JP2000065747A - Apparatus and method for inspection of defect of pattern - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an apparatus and a method for the inspection of the defect of a pattern, in which even the defect having a small change in a brightness level can be detected satisfactorily. SOLUTION: In the pattern-defect inspection apparatus 20, inspection data which is obtained by picking up the image of an object to be inspected is compared with its reference data, and a defect on the object to be inspected is inspected. The pattern-defect inspection apparatus 20 is featured in such a way that an edge-direction detecting means 26 which detects the edge direction of a pattern on the object to be inspected is provided. In addition, data fetch means 24, 27 wherein a pixel data group having a pixel, to be noticed, which is directed to a direction perpendicular to the detected edge direction and whose edge stretches over any pixel is fetched regarding the inspection data and the reference data are provided. In addition, white-ratio conversion means 30, 31 by which the ratio (white ratio) of a dark part to a bright part in respective pixels in the fetched pixel data group is calculated are provided. In addition, a defect judgment means 32 by which whether a defect exists on the object to be inspected or not is judged on the basis of the calculated white ratio and on the basis of a preset threshold value.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] 1. Field of the Invention [0002] The present invention relates to a pattern defect inspection apparatus and a pattern defect inspection method for detecting pattern shape defects in a photomask.

[0002]

2. Description of the Related Art FIG. 11 is a block diagram of a pattern defect inspection apparatus. The imaging device 1 includes a line sensor and the like. The imaging apparatus 1 captures an image of a pattern of a photomask as an object to be inspected and outputs an image signal thereof.
This image signal is digitally converted by the A / D converter 2 and sent to the comparison circuit 3 as inspection data.

On the other hand, a reference data generation circuit 4 develops design data of a photomask pattern into a bit pattern,
In addition, position data indicating an image capturing position on the object to be inspected by the imaging apparatus 1 is input, and reference data corresponding to a photomask pattern to be inspected is generated using the position data.

The position data on the object to be inspected is obtained by measuring the moving position of the imaging device 1 by using a laser interferometer. The comparison circuit 3 compares the inspection data with the reference data and outputs a defective portion on the object to be inspected, and performs the following operation.

[0005] The differentiating circuit 5 has an X direction, a Y direction, and ± 4
It has each differential parameter in the 5-degree direction, and differentiates the inspection data with these differential parameters. The edge direction detection circuit 6 has differentiating parameters in the X direction, the Y direction, and ± 45 degrees, and indicates the minimum value among the differentiating values obtained by differentiating the reference data using these differentiating parameters. The direction of the parameter is detected as the edge direction of the pattern formed on the inspection object.

[0006] In response to the detection result of the edge direction, the selector 7 selects a differentiation result by a differentiation parameter in the same direction as the edge direction among the differentiation results of the differentiation circuit 5, and sends the differentiation result to the subtraction circuit 8.

The peripheral pixel edge direction differentiating circuit 9 differentiates the eight pixels existing around the pixel for which the differential value is determined in the reference data in the same direction as the pattern edge direction detected by the edge direction detecting circuit 6. Differentiate by parameters.

The maximum value detection circuit 10 detects the maximum value of the absolute values of the eight differential values of the peripheral pixels obtained by the peripheral pixel edge direction differentiating circuit 9 and sends the detected absolute value to the subtraction circuit 8.
The subtraction circuit 8 receives the differential result by the differential parameter in the same direction as the edge direction selected by the selector 7 and the maximum value of each differential value of the peripheral pixel. The maximum value of each differential value is subtracted, and the result of the subtraction is sent to the first determination circuit 11.

The first determination circuit 11 compares the subtraction result with the threshold value, and determines a portion where the subtraction result is larger than the threshold value as a defect. On the other hand, the direct comparison circuit 12
The inspection data and the reference data are input, and a value obtained by subtracting the reference data from the inspection data, that is, a level difference between the data is obtained and sent to the second determination circuit 13.

