JP2001165633A - Method for detecting pattern defect - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a method for detecting pattern defect in which the processing quantity in the detection of defect by the comparison of a pattern to be inspected with a non-defective pattern is reduced to improve the inspection speed. SOLUTION: The inputted image of a pattern to be inspected is compared with the image of a non-defective pattern to detect the incompatible area, and the horizontal scanning directional and vertical scanning directional continuity of the pattern is also detected. The area where one end of the continuous area of the pattern is close to the compatible area, and the other end is not close to the non-defective pattern or the inspecting pattern is extracted, the area where the length of the continuous direction is less than a prescribed value is included in the compatible area to perform the processing, and the kind of defective is specified from the area where the length of the continuous area is the prescribed value or more.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for detecting a pattern defect used for inspecting a circuit pattern formed on an electronic circuit board.

[0002]

2. Description of the Related Art As a method of detecting a defect of a pattern such as a circuit pattern, an image of a pattern to be inspected is compared with an image of a non-defective pattern determined as a non-defective item, and a defective area which does not match the non-defective pattern is determined. A comparative inspection method for detecting is known. In this comparison inspection, simply comparing the two patterns causes a large amount of mismatched areas around the contour of the pattern due to problems such as pattern variations and alignment accuracy of the two patterns. This leads to speed delays.

[0003] Then, the image of the pattern to be inspected is compared with the image of the non-defective pattern, and as shown in FIG.
For 0, an area in which the length of the area of the pixel continuous in the horizontal scanning direction of the image is shorter than a preset value (in the illustrated example, 2)
9 (b), only the mismatched area 30a having a large value is extracted, and only the characteristic portion of the defect is aggregated to reduce the processing area, as shown in FIG. 9B. The type of defect is specified for 30a. This processing reduces the amount of memory and can improve the inspection processing speed.

[0004]

However, when the area shorter than the set value is deleted in the above-described conventional method, as shown in FIG. If there is a mismatched area 31 in which a short area in the horizontal scanning direction is connected in the vertical scanning direction, the mismatched area 31 will be deleted, and as shown in FIG. The region 31 does not exist. In the state shown in FIG. 10A, the patterns are short-circuited in the minute non-coincidence region 31, and the short-circuit defect is overlooked.

Further, as shown in FIG. 11A, when the pattern 33a and the pattern 33b are short-circuited at the mismatched area 34 as shown in the figure, the area shorter than the set value is deleted. When the processing is performed, FIG.
As shown in (1), only the non-coincidence area 34a remains, the one to be determined as a short-circuit defect is overlooked, and the type of defect cannot be determined.

It is an object of the present invention to provide a pattern defect detection method capable of reducing the processing amount for defect detection while maintaining the capability of defect detection.

[0007]

According to a first aspect of the present invention, there is provided an inspection target pattern image obtained by imaging a pattern to be inspected, which is adjacent to both sides of each pixel in a horizontal scanning direction. The continuity of the pattern is detected between pixels, and the continuity of the pattern is detected between adjacent pixels on both sides of each pixel in the vertical scanning direction. In the pattern defect detection method for detecting a defect of a pattern to be inspected by extracting a non-coincidence area, one continuous end of a pattern in the non-coincidence area is in contact with a coincidence area where a pattern to be inspected matches a non-defective pattern, and When the end of the pattern is in contact with the area where the non-defective pattern and the pattern to be inspected do not exist, and the length of the pattern continuous from one end to the other end is shorter than a predetermined value, The non-coincidence area is processed as being included in the coincidence area, and a defect is detected from an area where the length of a continuous pattern of the non-coincidence area is longer than a predetermined value and an area where continuity is detected in the horizontal scanning direction and the vertical scanning direction. The small area in the non-matching area is processed as included in the matching area, so that the amount of data stored in the memory is reduced. The throughput can be reduced and the inspection speed can be improved.

