JPH07190956A - Method and apparatus for detecting defect in structure and/or pattern of material - Google Patents

Abstract

PURPOSE:To provide a highly reliable defect detecting apparatus in which various linear defective materials can be separated perfectly from a flawless linear material. CONSTITUTION:The image of an object to be inspected, e.g. a wired 11 plate glass 10, is picked up while carrying the object in a predetermined direction (arrow Y direction) by an image pickup means which then delivers a one- dimensional binary pixel data along the scanning line every time when the object is scanned. A processing data obtained from a processing line L1 is compared with the pixel data of a processing line L0, i.e. one scanning line for which the image is picked up immediately before, which has finished labeling for every pixel having substantially identical address in the direction of scanning line (arrow X direction). Each pixel of processing data is then labeled with a line 11 or various defect 12, 13, 14, 16 depending on the rule adaptable to the comparison results.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method and apparatus for detecting a defect relating to the structure and / or pattern of an object to be inspected from a plurality of one-dimensional pixel data which are sequentially obtained corresponding to various objects to be inspected. , Especially suitable for defect detection on lined or meshed flat glass.

[0002]

2. Description of the Related Art A conventional defect detecting apparatus for inspecting a locally existing defect in a light-transmissive plate-like body such as an ordinary plate glass is designed to convey an ordinary plate glass, which is an object to be inspected, in a fixed direction. Light is emitted from one surface of the plate glass over its entire width, and the light transmitted through this plate glass is received by the light receiving portion of the one-dimensional line sensor arranged at a position appropriately separated from the other surface. There is. Then, by photoelectrically converting the light received by the light receiving unit, the read pixel signal is taken out every one reading scan.

In this case, since the light radiated to the plate glass is almost entirely transmitted through the plate glass in a normal region having no defect in the plate glass and is emitted from the plate glass, a large amount of light is transmitted in such a normal region. Is obtained. However, since light is absorbed or reflected by the light-shielding defect in a region having a light-shielding defect such as a dot, a line, or a ring formed of bubbles inside the plate glass or an undissolved component of the plate glass, The amount of transmitted light is small in the defective area. Therefore, the value of the read pixel signal is reduced in the defective area of the plate glass as compared with the normal area. Therefore, when the value of the read pixel signal is more than a certain reference value, it is normal, and when it is less than the reference value, it is possible to easily detect the light shielding defect of the plate glass by determining that there is the light shielding defect. it can.

[0004]

However, not the ordinary plate glass as described above, but a lined plate glass in which a plurality of linear wires that are approximately parallel to each other are embedded, or a grid-like or mesh-like wire is embedded. When detecting the same defect as the above-mentioned conventional defect detection device using the mesh plate glass that has been subjected to inspection as the object to be inspected, the scanning line of the one-dimensional line sensor intersects the linear wire and the mesh wire. The value of the read pixel signal is lowered by the linear wire and the mesh wire at the point where the light is read. Since this decrease is very similar to the decrease in the value of the read pixel signal due to the light-shielding defect, the decrease in the value of the read pixel signal caused by the light-shielding defect is caused by the linear wire and the mesh wire. It is very difficult to accurately distinguish it from the decrease in the value of the pixel signal.

In the above-mentioned case, it is conceivable to label by a conventionally known labeling method. In this case, however, the following problems occur, and thus such conventional labeling is performed. The law cannot be used.

That is, when a lined plate glass in which, for example, seven wires which are approximately parallel to each other are embedded is labeled by a conventional labeling method, a wire of one end is labeled with a label of 1 L at the first scanning. Further, the 5 wires in the middle part are sequentially labeled with 2L to 6L, and the wire at the other end is labeled with 7L.
Further, each time a new scan is performed thereafter, the 7 wires are labeled with 1L to 7L.
Then, when such scanning is also performed on the light-shielding isolated defect, a new label (for example, a label of 8L) which has not been used so far is attached to the light-shielding isolated defect. However, the same label as this wire is attached to the light-shielding contact defect overlapping with the wire and the light-shielding crossing defect intersecting with the wire. No detection can be done.

Further, it is very difficult to detect not only the light-shielding defect but also the defect having no light-shielding function such as the broken wire portion by the conventional labeling method.

The present invention has been invented in order to solve the above-mentioned problems, and has a normal structure in which the object to be inspected originally has a light-shielding defect and a line break defect included in the object to be inspected. An object of the present invention is to provide a defect detection method and a defect detection device capable of surely and / or rapidly distinguishing from a pattern and / or a pattern.

[0009]

In order to solve the above-mentioned problems, the present invention provides the defect detection method and the defect detection apparatus described at the beginning, wherein the labeling of each pixel of the one-dimensional pixel data is performed before that. This is performed according to a first label attached to each pixel of the one-dimensional pixel data and a second label attached to a pixel different from the pixel to be labeled in the one-dimensional pixel data. The feature amount of at least the defective portion of the pixel is compared with the reference value.

[0010]

According to the normal configuration and / or pattern of the object to be inspected and the defect contained in the object to be inspected, each pixel of the one-dimensional pixel data can be labeled accurately and quickly.

[0011]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment in which the present invention is applied to a defect detection device for detecting defects in a lined sheet glass will be described with reference to the drawings.