The second determination circuit 13 compares the level difference between each data with a threshold value, and determines a portion where the level difference is larger than the threshold value as a defect. After a while, the first
The respective determination results of the second determination circuits 11 and 13 are output as defect data by passing through the OR gate 14.

The reference data generation circuit 4 in such a pattern defect inspection shows an example of a photomask pattern in which a light-shielding pattern 15 of a Cr film is formed on a photomask.

For example, FIG. 12 shows a light shielding pattern 1 of a Cr film.
5 shows an example of a photomask pattern on which No. 5 is formed. Inspection data on the line AA for such a photomask pattern is “0” in a portion where the light shielding pattern 15 is formed as shown in FIG.
It becomes a dark level in the vicinity, and becomes a light level in the vicinity of, for example, “224” in a portion where the light shielding pattern 15 does not exist.

Further, as shown in FIG. 13B, the reference data is generated with the dark level “0” and the light level being, for example, “224” corresponding to the presence or absence of the light-shielding pattern 15 of the inspection data.

Therefore, the dark level "0" matches the light level "224" between the inspection data and the reference data, and the comparator 3 can perform defect determination by comparing the inspection data with the reference data without erroneous recognition. .

[0015]

However, in such a pattern defect inspection apparatus, even if the size of the defect is the same, the brightness of the defect is not always the same. A defect with a small change in level has a lower detection sensitivity than a defect with a large change in brightness level, so that a defect occurs that the defect is missed.

In this case, as shown in FIG. 14, since the output level of the defect signal is large in a normal defect,
Although the defect can be detected satisfactorily, the defect cannot be detected satisfactorily for a defect having a small change in brightness level because the output level of the defect signal is small.

That is, even if there is a defect having a small change in the brightness level such as a slightly bright portion or a slightly dark portion in the pattern shape of the object to be inspected, it causes a defect in the photomask. Such a defect having a small change in brightness level cannot detect a defect that does not reach a certain value or more, and thus has a low detection sensitivity.

[0018] Even in such a defect having a small change in brightness level, it is desired to find out a defect in a photomask pattern in which miniaturization is further advanced. The present invention has been made based on the above circumstances, and an object of the present invention is to provide a pattern defect inspection apparatus and a pattern defect inspection method capable of detecting a defect with a small change in brightness level. It is.

[0019]

In order to solve the above-mentioned problems, the invention according to claim 1 compares inspection data obtained by imaging an object to be inspected with reference data thereof, and compares the inspection data with the reference data. In a pattern defect inspection apparatus for inspecting a defect, an edge direction detecting means for detecting an edge direction of the pattern of the object to be inspected, and a direction perpendicular to the edge direction detected by the edge direction detecting means, any one of pixels A data extraction unit that extracts a pixel data group having a pixel of interest over which the edge of the pattern of the object to be inspected crosses over the inspection data and the reference data, and a pixel data group and a reference data of the inspection data that are extracted by the data extraction unit. A white ratio conversion means for calculating a ratio (white ratio) between a dark part and a light part in each pixel of the pixel data group; It is white ratio of inspection data, based on the white ratio and a preset threshold value of the reference data,
A defect determining unit that determines the presence or absence of a defect at a predetermined position of the object to be inspected.

According to a second aspect of the present invention, in the pattern defect inspection apparatus according to the first aspect, the edge direction of the inspection object is recognized by detecting an edge direction of a pattern of reference data. is there.

According to a third aspect of the present invention, the pixel data group is composed of three pixel data in which a pixel of interest straddling an edge of an object to be inspected is located at the center while being provided in series. A pattern defect inspection apparatus according to claim 1 or 2, wherein:

According to a fourth aspect of the present invention, the defect judging means compares a difference between the white ratio of the inspection data and the white ratio of the reference data with a predetermined threshold value to determine the difference between the white ratios. 4. The pattern defect inspection apparatus according to claim 1, wherein a defect is determined when a value exceeds a predetermined threshold value.