In the second invention of the present application, for an image obtained by capturing each pattern of a repetitive pattern in which the same pattern is repeated, the continuity of the image is determined for each pixel in the horizontal scanning direction between adjacent pixels on both sides thereof. At the same time, each pixel in the vertical scanning direction is detected for each pixel in the vertical scanning direction to detect the continuity of an image between adjacent pixels on both sides thereof, and a pattern defect is detected by extracting a non-coincidence area where both adjacent patterns do not match. In the pattern defect detection method, for the non-matching area, one end of a continuous image contacts a matching area where both patterns match, the other end touches an area where neither pattern exists, and one end to the other end. When the length of an image that is continuous in the area is shorter than a predetermined value, this mismatched area is processed as being included in the matched area, and the The presence / absence and type of a defect are specified from an area longer than the value and an area where continuity is detected in the horizontal scanning direction and the vertical scanning direction. By processing as included, the amount of data stored in the memory is reduced, so that the processing amount for defect detection is reduced and the inspection speed can be improved.

[0009]

Embodiments of the present invention will be described below with reference to the accompanying drawings to provide an understanding of the present invention.
The embodiment described below is an example embodying the present invention, and does not limit the technical scope of the present invention.

FIG. 1 shows a processing procedure of a pattern defect detecting method according to an embodiment of the present invention. Hereinafter, each processing procedure will be described with reference to FIGS.

First, an inspection target pattern image obtained by imaging a pattern to be inspected is input (S1), and is compared with a non-defective pattern image determined as a non-defective product of the inspection target pattern (S2). By this comparison, an unmatched area of the pattern to be inspected that does not match the non-defective pattern is detected (S3). For example, when the non-defective pattern is the pattern shown in FIG. 2A and the inspection target pattern is the pattern shown in FIG. 2B, as shown in FIG. A mismatch area 1 is detected.

In the detection of the mismatched area 1, the operation of detecting the presence or absence of a pattern for each pixel and the pixels adjacent on both sides in the horizontal scanning direction of the inspection target pattern image is repeated by scanning the entire image. The operation of detecting the presence / absence of a pattern is repeated for each pixel in the vertical scanning direction and pixels adjacent on both sides of the pixel.
In addition, a pixel having a pattern at a coordinate position where there is no pattern in the non-defective pattern image is detected as the non-coincidence area 1. Further, by this scanning, the continuity of the pattern in the mismatch area 1 in the horizontal scanning direction and the vertical scanning direction is detected (S
4).

Next, as shown in FIG. 4, one end of a pattern which is continuous in the horizontal scanning direction or the vertical scanning direction in the non-matching area 1 is in contact with the matching area 2 (pixel 1a shown in FIG. 4). As shown in FIG. 5, the other end is in contact with a region where neither a good pattern nor a pattern to be inspected exists (pixel 1b shown in FIG. 5), and a region 3a shown in FIG.
It is extracted as 3b (S5). In the examples shown in FIG. 4, FIG. 5, and FIG. 6, a pixel row continuous in the horizontal scanning direction is extracted.

In the extracted areas 3a and 3b shown in FIG.
An area whose length in the horizontal scanning direction is equal to or less than a predetermined value (here, two pixels), that is, areas 3a and 3a in the illustrated example are treated as being included in the matching area 1 as a minute defect as shown in FIG. Is performed (S6). The areas 3b and 3c whose lengths in the horizontal scanning direction are equal to or more than a predetermined value are stored in the memory as data for determining the presence or absence and type of a defect as a relatively large defect portion.

The continuity of the image in the mismatch area 1 in the horizontal scanning direction and the vertical scanning direction is also stored in the memory.

As described above, when the mismatched area 1 is detected by comparing the pattern image to be inspected with the non-defective pattern image, all the data in the mismatched area 1 is not stored in the memory, so that it is stored in the memory. The amount of data to be stored is reduced, and a large memory capacity is not required. Therefore, the amount of data processing for determining the presence / absence and type of a defect from the data stored in the memory is small, and a reduction in inspection speed can be suppressed. .

From the continuity of the image in the non-coincidence area 1 where the length of the non-coincidence area 1 in the horizontal scanning direction is greater than or equal to a predetermined value and the continuity of the image in the non-coincidence area 1 in the horizontal and vertical scanning directions,
As shown in FIG. 8, a projection 5a (corresponding to the area 3b) and a short-circuit 5b (corresponding to the areas 3a and 3a defined as the coincident area 2) which are not included in the non-defective pattern are not the inspection target patterns. Continuity) can be detected (S7).

The above description is an example of an inspection target pattern having a defect in the increasing direction with respect to the non-defective pattern, but the inspection target pattern having a defect in the decreasing direction with respect to the non-defective pattern, that is, chipping or disconnection of the pattern. In this case, the defective portion is detected as the non-coincidence area 1. Therefore, by performing the processing shown in FIG. 1 on the non-coincidence area 1, a defect can be detected.