In FIG. 1, an object to be inspected is a lined plate glass 10 in which a large number of wires (hereinafter referred to as "lines") 11 which are substantially parallel to each other and do not intersect each other are embedded. The lined glass sheet 10 has various defects as shown in FIG. 1, and these various defects include a light-blocking contact defect 12 in contact with the line 11 and a light-blocking property intersecting with the line 11. It includes a crossing defect 13, a light-shielding isolated defect 14 isolated from the line 11, and a line cut portion 16 in which a part of the line 11 is broken and both ends thereof are separated. In the present embodiment, the "light-shielding defect" is simply referred to as "defect", and the "light-shielding contact defect" and "light-shielding crossing defect" are hereinafter referred to.
And "light-shielding isolated defect" as "contact defect",
These are called "crossing faults" and "isolation faults".

The lined plate glass 10 is moved to a predetermined scanning line L
A one-dimensional line sensor LS is arranged so as to face one surface of the lined plate glass 10 in order to capture an image along, and an illumination device (not shown) is arranged so as to face the other surface. It is arranged. The light from the illumination device is configured to pass through the lined plate glass 10 and then enter the light receiving portion of the one-dimensional line sensor LS. In FIG. 1, an arrow X indicates the scanning direction in the scanning line L, and an arrow Y indicates the conveyance direction of the lined plate glass 10, which is substantially orthogonal to the arrow X.

The one-dimensional line sensor LS has one pixel in the arrow Y direction and, for example, 2048 pixels in the arrow X direction,
It can be configured to output a one-dimensional pixel signal composed of 2048 pixels for each scanning. The one-dimensional line sensor LS is connected to a binarization circuit 17 for converting the one-dimensional pixel signal output from the one-dimensional line sensor LS into binary data. Is connected to a run length encoding circuit 18 for converting the binary data into run length data. The run-length encoding circuit 18 also includes a separation processing circuit 19 for separating the defects 12 and 13 and the line 11 from each other in a procedure described later, and a line break 16 in a procedure described later. The line break detection circuit 21 for performing the operation is sequentially connected. Then, the separation processing circuit 19 and the line break detection circuit 21 have basic processing circuits 19a and 2a.
1a and correction processing circuits 19b and 21b, respectively. Further, the run-length encoding circuit 18 is connected to a line-spacing determination circuit 25 for determining the spacing between lines in order to determine the inspection start timing. The output of the line interval determination circuit 25 is supplied to the circuits 19 and 21 and the like in order to notify the timing when the circuits 19 and 21 start to operate.

The line break detection circuit 21 corrects the overlapping portion of the line 11 and the defects 12 and 13, and
An overlap correction processing circuit 22 is connected to statistically calculate the width of the line 11 (that is, the length of one line 11 in the arrow X direction) to obtain a line width threshold value, which will be described later. There is. Then, the overlap correction processing circuit 22 receives the connector 2 in order to supply the final reference label data.
6 are connected to the basic processing circuits 19a and 21a, respectively. Further, an inspection device 23 for outputting an inspection result is connected to the overlap correction processing circuit 22.

The inspection device 23 includes a line / defect separation circuit 27 for separating line data and defect data, a defect judgment circuit 28 for making a judgment regarding a defect, and a defect for a line break. A line break defect determination circuit 29 for performing the determination, and a comprehensive determination circuit 3 for performing a comprehensive determination based on the determination results of these two determination circuits 28 and 29.
Contains 0 and. Then, the line break defect determination circuit 29
Is connected by the connector 24 to the correction processing circuits 19b and 21 in order to supply data relating to the line width threshold value.
b respectively connected. In FIG. 1, a scanning line L ₁ indicated by a solid line indicates a scanning line for one scanning (that is, a processing line) currently being scanned by the one-dimensional line sensor LS, and a scanning line L indicated by a dashed line. ₀ indicates a scan line for one scan that is scanned immediately before the processing line L ₁ . Hereinafter, the operation of the defect detection apparatus of this embodiment configured as described above will be described.

In FIG. 1, the line-shaped plate glass 10 is imaged by the one-dimensional line sensor LS while the line-made plate glass 10 continuously produced is conveyed in the arrow Y direction.
A one-dimensional pixel signal is obtained for each scan, and the one-dimensional pixel signal is converted into binary pixel data by the binarization circuit 17. In the binary pixel data, "1" corresponds to the pixel facing the light shield such as the line 11 and defects 12, 13, and 14, and "0" corresponds to the pixel not facing the light shield. is doing. Further, this binary pixel data is supplied to the separation processing circuit 19 after being converted into run length data for each scanning in the run length encoding circuit 17.

When the defect detecting apparatus of this embodiment is started, first, the one-dimensional scanning is performed by the one-dimensional line sensor LS on the front end portion of the lined glass sheet 10, and binary pixel data for one scanning is obtained. Therefore, the line spacing determination circuit 25
Is a pixel of “1” in the binary pixel data for the first scan (that is, a pixel corresponding to the light shielding object).
Unconditionally label all with "line". As shown in FIG. 1, in the lined plate glass 10, since the multiple lines 11 are arranged at a substantially constant interval from each other, the “lines” attached to the multiple lines 11 are The spacing between the labels should be essentially constant with each other immediately.