According to a fifth aspect of the present invention, the threshold value is:
The pattern defect inspection apparatus according to any one of claims 1 to 4, wherein the pattern defect inspection apparatus has a value closer to a light part side than an intermediate value between a light part level and a dark part level. .

According to a sixth aspect of the present invention, the threshold value is:
5. The pattern defect inspection apparatus according to claim 1, wherein the pattern defect inspection apparatus has a value closer to a dark part side than a value intermediate between a light part level and a dark part level.

According to a seventh aspect of the present invention, the threshold value is:
It has a value closer to the light side than the intermediate value between the light level and the dark level, and also has a value closer to the dark side than this intermediate value. A pattern defect inspection apparatus according to any one of claims 1 to 4.

In a preferred embodiment of the present invention, the defect judging means determines a defect from the difference between the white ratio obtained by the subtracting means and a threshold value, the subtracting means obtaining a difference in white ratio between the inspection data and the reference data. The pattern defect inspection apparatus according to any one of claims 1 to 7, further comprising a determination unit.

According to a ninth aspect of the present invention, in the pattern defect inspection method for inspecting a defect on the inspection object by comparing inspection data obtained by imaging the inspection object with reference data, An edge direction detection step of detecting the edge direction of the pattern, and a data cutout step of cutting out a pixel data group having a pixel of interest in which the edge of the pattern of the pattern to be inspected straddles any one of the test data and the reference data, A white ratio conversion step of calculating a ratio (white ratio) between a dark part and a light part in each pixel of the pixel data group of the inspection data and the pixel data group of the reference data cut out in the data cutting step; Based on the white ratio of the inspection data calculated in the process, the white ratio of the reference data, and a preset threshold value, the predetermined position of the inspection object is determined. A pattern defect inspection method characterized by comprising the defect determination step of determining presence or absence of a defect that, the.

According to the first aspect of the present invention, the edge direction detecting means for detecting the edge direction of the pattern of the object to be inspected, and the direction perpendicular to the edge direction detected by the edge direction detecting means, A data cutout unit for cutting out a pixel data group having a target pixel straddling an edge of a pattern of an object to be inspected with respect to test data and reference data, respectively, and a pixel data group of test data and pixel data of reference data cut out by the data cutout unit A white ratio conversion means for calculating a ratio (white ratio) between a dark part and a light part in each pixel of the group; a white ratio of the inspection data, a white ratio of the reference data, and a preset white ratio calculated by the white ratio conversion means. A defect determining means for determining the presence or absence of a defect at a predetermined position of the inspection object based on the threshold value. Inspection data and reference data are cut out in a direction perpendicular to the direction, and the degree of brightness (white ratio) at the pixel of interest is compared with the cut-out image data group by white ratio conversion means. It can be calculated.

Here, when the white ratio is calculated, a minute difference in the brightness level can be reflected. Then, based on the calculated white ratio of the inspection data, the white ratio of the reference data, and a preset threshold value, it is determined whether or not the brightness level corresponds to a defect. From the setting of the threshold value and the white ratio reflecting the minute brightness, it is possible to satisfactorily find even a defect with a small change in the brightness level.

According to the second aspect of the present invention, since the edge direction of the inspection object is recognized by detecting the edge direction of the pattern of the reference data, the more accurate detection of the edge direction can be performed by using the reference data. It is possible to do it.

According to the third aspect of the present invention, the pixel data group is provided in series and is composed of pixel data of three pixels in which the target pixel straddling the edge of the object to be inspected is located at the center. By giving the brightness level of each pixel and further giving a threshold value, it is possible to calculate the white ratio of the target pixel.

According to the fourth aspect of the present invention, the defect determining means compares the difference between the white ratio of the inspection data and the white ratio of the reference data with a predetermined threshold value, and determines that the difference between the white ratios is a predetermined value. Since a defect is determined when the threshold value is crossed, it is easy to detect a minute difference in brightness level by comparing the difference in white ratio. That is, even when the change in the brightness level is small, it is possible to satisfactorily find such a defect having a small change in the brightness level by taking the difference between the white ratio of the inspection data and the white ratio of the reference data. Has become.