In the case of repeating the same pattern in which the same circuit pattern is continuously arranged on one substrate, adjacent patterns are compared, and the same processing is performed from the detection of the mismatch area 1. , A defective pattern can be detected.

[0020]

As described above, according to the pattern defect detection method of the present invention, the amount of mismatched regions can be reduced, so that the processing amount can be reduced while maintaining the inspection accuracy, and defect detection can be performed efficiently. It can be carried out.

[Brief description of the drawings]

FIG. 1 is a flowchart showing a processing procedure of a pattern defect detection method according to an embodiment.

2A is a non-defective pattern, and FIG. 2B is an inspection target pattern diagram.

FIG. 3 is a diagram of a mismatch area detected by comparing a non-defective pattern with a pattern to be inspected;

FIG. 4 is a view for explaining extraction of pixels in contact with a matching area of a mismatching area.

FIG. 5 is a view for explaining extraction of pixels in contact with an area where neither a good pattern nor an inspection target pattern exists in a mismatched area;

FIG. 6 is a diagram illustrating an extracted region.

FIG. 7 is a view for explaining a state in which a region equal to or less than a predetermined value from an extracted region is included in a matching region;

FIG. 8 is a view for explaining a type of a defect specified from an area having a predetermined value or more from an extracted area and continuity of an image.

9A and 9B show a conventional processing method, in which FIG. 9A shows a detection state of a mismatched area, and FIG. 9B shows a state after processing.

FIG. 10 is a diagram illustrating a problem of a conventional processing method.

FIG. 11 is a diagram illustrating a problem of a conventional processing method.

[Explanation of symbols]

 1 Unmatched area 2 Matched area

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Hajime Kawano 1006 Kazuma Kadoma, Osaka Prefecture Matsushita Electric Industrial Co., Ltd. Term (reference) 2F065 AA49 AA56 BB02 BB18 CC25 DD06 DD07 FF04 FF41 FF61 JJ03 MM22 QQ00 QQ21 QQ23 QQ25 RR05 RR09 2G051 AA65 AB02 AC21 EA11 EB01 EC01 4M106 CA40 DB04 DB21 DJ17 DJ18 DC21 DA03 CC04

Claims (2)

[Claims] 1. An image of a pattern to be inspected, which is obtained by imaging a pattern to be inspected, detects the continuity of the pattern between adjacent pixels on both sides of each pixel in the horizontal scanning direction and detects the pattern continuity in the vertical scanning direction. In a pattern defect detection method for detecting continuity of a pattern between adjacent pixels on both sides of each pixel and detecting a defect of the inspection target pattern by extracting a non-matching area that does not match a non-defective pattern of the inspection target pattern. Regarding the pattern in the non-coincidence area, one end of the continuous pattern is in contact with the matching area where the pattern to be inspected matches the non-defective pattern, and the other end is in contact with the area where the non-defective pattern and the pattern to be inspected are not present. When the length of the pattern continuing from one end to the other end is shorter than a predetermined value, the mismatched area is included in the matched area. Wherein the type of defect is identified from a region where the continuous length of the pattern in the mismatched region is longer than a predetermined value and a region where continuity is detected in the horizontal scanning direction and the vertical scanning direction. Pattern defect detection method. 2. In an image obtained by capturing each pattern of a repetitive pattern in which the same pattern is repeated, the continuity of the pattern between each pixel in the horizontal scanning direction and adjacent pixels on both sides thereof is detected, and the vertical scanning direction is detected. A pattern defect detection method for detecting pattern continuity between adjacent pixels on both sides thereof for each pixel, and detecting a pattern defect by extracting a non-coincidence area where both adjacent patterns do not coincide with each other; For a pattern in a non-matching area, one end of a continuous pattern touches a matching area where both patterns match, the other end touches an area where neither pattern exists, and a pattern continuing from one end to the other end When the length of the pattern is shorter than a predetermined value, the mismatched area is processed as included in the matched area, and the pattern in the mismatched area is processed. And long area length than a predetermined value continues, a pattern defect detecting method characterized by continuity in the horizontal scanning direction and the vertical scanning direction is to identify the type of defect from the detected region.