However, if the front end portion of the lined sheet glass 10 contains a defect, if the "line" is unconditionally labeled as described above, this defect is also labeled. Therefore, in the scanning line L of the front end portion including this defect, a large number of lines 1
The spacing of the "line" labels attached to one from one another does not immediately become approximately constant with one another as described above. Therefore, in the first several or tens of scans, the line-spacing determination circuit 25 monitors the spacing between the "line" labels and the spacing between the "line" labels is substantially constant. I'm waiting for you. Then, the label data for one scan when the line interval determination circuit 25 detects that the intervals between the "line" labels are substantially constant is determined by the separation processing circuit 19 and the line break detection circuit 21. In the basic processing circuits 19a and 21a, they are respectively used as data corresponding to the first scanning line L (that is, first reference data). When the line spacing determination circuit 25 detects that the spacing between the "line" labels is approximately constant, the line spacing determination circuit 25 detects the separation processing circuit 19, the line breakage detection circuit 21, and the like. The above operation is started, and the operation thereafter is stopped by itself.

In the following processing, run length data is actually used, but in order to simplify the explanation, it will be explained using pixels. In the basic processing circuit 19a of the separation processing circuit 19, the binary pixel data (that is, the data to be processed) corresponding to the processing line L ₁ immediately below the line sensor LS is the scanning line L for one scan immediately before that.
Label data corresponding to ₀ (that is, reference data)
The pixel 20 of “1” in the processing line L ₁ (see FIG. 2, which will be described later) is labeled according to the rule shown in Table 1 below.

Table 1 shows rules when the pixel 20 in the binary pixel data of the processing line L ₁ is "1". Then, in the uppermost column of Table 1, the scanning line types L ₀ and L _{1 are shown.}
The type of label attached to each pixel 20 in the label data of the scanning line L ₀ is shown in the left column except the top column, and the label shown in the left column is shown in the right column except the top column. A processing line L having the same address as each pixel 20 in the scan line L ₀ marked with (that is, the value of the address (0 to 2048) along the arrow X direction is the same).
The label type of each pixel 20 in the label data of ₁ is shown. The types of labels shown in the right columns except the top column are attached according to the label of each pixel 20 (described in the left column except the top column) in the label data of the scanning line L ₀ .

[0022]

[Table 1]

In Table 1, when one pixel 20 of the processing line L ₁ is "1" and the label of the pixel 20 of the scanning line L ₀ having the same address as this pixel 20 is "0". Is one pixel 2 of the processing line L _1.
The zeros are labeled "defects". That is, if a light shield suddenly appears at an address in the scanning direction X where there is no light shield until then, the light shield is determined to be a defect. If one pixel 20 of the processing line L ₁ is “1” and the label of the pixel 20 of the scanning line L ₀ having the same address as this pixel 20 is “line”, the processing line L ₁ The one pixel 20 of is labeled "line." That is, if a light-blocking object is detected following the address in the scanning direction X in which the line 11 was present, the line 11 is determined to continue.

Further, one pixel 2 on the processing line L ₁
If 0 is “1” and the label of the pixel 20 of the scan line L ₀ having the same address as this pixel 20 is “line break” meaning the line break portion 16 of the line 11, the processing line L
_{The one} pixel 20 of 1 is labeled "line". That is, if a light-blocking object appears at an address in the scanning direction X where there was a line break portion 16 until then, it is determined that the light-blocking object is the line 11. This is a disadvantage 1 when the line 11 is cut once and then the same thing appears again.
This is because it should be determined that the line is 11 instead of 2, 13, and 14. Further, if one pixel 20 of the processing line L ₁ is “1” and the label of the pixel 20 of the scanning line L ₀ having the same address as this pixel 20 is “defective”, the processing line L ₁ The one pixel 20 of is labeled "defect". That is, defect 1
If a light-blocking object is detected following the address of 2, 13 or 14 in the scanning direction X, the defect is determined to continue.

Further, one pixel 2 on the processing line L ₁
When 0 is “1” and the label of the pixel 20 of the scanning line L ₀ having the same address as this pixel 20 is “line and defect” as described later, the one pixel 20 of the processing line L ₁ is Is labeled with a "line". That is, if a light-blocking object is detected following the address in the scanning direction X where there was a pixel 20 of “line and defect” until then, it is determined that the light-blocking object is the line 11.

In the basic processing circuit 19a, the conversion from FIG. 2A to FIG. 2B is performed according to the rules shown in Table 1. That is, since each pixel 20 of the binary pixel data of the processing line L ₁ is labeled according to the rule shown in Table 1 and the contact defect 12 is labeled differently from the line 11, the contact defect 12 and the line 11 are labeled. Label data in which and are discriminated (that is, separated) from each other is obtained. Then, the label data is sequentially supplied from the basic processing circuit 19a to the correction processing circuit 19b for each scanning.

When there is a contact defect 12 or a crossing defect 13 in the processing line L ₁ , the label data which has been processed by the basic processing circuit 19a is processed in the processing line L _{1 as} shown in FIG. 4 (A). There may be an interruption of pixels 20 labeled "line" within the width of one line 11 or so. This point will be specifically described with reference to FIG. 4A. In the label data of the processing line L ₁ , the four pixels 20 labeled with “line” and the “line” are included.
One pixel 20 labeled "0" is interposed between the pixel 20 labeled "1" and the pixel 20 labeled "1". Therefore, the four pixels 20 labeled with "line" and the one pixel 20 labeled with "line" are arranged along the scanning direction X among the six pixels 20 having the width of the line width. It is intermittent.