According to the fifth aspect of the present invention, since the threshold value has a value closer to the light portion side than the intermediate value between the level of the bright portion and the level of the dark portion, the threshold value is slightly brighter than the bright portion. It is possible to detect even a defect with a low quality. This makes it possible to improve the detection sensitivity on the side with the higher brightness level.

According to the sixth aspect of the present invention, since the threshold value has a value closer to the dark portion side than the intermediate value between the level of the bright portion and the level of the dark portion, a defect slightly brighter than the dark portion can be obtained. Even if there is, it is possible to discover well. This makes it possible to improve the detection sensitivity on the side with the lower brightness level.

According to the seventh aspect of the present invention, the threshold value has a value closer to the light portion side than the intermediate value between the bright portion level and the dark portion level, and has a higher value than the intermediate value. Since it also has a value that is closer to the dark part side, it is possible to satisfactorily find a bright defect or a dark defect as long as the defect has a small change in brightness level. As described above, since the threshold values are provided closer to the light part side and the dark part side, respectively,
Even for a defect with a small change in brightness level, the detection range is widened, thereby making it possible to provide a pattern defect inspection device with improved detection sensitivity.

According to the eighth aspect of the present invention, the defect judging means judges the defect based on the difference between the white ratio obtained by the subtracting means and the threshold value, and the subtraction means for obtaining the difference between the white ratio of the inspection data and the reference data. By determining the difference between the white ratios by the subtracting means, the presence or absence of a defect having a small change in brightness level can be determined. The determination as to whether or not a defect is present can be made by the determination means.

According to the ninth aspect of the present invention, the edge direction detecting step of detecting the edge direction of the pattern of the object to be inspected, and the pixel data group having the target pixel in which the edge of the pattern of the object to be inspected straddles any pixel For extracting inspection data and reference data, respectively, and the ratio (white ratio) of the dark part and the light part in each pixel of the pixel data group of the inspection data and the pixel data group of the reference data extracted in the data extraction step Is calculated based on the white ratio of the inspection data, the white ratio of the reference data, and the threshold value set in advance based on the white ratio of the inspection data calculated in the white ratio conversion step. And a defect determining step of determining the presence / absence of an image data group including the pixel of interest. Calculate the white ratio from the threshold value and calculate the change in brightness level based on the white ratio and the threshold value at which the change in brightness level is easy to be clarified. Is possible. This makes it possible to improve the detection sensitivity.

[0038]

DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be described below with reference to FIGS. FIG.
1 is a configuration diagram of a pattern defect inspection device 20 of the present invention.

The pattern defect inspection device 20 includes the image pickup device 2
One. The image pickup apparatus 21 includes a pattern 22 to be inspected such as a photomask as shown in FIGS. 3, 4, and 8.
Is input as an image and captured as an image signal, and is configured by a line sensor or the like. An A / D converter 23 is connected to the imaging device 21. A
The / D converter 23 performs A / D conversion on the image signal, whereby the image signal, which is an analog signal, is digitized and converted into inspection data.

The A / D converter 23 is connected to a first data extracting circuit 24. The first data cut-out circuit 24 cuts out the inspection data in a direction perpendicular to the edge direction of the pattern 22 to be inspected, whereby the first white ratio conversion circuit 30 cuts out the inspection data. The white ratio is calculated.

Further, a reference data generation circuit 25 is provided for obtaining reference data to be compared with the inspection data. The reference data generation circuit 25 inputs position data corresponding to the position where the image is captured by the imaging device 21 and generates reference data by electrically developing CAD data corresponding to the inspection data at each position. It is.

