incorporating a computer into an edge defect detection apparatus having imaging pixels finer than the display pixels of the display image of the display panel to be inspected and having the imaging means for imaging the display image,
A luminance value calculation means for calculating a luminance value L _i of each of the imaging pixels based on the imaging data imaged by the imaging means,
The imaging pixel in the imaging data obtained by imaging the display pixel is scanned with attention paid to the imaging pixel in order along a predetermined direction, with the base end side in the scanning direction being up and the tip side being down. when the luminance value L _i of the target imaging pixels between the luminance values L _i-1 of the imaging pixel located immediately above the interest imaging pixel of the luminance value L _i Toko of interest imaging pixel is focused when a predetermined condition is satisfied Difference value calculating means for calculating the absolute value of the difference between the brightness value L _i-1 of the imaging pixel immediately above and the difference value,
Maximum difference value search means for searching an image pickup pixel adjacent to the image pickup pixel of _interest and searching for the maximum difference value _Dmax ;
And emphasizing the difference value calculating means for calculating the enhancement difference value _D'i by adding the difference value D _max of the maximum that is searched by the maximum difference value search means the difference value D _i of the noted imaging pixels,
Detecting the image pickup pixel the enhancement difference value _D'i is the luminance difference threshold _D's or the enhancement difference value _D'i is compared with a predetermined luminance difference threshold _D's of each imaging pixel as a defective portion A computer-readable edge defect detection program that functions as a defect detection means.   The recording medium which recorded the edge defect detection program of Claim 6 so that computer reading was possible.

Images