In the state shown in FIG. 4A as described above, the threshold value of the line width is too large as compared with the entire width of the line 11,
This is likely to occur when the width of the continuous portion labeled “line” in the scan line L ₀ becomes too large. That is,
If the line width threshold is not too large compared to the total width of the line 11 and is an appropriate value, the pixel 20 labeled "line"
As in the case of the label data of the scanning line L ₀ shown in FIG. 4A, there are no 6 consecutive lines, and FIGS.
As will be described later with reference to (B), only the number of pixels corresponding to the threshold value of the line width is continuous, and the remaining pixels 20 are labeled as "defects". Therefore, in this case, the one pixel 20 labeled with the "line" in the label data of the processing line L ₁ shown in FIG. 4A is labeled with the "defect" in advance by the basic processing circuit 19a. Therefore, the pixels 20 labeled with "line" are not intermittent as in the processing line L ₁ shown in FIG. 4 (A).

When the pixel 20 labeled as "line" in the label data of the processing line L ₁ in FIG. 4A is intermittent, as described above, the correction processing circuit 19b selects one of the intermittent pixels 20. Consecutive pixels 20
, The continuous portion having the largest number of pixels is determined to be the line 11, and the other pixels 20 are determined to be “defects” and are determined to be the “defects”. The 20 label is changed from "line" to "defect".
In FIG. 4 (A), 4 consecutive lines in the processing line L ₁
It is determined that the “line” label of one pixel 20 corresponds to the line 11, and the “line” label of the other one pixel 20 is “line”.
Are replaced with "defects" and are corrected as shown in FIG.

In the correction processing circuit 19b, further,
Based on the threshold of the line width of the line 11, the correction shown in FIGS. 5A to 5B is performed. Ie line 11
When the threshold of the changing line width of 4 pixels is 4 pixels, for example, as shown on the left side of FIG. If the label is in the processing line L ₁ , then the sum of both is 4
Since it becomes a pixel and is within the above threshold, the label of "defect" of this one pixel is "defect" as shown in FIG. 5 (B).
Is changed to a “line” and the width of the line 11a in the processing line L ₁ is set to 4 pixels. On the other hand, FIG. 5 (A)
As shown on the right side of the figure, if there is a label of "defective" of 2 pixels adjacent to the label of "line" of 3 consecutive pixels, the total of both of them becomes 5 pixels, which exceeds the above threshold value. , The two-pixel "fault" label is never relabeled. Therefore, in this case, the width of the line 11b in the processing line L ₁ is 3 pixels or less. In addition, the above-mentioned label data is used for the correction processing in the plate glass 1 with a line.
Performed to more faithfully describe the entity of zero. Further, the line width of the detected line 11 may change due to the fluctuation of the illumination by the lighting device, and further, the line width may vary among the individual lines 11.
It is preferable to recognize the width of each line such as 11b by the correction processing method as described above.

The label data for one scan which has been subjected to the correction processing in the correction processing circuit 19b is supplied to the basic processing circuit 21a of the line break detection circuit 21. Then, in the basic processing circuit 21a, the line data of the processing line L ₁ (that is, the data to be processed) and the label data of the scanning line L ₀ (that is, the reference data) which is one scan immediately before the processing are sequentially performed. Pixels 20 of “0” in the processing line L ₁ that are compared with each other are labeled according to the rules shown in Table 2 below. The meanings of the contents of each column in this Table 2 are the same as those in the above Table 1, and this Table 2 shows that one pixel 20 in the processing line L ₁
Is a rule when is "0".

[0032]

[Table 2]

In Table 2, when one pixel 20 of the processing line L ₁ is "0" and the label of the pixel 20 of the scanning line L ₀ having the same address as this pixel 20 is "line". Is one pixel 2 of the processing line L _1.
The 0 is labeled "broken". That is,
If the light shield suddenly disappears from the address in the scanning direction X, which was the line 11 until then, it is determined that there is a line break portion 16 there. This is because the line 11 cannot be cut in products such as the lined sheet glass 10, and therefore the line cut portion 16
This is because it is necessary to reliably detect Further, a certain pixel 20 on the processing line L ₁ is “0”, and this pixel 2
Pixel 20 of processing line L ₀ having the same address as ₀
If the label is “broken”, the pixel 20
Is labeled "broken." That is, if the light-blocking object does not appear at the address in the scanning direction X at which the line break 16 was present, it is determined that the line break 16 continues. If one pixel 20 of the processing line L ₁ is “0” and the label of the pixel 20 of the processing line L ₀ having the same address as this pixel 20 is “line and defect”, the above 1 One pixel 20 has a “broken line”
Is labeled. That is, when the light-blocking object suddenly disappears from the address in the scanning direction X, which has had "lines and defects" until then, it is determined that there is a line break portion 16 there.