The reference data generating circuit 25 is connected to an edge direction detecting circuit 26 and a second data extracting circuit 27. As shown in FIG. 2, the edge direction detecting circuit 26 has a differentiating circuit 28, which differentiates the reference data into respective differential parameters in the X direction, the Y direction, and ± 45 degrees, and calculates the minimum value by the differentiation. Is detected as an edge direction. That is, the direction in which the brightness change is the least in the reference data is detected as the edge direction.

Here, FIG. 3 shows an example of the definition of the differentiation in the X direction, the Y direction, and the ± 45 ° direction calculated from the 3 × 3 neighboring pixels which are the target pixel. In this figure, | 2 is the vertical direction (Y direction) with respect to the edge of the pattern 22 to be inspected.
When the differential processing is performed toward the −8 | pixel, the differential value in the Y direction becomes 0, and the edge direction becomes the Y direction. In the case where the differential processing is performed toward another direction, for example, the pixel of | 4-6 | which is the X direction, or the differential processing is performed toward the pixels of | 1-9 | and | 3-7 | When performing, because there is a difference in brightness change,
In any case, the differentiated value does not become 0, and the edge direction is determined as described above.

The edge direction detection circuit 26 is connected to first and second data cutout circuits 24 and 27. Based on the detection result of the edge direction by the edge direction detection circuit 26, The first and second data extracting circuits 24 and 27 extract data.

The first and second data cutout circuits 24 and 27 are respectively provided with the first and second white ratio conversion circuits 3
0,31. These first and second white ratio conversion circuits 30 and 31 convert the cut-out data into a white ratio in pixel units. Here, the white ratio is
As shown in FIG. 4, when the three pixel data groups provided in series are cut out, the brightness ratio of the pixel located at the center, which is the target pixel, in the cut out pixel data group is shown. is there. For example, when the edge is at the center of the pixel of interest, the white ratio is 0.5. When the bright and dark portions occupy 80% of the target pixel, the white ratios are 0.8 and 0.2, respectively.

Here, the white ratio can be obtained by calculation by giving data of three pixels of the extracted pixel data group, in other words, a brightness level and a threshold level existing between a white portion and a black portion. is there.

The method of calculating the white ratio will be described with reference to FIG. First, as shown in FIG. 5A, the brightness level data of each of the cut-out serial pixel data groups P ₁ , P ₂ , and P ₃ is sequentially expressed as P ₁ , P ₂ , P
₃ and when the cross section profile showing a brightness level of P _1, P _2, P ₃ are as shown in FIG.

Here, when the cross-sectional profile of the brightness level changes as shown in (i), when the threshold value is set to th shown in FIG. Since the level is small, the white ratio is 0.

When the sectional profile of the brightness level changes as in (ii), P ₁ <th, P ₂
Since <th, P ₃ ≧ th, the white ratio is as follows: white ratio = １ ／ · (P ₃ −th) / (P ₃ −P ₂ ). Hereinafter, the brightness level is similarly obtained for each cross-sectional profile. FIG. 5 shows how to determine the white ratio in each of the cross-sectional profiles in FIG.

In FIG. 5, the calculation method is described not only when the brightness changes on the left and right sides of the edge pattern but also when the brightness changes only in the center pixel.

The first and second white ratio conversion circuits 3
According to 0 and 31, the data of the white ratio obtained in each of the inspection data and the reference data is input to the defect determination circuit 32. The defect determination circuit 32 has a subtraction circuit 33 as shown in FIG. 7, and the data of the white ratio obtained by the first and second white ratio conversion circuits 30 and 31 are input to the subtraction circuit 33, respectively. Then, the difference between the white ratio of the inspection data and the reference data is calculated.

The defect determining circuit 32 is provided with a determining circuit 34.
The result of the difference between the white ratios calculated by the subtraction circuit 33 is sent, and the set predetermined threshold value is input. Threshold, as shown in FIG. 8 (b), in the pixel of interest P _2, when both bright and the difference is small, by setting the threshold slightly lower than the bright level "224" of bright values may be put, in the pixel of interest P ₂ Conversely, when neither dark the difference is small, it is sufficient to set the threshold to a value slightly higher than the dark level "0" of the dark values.

Here, the threshold value input to the judgment circuit 34 is different from the threshold value in the cross-sectional profile.
As shown in FIG. 10, when the difference between the white ratios is equal to or greater than a predetermined value, the difference is determined as a defect. That is, if the threshold value is exceeded, this is automatically determined as a defect. However, it is also possible to provide a margin for this threshold value within a predetermined range, and to set the threshold value not to be judged as a defect if the margin is within this margin range.

Then, based on the difference between the threshold value, the white ratio of the inspection data, and the white ratio of the reference data, the detection of the defect occurring in the pattern 22 to be inspected is determined. In addition,
In the cross-sectional profile, thresholds are usually set at two levels, a bright value and a dark value, so that a comparison can be made at both the bright level and the dark level. This makes it possible to make a determination using the threshold values of both the bright level and the dark level at the stage of obtaining the white ratio.

However, it is not always necessary to set the threshold value in the cross-sectional profile to two levels. The threshold value is a light value slightly lower than the light level “224” or a light value slightly higher than the dark level “0”. A configuration in which the threshold value is set to only one of the dark values may be used.

The above-described pattern defect inspection apparatus 20 has first and second data cutout circuits 24 and 27 and is different from the conventional pattern defect inspection apparatus. The first and second
, And a defect determination circuit 32 connected to the first and second white ratio conversion circuits 30 and 31.

And especially in this characteristic part,
It has an effect that has not existed before. Hereinafter, this point will be described. The first and second data extraction circuits 24,
In 27, pixel data groups P ₁ , P ₃ , P ₃ , P ₃ , P ₃ , P ₃ , P ₃ , P ₃ , P ₃ , and P 3 at the center pixel of interest P ₂ are pixel data having a bright portion and a dark portion at an edge. Cut out P ₂ and P ₃ .

The extracted pixel data groups P ₁ , P
₂ and P ₃ , the first and second white ratio conversion circuits 3
At 0 and 31, a threshold level is given, and when the values of P _{1 to} P ₃ exist across this threshold, the white ratio is set to any value between 0 and 1 Become. Conversely, when the values of P _{1 to} P ₃ exist without straddling this threshold value, the white ratio becomes either 0 or 1.

Here, a problem is a defect having a small difference in brightness level, that is, P ₁ and P ₂ or P _2.
As is the case with little difference in brightness level between the P _3. In this case, the conventional pattern defect inspection apparatus may not determine a defect unless the defect is equal to or more than the threshold value or less than the threshold value.

FIG. 8 shows an example of a cross-sectional profile when the difference between the brightness levels is small and a pattern to be inspected showing such a cross-sectional profile. As shown in this cross-sectional profile, when the difference in the brightness level is small, the gradient of the cross-sectional profile is closer to the reference data side than the defect of the normal pattern to be inspected, and the gradient angle is gentle. Become. That is, in this figure, the side of the bright level is closer to the reference data side, so that the gradient angle is gentle.

However, as shown in FIG. 8, by setting the threshold level in this case to a sufficiently low value and slightly higher than the dark level “0”, P ₂ is set to be equal to or more than the threshold value. Accordingly, even if the difference in brightness level is small, the defect determination circuit 32 can determine whether or not a defect exists.

As shown in FIG. 9, even in the case of a defect having a small difference in brightness level, the output signal can be output to a signal level as a normal defect.

According to the pattern defect inspection apparatus 20 described above, after the edge direction detection circuit 26 detects the edge direction of the pattern 22 to be inspected, the inspection data
Data extraction circuit 24, and reference data
Is a configuration in which a pixel data group P ₁ , P ₂ , P ₃ composed of three pixels is extracted by the data extraction circuit 27.
Thereby, the first and second white ratio conversion circuits 30,
Since the calculation of the white ratio is performed at 31, it is possible to reflect a minute difference in the brightness level from the calculation of the white ratio.