In the basic processing circuit 21a, the conversion from FIG. 3A to FIG. 3B is performed according to the rules shown in Table 2. At this time, simultaneously, from FIG. 2 (B) to FIG.
4B to FIG. 4C, and FIG. 5B to FIG. 5C. Further, when the detected line width of the line 11 changes due to the above-described change in illumination, the address of the scan line L _{1 that} is the same as the address labeled as “line” in the scan line L ₀ . In some cases, a "0" pixel 20 may appear. And
When this "0" pixel 20 is labeled according to the rule shown in Table 2, it is labeled "broken line" as shown in FIG. 6 (A). However, the width of the pixel labeled with this "broken line" can only be smaller than the total width of one line 11.

The line data thus labeled according to the rules of Table 2 is supplied from the basic processing circuit 21a to the correction processing circuit 21b. Then, in the correction processing circuit 21b, as shown in FIG.
From FIG. 5C to FIG. 5D, from FIG. 5C to FIG.
6A, and the conversion from FIG. 6A to FIG. 6B is performed respectively, and all "broken" labels that do not span the entire width of one line 11 become "0" labels. Can be replaced.
Such processing is performed in a completely "line break" state (that is, one line 11 is completely broken over its entire width) to be detected in this embodiment, and other states. This is performed in order to accurately determine the state of partial division of the line 11 which is

When there is a complete "line break", as is clear from Table 2, the total width of the line break portion 16 is the same as the detected line width of the line 11 immediately before the "line break" occurs. Match. The correction processing circuit 21b is shown in FIG.
As shown in FIG. 7, when the number of pixels (that is, the total width) of the continuous portion labeled as “broken line” is smaller than a predetermined value (for example, 5 pixels) that exceeds the threshold value of the line width, FIG. As shown in B), the entire width of the continuous portion labeled as "broken line" is expanded to the above predetermined value. Therefore, even if the tip 11p (see FIG. 1) of the line 11 continues to be slightly displaced in the X direction following these line breaks 16 after several scans, the "line" is determined according to the rule of Table 1 above. Can be accurately determined. Further, if there is a defect near the line break portion 16, this defect may be mistakenly recognized as the line 11. However, even if such an erroneous recognition occurs, the "broken line" itself is a decisive defect, and therefore such an erroneous recognition does not usually cause a problem. In the case where such erroneous recognition is a problem because a high degree of strictness is required, the expansion up to the above-mentioned predetermined value is performed by right and left every time new label data is sent for one scanning. The label number of "line break" may be increased by one pixel (that is, in the arrow X direction). By doing so, it is possible to further reduce the possibility of erroneously recognizing a defect existing in the vicinity of the “broken line” as a line.

In this embodiment, the threshold value of the width of the line break portion 16 is set to a value one pixel larger than the threshold value of the line width. However, this is not always necessary, and the line break portion 16 is continued. It may be set to a sufficiently large value so that the tip 11p (see FIG. 1) of the appearing line 11 can be smoothly connected. Incidentally, the length of the line break portion 16 in the Y direction usually does not matter, but the length in the X direction does not matter (even if it is necessary to make this a problem. Instead, it is sufficient to consider the width of the line 11), so its value can be chosen quite freely.

The label data for one scan which has undergone the correction processing in the correction processing circuit 21b overlaps the correction processing circuit 22.
Is supplied to. Then, in the overlap correction processing circuit 22, as shown in FIG. 8 or 9, the label data for one scan is corrected. That is, in FIG.
Defect 12 or 1 on either side of line 11
When 3 is in contact, the "line" label is replaced with the "line and defect" label. Further, in FIG. 9, only when the contact defects 12 are in contact with the left and right sides of the line 11, the label of “line” is replaced with the label of “line and defect”, and only one of the left and right sides is contacted. If so, the "line" label is left as is. When the correction processing circuit 22 performs the correction processing, the one-dimensional line sensor L having 2048 pixels is used in this embodiment.
The method shown in either FIG. 8 or FIG. 9 can be adopted according to various conditions such as the resolution of S and the stability of illumination by the illumination device. For example, when the width of the defect 12 or 13 to be detected in the direction of the arrow X is generally sufficiently large with respect to the width of the line 11, the correction method of FIG. 8 is adopted, and in other cases, the correction method of FIG. Nine correction methods can be adopted.

The final label data (that is, the correction data) which has been subjected to the correction processing in the correction processing circuit 22.
Is supplied to the inspection device 23 every one scan, and as label data (that is, reference data) assigned to the scan line L ₀ in Tables 1 and 2 above,
Basic processing circuits 19a and 21a via the connector 26
Is supplied to each. In the line / defect determination circuit 29 of the inspection device 23, the line width threshold value is statistically calculated based on the label data of the final "line" portion for several scans. Data relating to the threshold value is supplied to the correction processing circuits 19b and 21b via the connector 24, respectively. Therefore, in this case, if necessary, the threshold value of the line width is changed from 4 pixels to another number of pixels according to this statistically calculated value. Since the accuracy of separation for separating the defect 12 or 13 and the line 11 from each other changes depending on the threshold value of the line width, the threshold value of the line width is detected by the fluctuation of the illumination by the illumination device as described above. It must be set with high accuracy so that it is always appropriate even if there is a change in the line width.

In the inspection device 23, first, the line / defect separation circuit 27 causes the data of the lined sheet glass 10 to be "line" and "line break" data as shown in FIG. They are separated from each other by the "defect" data as shown in B). For example, the data of FIG. 10A is supplied to the line break defect determination circuit 29, and the data of FIG. 10B is supplied to the defect determination circuit 28. Then, in the line / defect determination circuit 29, based on the data of FIG. 10A, the line width of the line 11 and the variation of the mutual intervals, the meandering of the line 11, the line 11 such as the broken line portion 16 of the line 11 are related. The feature amount is detected. In addition, the defect determination circuit 28 groups the individual defects 12 to 14 by performing labeling processing (that is, labeling processing) by the conventional labeling method as described above on the data of FIG. 10B. Then, for each of the defects 12 to 14, the feature amount such as the length in the X direction, the length in the Y direction, and the area is calculated. Further, the comprehensive determination circuit 31 compares the feature amount and the feature amount of the line 11 with a predetermined reference value to determine the lined flat glass 1
Judge overall pass / fail of products such as 0.

In the above-described embodiment, the lined plate glass 1
On one scanning line L over the entire width of 0, the line 11 and various defects 12, 13, 14 and the like are usually present in the order of several to several tens.
9 and the processing in the line break detection circuit 21,
It need not be done for all 48 pixels. Ie 2
Since most of the 048 pixels are "0" pixels 20, the processing of the "0" pixels 20 can be omitted. By doing so, the processing time can be significantly shortened.

Further, in the above-mentioned embodiment, the overlap correction processing circuit 22 may be omitted, but in this case, the pixel 20 in the scanning line L ₀ is not labeled with “line and defect”. Therefore, instead of the above Table 1, the following Table 3
And the following Table 4 can be used instead of Table 2 above to simplify Table 1 and Table 2. However, of course, even if the correction processing circuit 22 is omitted, the above Tables 1 and 2 may be used as they are.

[0043]

[Table 3]

[0044]

[Table 4]

In the above embodiment, the case where the object to be inspected is the lined plate glass 10 as shown in FIG. 1 has been described. However, the present invention is substantially parallel to each other and has a large number of linear patterns. It can also be applied to defect detection of an object (for example, printed matter) on which is printed. Further, in the above-described embodiment, the line width of the line 11 is about 3 to 4 pixels, but in the actual defect detection device, the line width of the line 11 is ten or more pixels or more. It is preferable to set the conditions such as the number of pixels of the one-dimensional line sensor LS so as to correspond to the pixels because the accuracy becomes higher. Further, in the above-described embodiment, when the overlap correction processing circuit 22 and / or the line break detection circuit 21 is omitted, or in addition to the omission of these circuits 22 and 21, the correction processing circuit 19b of the separation processing circuit 19 is further added. Even if omitted, line 1
It is possible to separate 1 and the defects 12, 13 from one another.

Further, in the present invention, the line 11 is the arrow Y.
Even if it is tilted diagonally instead of being almost parallel to the direction,
FIG. 5A to FIG. 5B in the correction processing circuit 19b
It is clear that the same good defect detection can be performed by using the conversion to the same as in the above-described embodiment. In addition, in the present invention, not only when the plurality of lines 11 are parallel to each other but also when the lines 11 intersect each other in a lattice shape such as a net, the grid portion and the drawback of being in contact with the grid portion are Can be separated from each other,
It is also clear that broken lines can be detected.

Further, in the above-described embodiment, since the predetermined processing can be performed every time the binary pixel data for one scan is obtained, the two-dimensional image memory is not particularly required. Further, although the one-dimensional line sensor LS is used in the above-described embodiment, the pixel data for each scanning is extracted from the two-dimensional image obtained by imaging the inspection target object with the two-dimensional imaging camera. You may

In the above embodiment, the labeling of each pixel of the one-dimensional pixel data corresponds to the pixel of the one-dimensional pixel data to be labeled among the pixels of the one-dimensional pixel data immediately before it. The second label is applied according to the first label attached to the pixel having the address, and the labeled label is attached to the pixel other than the pixel to be labeled in the one-dimensional pixel data. I am trying to fix it accordingly. However, this labeled label is used as a third label attached to each pixel of the subsequent one-dimensional pixel data (particularly, the labeling of the one-dimensional pixel data among the pixels of the one-dimensional pixel data immediately after that). Further correction may be made according to the label attached to the pixel having the address corresponding to the target pixel).

[0049]

According to the present invention, each pixel of one-dimensional pixel data is labeled with a first label attached to each pixel of the previous one-dimensional pixel data and the one-dimensional pixel data. This is performed according to the pixel to be labeled and the second label attached to another pixel, and the feature amount of at least the defective portion of the labeled pixel is compared with the reference value. Therefore, each pixel of the one-dimensional pixel data can be labeled accurately and quickly according to the normal configuration and / or pattern of the inspection target object and the defect included in the inspection target object. The defects contained in can be reliably and quickly distinguished from their normal configuration and / or pattern, which allows reliable detection of defects at high speed.

[Brief description of drawings]

FIG. 1 is a block diagram of an embodiment in which the present invention is applied to a defect detection device that detects a defect in a lined sheet glass.