Then, a difference between the calculated white ratio of the inspection data and the white ratio of the reference data is calculated, and based on the difference between the calculated white ratio and a predetermined threshold value, whether or not this brightness level corresponds to a defect is determined. Is determined by the defect determination circuit 32, so that it is possible to satisfactorily detect whether or not there is a defect even with this slight difference in brightness level.

Since the edge direction is detected using the reference data in the edge direction detection circuit 26, the edge direction can be accurately detected. Further, the pixel data groups P ₁ , P ₂ , and P ₃ have a target pixel P ₂ at the center, and the white ratio can be easily derived from the pixel data before and after the target pixel P _2. ing.

In the cross-sectional profile, the brightness level is set to two levels, a level slightly higher than the dark level “0” and a level slightly lower than the light level “224”, so that the brightness level is low. The detection sensitivity of the defect on the side and on the side with the higher brightness level is improved, thereby enabling good defect detection.
That is, it is possible to improve the detection sensitivity of a defect having a small change in the brightness level.

While the embodiment of the present invention has been described above, the present invention can be variously modified. This is described below. In the above embodiment, 3
Pixel data groups P ₁ , P
While using a _2, P _3, not limited to this pixel data group, it may be configured to use three or more pixels of the pixel data group.

In the above embodiment, the first and second data cutout circuits 24 and 27 and the first and second white ratio conversion circuits 30 and 31 corresponding thereto are used. Alternatively, these two circuits may be unified. In addition, various modifications can be made without departing from the scope of the present invention.

[0069]

As described above, according to the present invention,
The edge direction is determined, inspection data and reference data are cut out in a direction perpendicular to the edge direction, and the degree of brightness (white ratio) at the pixel of interest is compared with the cut-out image data group by a white ratio conversion unit. Can be calculated.

Here, by calculating the white ratio, a minute difference in the brightness level can be reflected. Then, based on the calculated white ratio of the inspection data, the white ratio of the reference data, and a preset threshold value, it is determined whether or not the brightness level corresponds to a defect. From the setting of the threshold value and the white ratio reflecting the minute brightness, it is possible to satisfactorily find even a defect with a small change in the brightness level.

[Brief description of the drawings]

FIG. 1 is a diagram showing a configuration of a pattern defect inspection apparatus according to an embodiment of the present invention.

FIG. 2 is a diagram showing a configuration of an edge direction detection circuit according to the embodiment;

FIG. 3 shows an X direction, a Y direction, and ± 4 according to the embodiment.
The figure which shows an example of the definition of the differentiation of the 5 degree direction.

FIG. 4 is a view showing a state of a white ratio and a cross-sectional profile of pixel data cut out from the edge shape of the pattern to be inspected according to the embodiment;

5A is a diagram showing a pixel data group in a direction perpendicular to an edge direction, and FIG. 5B is a diagram showing a calculation method of a white ratio according to the embodiment;

FIG. 6 is a view showing a state of a cross-sectional profile in each case of the calculation method of the white ratio in FIG. 5 according to the embodiment.

FIG. 7 is a diagram showing a configuration of a defect determination circuit according to the embodiment.

FIG. 8 is a diagram showing a relationship between a defect and a cross-sectional profile of a pattern to be inspected and a threshold value according to the embodiment;
(A) shows a case of a normal defect, and (b) shows a case of a small difference in brightness level.

FIG. 9 is a diagram showing a defect signal level in a cross-sectional profile in FIG. 9 according to the embodiment.

FIG. 10 is a diagram showing a relationship between a threshold value in the determination circuit according to the embodiment and a difference in white ratio between inspection data and reference data.

FIG. 11 is a diagram showing a configuration of a conventional pattern defect inspection apparatus.

FIG. 12 is a schematic view showing a mask pattern of a conventional Cr film.

FIG. 13 is a diagram showing levels of inspection and reference data for a conventional Cr film mask pattern.

FIG. 14 is a diagram showing a state of a signal level when a change in signal level and brightness level of a conventional normal defect is small.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 20 ... Pattern defect inspection apparatus 21 ... Imaging apparatus 22 ... Inspection pattern 23 ... A / D converter 24 ... First data extraction circuit 25 ... Reference data generation circuit 26 ... Edge direction detection circuit 27 ... Second data extraction circuit 28 differentiating circuit 30 first white ratio converting circuit 31 second white ratio converting circuit 32 defect determining circuit 33 subtracting circuit

Claims (9)

[Claims] 1. A pattern defect inspection apparatus for inspecting a defect on an object to be inspected by comparing inspection data obtained by imaging an object to be inspected with reference data thereof, wherein an edge direction of a pattern of the object to be inspected is An edge direction detecting means to be detected, and a pixel data group having a target pixel which is directed in a direction perpendicular to the edge direction detected by the edge direction detecting means and has an edge of a pattern of an object to be inspected straddling any pixel. Data extraction means for extracting inspection data and reference data; and a ratio (white ratio) between a dark part and a bright part in each pixel of the pixel data group of the inspection data and the pixel data group of the reference data extracted by the data extraction means. A white ratio conversion unit to calculate, a white ratio of the inspection data, a white ratio of the reference data, and a preset white ratio calculated by the white ratio conversion unit. Was based on a threshold, the pattern defect inspection apparatus characterized by comprising: a defect determining unit determines the presence or absence of defects at a predetermined position of the device under test. 2. The pattern defect inspection apparatus according to claim 1, wherein the edge direction of the inspection object is recognized by detecting an edge direction of a pattern of reference data. 3. The pixel data group according to claim 3, wherein the pixel data group is provided in series, and includes pixel data of three pixels in which a target pixel straddling an edge of an object to be inspected is located at a center. 3. The pattern defect inspection device according to claim 1 or 2. 4. The defect determining means compares a difference between a white ratio of the inspection data and a white ratio of the reference data with a predetermined threshold value, and determines that the difference between the white ratios is a predetermined threshold value. The pattern defect inspection apparatus according to claim 1, wherein the pattern is determined to be a defect when the pattern is crossed. 5. The method according to claim 1, wherein the threshold value has a value closer to the light portion side than an intermediate value between the level of the light portion and the level of the dark portion. The pattern defect inspection device according to any one of the above. 6. The method according to claim 1, wherein the threshold value has a value closer to the dark part side than a value intermediate between the level of the light part and the level of the dark part. A pattern defect inspection apparatus according to any one of the above. 7. The threshold value has a value closer to the light part side than an intermediate value between the light part level and the dark part level, and has a value closer to the dark part side than the intermediate value. The pattern defect inspection apparatus according to claim 1, further comprising a value. 8. A defect determining means for determining a difference between the white ratio of the inspection data and the reference data, a determining means for determining a defect based on the difference between the white ratio and a threshold value obtained by the subtracting means, The pattern defect inspection apparatus according to any one of claims 1 to 7, comprising: 9. A pattern defect inspection method for inspecting a defect on the inspection object by comparing inspection data obtained by imaging the inspection object with reference data, wherein the edge direction of the pattern of the inspection object is An edge direction detecting step of detecting, a data cutout step of cutting out a pixel data group having a pixel of interest in which the edge of the pattern of the pattern to be inspected straddles any pixel for the test data and the reference data, respectively, A white ratio conversion step of calculating a ratio (white ratio) between a dark part and a light part in each pixel of the pixel data group of the inspection data and the pixel data group of the reference data, and the inspection calculated by the white ratio conversion step Based on the white ratio of the data, the white ratio of the reference data, and a preset threshold, the presence or absence of a defect at a predetermined position on the inspection object is determined. Pattern defect inspection method characterized by comprising the defect determination step of, a.