FIG. 2 is a diagram for generally explaining a procedure of separating a “line” from a “defect” and a procedure for detecting a line break in binary pixel data obtained from the run-length encoding circuit shown in FIG. 1. 2A is a diagram showing binary pixel data of a processing line obtained from the run-length encoding circuit shown in FIG. 1 in comparison with label data of a scanning line immediately before, and FIG. FIG. 3 is a diagram showing label data of a processing line obtained from a basic processing circuit of the separation processing circuit shown in FIG. 1 in comparison with label data of a scanning line immediately before the processing line;
FIG. 3C is a diagram showing label data of a processing line obtained from the basic processing circuit of the line break detection circuit shown in FIG. 1 in comparison with label data of a scan line immediately before it.
FIG. 3D is a diagram showing the label data of the processing line obtained from the correction processing circuit of the line break detection circuit shown in FIG. 1 in comparison with the label data of the scanning line immediately before it.

3 is a view for explaining a procedure of line break detection by a basic processing circuit of the line break detection circuit shown in FIG.
FIG. 3A is a diagram showing label data of a processing line obtained from the correction processing circuit of the separation processing circuit shown in FIG. 1 in comparison with label data of a scanning line immediately before it, and FIG.
FIG. 3 is a diagram showing label data of a processing line obtained from the basic processing circuit of the line break detection circuit shown in FIG. 1 in comparison with label data of a scanning line immediately before it.

FIG. 4 is a view for explaining a procedure for discriminating a “line” label and a “defect” label from each other,
2B is a diagram showing the label data of the processing line obtained from the basic processing circuit of the separation processing circuit shown in FIG. 1 in comparison with the label data of the scanning line immediately before, and FIG. It is a figure which compares the label data of the processing line obtained from the correction processing circuit of a circuit with the label data of the processing line immediately before, and (C) is obtained from the basic processing circuit of the line break detection circuit shown in FIG. It is a figure which shows the label data of the processed line contrasted with the label data of the scanning line just before that, and (D) shows the label data of the processed line obtained from the correction processing circuit of the line break detection circuit shown in FIG. It is a figure shown in contrast with the label data of the scanning line just before that.

5A is a view for explaining a procedure of correcting a label of "line", in which (A) shows label data of a processing line obtained from a basic processing circuit of the separation processing circuit shown in FIG. 2B is a diagram showing the comparison with the label data of the scanning line of FIG. 1, and FIG. 7B is a diagram showing the comparison of the label data of the processing line obtained from the correction processing circuit of the separation processing circuit shown in FIG. FIG. 3C is a diagram showing the label data of the processing line obtained from the basic processing circuit of the line breakage detection circuit shown in FIG. 1 in comparison with the label data of the scanning line immediately before,
FIG. 3D is a diagram showing the label data of the processing line obtained from the correction processing circuit of the line break detection circuit shown in FIG. 1 in comparison with the label data of the scanning line immediately before it.

6A and 6B are views for explaining a procedure of correcting a "line break" label by a correction processing circuit of the line break detection circuit shown in FIG. 1, and FIG. 6A is a line break detection circuit shown in FIG. 2B is a diagram showing the label data of the processing line obtained from the basic processing circuit of FIG. 4 in comparison with the label data of the scanning line immediately before, and FIG. 3B is obtained from the correction processing circuit of the line break detection circuit shown in FIG. It is a figure which compares the label data of a process line with the label data of the scanning line just before that.

7 is a view for explaining a procedure for correcting the width of a continuous portion of a "broken line" label by the correction processing circuit of the broken line detection circuit shown in FIG. 1, and FIG. It is a figure which shows the label data of the processing line obtained from the basic processing circuit of the line break detection circuit shown in contrast with the label data of the scanning line immediately before, and (B) is a modification of the line break detection circuit shown in FIG. It is a figure which compares the label data of the processing line obtained from a processing circuit with the label data of the scanning line just before it.

FIG. 8 is a diagram for explaining an example of a procedure for correcting an overlapping portion of a “line” and a “defect” in label data by the overlapping correction processing circuit shown in FIG.

9 is a view similar to FIG. 8 for explaining another example of the procedure for correcting the overlapping portion of the “line” and the “defect” in the label data by the overlapping correction processing circuit shown in FIG. 1.

10A is a plan view of an object to be inspected showing data of “line” and “line break” separated from “defects” by the inspection apparatus shown in FIG. 1, and FIG. FIG. 3 is a plan view of an object to be inspected showing data of “defects” separated from “lines” and “broken lines” by the inspection device shown in FIG. 1.

[Explanation of symbols]

 10 Lined glass plate 11 Line 12 Contact defect 13 Crossing defect 14 Isolated defect 16 Line break 19 Separation processing circuit 20 Pixel 21 Line break detection circuit 22 Overlap correction processing circuit 23 Inspection device

[Procedure amendment]

[Submission date] February 15, 1994

[Procedure Amendment 1]

[Document name to be amended] Statement

[Name of item to be corrected] 0015

[Correction method] Change

[Correction content]

The line break detection circuit 21 is connected to an overlap correction processing circuit 22 which performs a correction process on the overlapping portion between the line 11 and the defects 12 and 13. Then, the overlap correction processing circuit 22 uses the connector 26 to supply the final reference label data to the basic processing circuit 19a.
And 21a, respectively. Further, an inspection device 23 for outputting an inspection result is connected to the overlap correction processing circuit 22.

[Procedure Amendment 2]

[Document name to be amended] Statement

[Correction target item name] 0016

[Correction method] Change

[Correction content]

The inspection device 23 relates to a line / defect separation circuit 27 for separating line data and defect data, a defect judgment circuit 28 for making a judgment regarding a defect, and a line. A line defect determination circuit 29 for determining a defect,
It includes a comprehensive judgment circuit 30 for making a comprehensive judgment based on the judgment results of these two judgment circuits 28 and 29. Then, the line defect determination circuit 29 determines the width of the line 11.
(That is, the length of one line 11 in the arrow X direction)
Correction processing circuits 19b and 21 via connector 24 to provide data relating to the line width threshold being
They are respectively connected to the b. In FIG. 1, a scanning line L ₁ indicated by a solid line indicates a scanning line (that is, a processing line) for one scanning currently being scanned by the one-dimensional line sensor LS, and a scanning line L indicated by a dashed line. ₀ indicates a scan line for one scan that is scanned immediately before the processing line L ₁ . Hereinafter, the operation of the defect detection apparatus of this embodiment configured as described above will be described.

[Procedure 3]

[Document name to be amended] Statement

[Correction target item name] 0039

[Correction method] Change

[Correction content]

The final label data (that is, the correction data) which has been subjected to the correction processing in the correction processing circuit 22.
Is supplied to the inspection device 23 every one scan, and as label data (that is, reference data) assigned to the scan line L ₀ in Tables 1 and 2 above,
Basic processing circuits 19a and 21a via the connector 26
Is supplied to each. Further, the line defect determination circuit 29 of the inspection device 23 statistically calculates the line width threshold value based on the label data of the final "line" portion for several scans, and this line width threshold value is determined. Value data is connector 2
4 to the correction processing circuits 19b and 21b, respectively. Therefore, in this case, if necessary, the threshold value of the line width is changed from 4 pixels to another number of pixels according to this statistically calculated value. Since the accuracy of separation for separating the defect 12 or 13 and the line 11 from each other changes depending on the threshold value of the line width, the threshold value of the line width is detected by the fluctuation of the illumination by the illumination device as described above. It must be set with high accuracy so that it is always appropriate even if there is a change in the line width.

[Procedure amendment 4]

[Document name to be amended] Statement

[Correction target item name] 0040

[Correction method] Change

[Correction content]

In the inspection device 23, first, the line / defect separation circuit 27 causes the data of the lined sheet glass 10 to be "line" and "line break" data as shown in FIG. They are separated from each other by the "defect" data as shown in B). For example, the data of FIG. 10A is supplied to the line defect determination circuit 29, and the data of FIG. 10B is supplied to the defect determination circuit 28. Then, in the line defect determination circuit 29, based on the data in FIG. 10A, the characteristics of the line 11 such as the variation of the line width of the line 11 and the mutual spacing, the meandering of the line 11, the broken portion 16 of the line 11, and the like. The amount is detected. In addition, the defect determination circuit 28 groups the individual defects 12 to 14 by performing labeling processing (that is, labeling processing) by the conventional labeling method as described above on the data of FIG. 10B. Then, for each of the defects 12 to 14, the length in the X direction, the length in the Y direction,
A feature amount such as an area is calculated. Furthermore, the comprehensive judgment circuit 3
1 compares the characteristic amount of the line 11 and the characteristic amount of the line 11 with a predetermined reference value to determine whether the product such as the lined plate glass 10 is totally passed or failed.

[Procedure Amendment 5]

[Document name to be corrected] Drawing

[Name of item to be corrected] Figure 1

[Correction method] Change

[Correction content]

[Figure 1]

Claims (5)

[Claims] 1. A method for detecting a defect relating to a structure and / or a pattern of an object to be inspected from a plurality of one-dimensional pixel data sequentially obtained corresponding to the object to be inspected, comprising: Labeling is performed on a first label attached to each pixel of the previous one-dimensional pixel data and a second label attached to a pixel different from the pixel to be labeled in the one-dimensional pixel data. A defect detection method comprising: performing a corresponding operation and comparing the feature amount of at least a defective portion of the labeled pixel with a reference value. 2. A first label for labeling each pixel of the one-dimensional pixel data, a first label attached to each pixel of the previous one-dimensional pixel data, and a pixel to be labeled in the one-dimensional pixel data. The method according to claim 1, wherein the method is performed according to a second label attached to a pixel different from the above and a third label attached to each pixel of the one-dimensional pixel data immediately thereafter. 3. The first label is a label attached to a pixel having an address corresponding to a pixel to be labeled with the one-dimensional pixel data, of the pixels of the one-dimensional pixel data before the first label. The method according to Item 1 or 2. 4. The third label is a label attached to a pixel having an address corresponding to a pixel to be labeled with the one-dimensional pixel data among the pixels of the one-dimensional pixel data after the third label. Item 4. The method according to Item 2 or 3. 5. An apparatus for detecting a defect relating to a structure and / or a pattern of an object to be inspected from a plurality of one-dimensional pixel data sequentially obtained corresponding to the object to be inspected, wherein each pixel of the one-dimensional pixel data is detected. Labeling is performed on a first label attached to each pixel of the previous one-dimensional pixel data and a second label attached to a pixel different from the pixel to be labeled in the one-dimensional pixel data. A defect detection apparatus comprising: a labeling unit that performs a response; and a comparing unit that compares a feature amount of at least a defective portion of the labeled pixel with a reference value.