Inspection equipment and inspection method

Abstract

The present invention provides an inspection device and inspection method that can improve the accuracy of defect detection. [Solution] The inspection apparatus 10 according to the present disclosure includes: an identification unit 11 that identifies a continuous region 53 in a sample 50 on which a pattern 51 is formed, in which substantially equivalent predetermined properties are continuous in a first direction; an acquisition unit 12 that acquires a difference image G30 based on a comparison between an imaged image G10 of the sample 50 and a reference image G20; and a determination unit 13 that determines a defect in a predetermined processing area within the region corresponding to the continuous region 53 in the difference image G30, based on parameters based on the difference image G30 of the central region, parameters based on the difference image G30 of a first peripheral region on one side of the central region in the first direction, and parameters based on the difference image G30 of a second peripheral region on the other side of the central region in the first direction.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to an inspection apparatus and an inspection method. BACKGROUND ART

[0002] Patent Document 1 describes an inspection apparatus that inspects defects in semiconductor-related samples such as masks using an optical device. PRIOR ART DOCUMENTS PATENT DOCUMENTS

[0003] PATENT DOCUMENT 1 Japanese Unexamined Patent Publication No. 2021-143840 SUMMARY OF THE INVENTION PROBLEMS TO BE SOLVED BY THE INVENTION

[0004] With the recent miniaturization and complication of the structures of semiconductor-related samples, the shapes of defects to be detected and the inspection target regions have diversified. Defects in semiconductor-related samples affect the yield and device operation of semiconductor devices, and thus need to be detected appropriately.

[0005] An object of the present disclosure is made to solve such a problem, and is to provide an inspection apparatus and an inspection method capable of improving defect detection accuracy. MEANS FOR SOLVING THE PROBLEMS

[0006] The inspection apparatus according to this disclosure includes: an identification unit that identifies a continuous region in a patterned sample in which substantially equivalent predetermined properties are continuous in a first direction; an acquisition unit that acquires a difference image based on a comparison between an image of the sample and a reference image; and a determination unit that determines a defect in a predetermined processing area within the region corresponding to the continuous region in the difference image, based on parameters of the difference image in the central region, parameters of the difference image in a first peripheral region on one side of the central region in the first direction, and parameters of the difference image in a second peripheral region on the other side of the central region in the first direction.

[0007] In the above inspection apparatus, the determination unit may identify a defect candidate region in the difference image that includes pixels having parameters within a predetermined range, and the region including the defect candidate region may be designated as the processing region.

[0008] In the inspection apparatus described above, the determination unit may convert the parameters into converted parameters by applying a conversion operator that converts the parameters of each pixel of the difference image to the processing area, and then determine the defect based on the converted parameters.

[0009] In the inspection apparatus described above, the conversion operator is a kernel having a central part containing one or more cells, a first peripheral part containing two or more cells on one side of the central part, and a second peripheral part containing two or more cells on the other side of the central part. The determination unit calculates a first defect signal from the parameters of a pixel of interest, which is a pixel of the difference image at a position corresponding to a cell in the central part, and a first statistical value of the parameters of a first peripheral region pixel group consisting of pixels of the difference image at positions corresponding to cells in the first peripheral part. The determination unit calculates a second defect signal from the parameters of the pixel of interest and a second statistical value of the parameters of a second peripheral region pixel group consisting of pixels of the difference image at positions corresponding to cells in the second peripheral part. Based on the conditions that the first defect signal and the second defect signal satisfy, the parameters of the pixel of interest may be converted to the conversion parameters.

[0010] In the above-described inspection device, the first statistical value and the second statistical value may include at least one of the mean and the standard deviation.

[0011] In the above-described inspection apparatus, the conditions may include a first condition in which the sign of the first defect signal and the sign of the second defect signal are equal, and a second condition in which the absolute value of the first defect signal is greater than the value obtained by dividing the absolute value of the second defect signal by a number greater than 1, and the absolute value of the second defect signal is greater than the value obtained by dividing the absolute value of the first defect signal by a number greater than 1.

[0012] In the inspection apparatus described above, the determination unit may convert the parameter of the pixel of interest into a defect signal sum, which is the sum of the first defect signal and the second defect signal, if both the first and second conditions are met, and convert the parameter of the pixel of interest into a specific value if at least one of the first and second conditions is not met.

[0013] In the inspection apparatus described above, the determination unit may determine the defect in the central region based on a first defect signal obtained by correcting the pixel parameters of the difference image in the central region based on a first statistical value of the pixel parameters of the difference image in the first peripheral region, and a second defect signal obtained by correcting the pixel parameters of the difference image in the central region based on a second statistical value of the pixel parameters of the difference image in the second peripheral region.

[0014] In the inspection apparatus described above, the determination unit may determine the defect in the central region based on a first condition which includes that the code of the first defect signal and the code of the second defect signal have a predetermined relationship.

[0015] In the inspection apparatus described above, the determination unit may determine the defect in the central region based on a second condition, which includes the fact that both the first defect signal and the second defect signal are of similarly large values.

[0016] In the inspection apparatus described above, the determination unit may determine the defect in the central region when both of the following conditions are met: a first condition, which includes that the sign of the first defect signal and the sign of the second defect signal have a predetermined relationship; and a second condition, which includes that both the first defect signal and the second defect signal have values ​​that are equally large.

[0017] In the inspection apparatus described above, the sample has a plurality of pairs of two patterns extending in a second direction intersecting the first direction, or pairs of two patterns extending in a second direction intersecting the first direction, where at least one of them is adjacent in the second direction with a gap between them, and these pairs are spaced apart in the first direction. The continuous region may include a plurality of gaps aligned in the first direction.

[0018] In the above inspection apparatus, the sample has multiple sets of circular patterns arranged in a second direction intersecting the first direction with gaps in between, spaced apart in the first direction, and the continuous region may include multiple gaps arranged in the first direction.

[0019] In the above inspection apparatus, the sample may contain the defect in the gap portion.

[0020] The inspection method according to this disclosure comprises the steps of: identifying a continuous region in a patterned sample in which substantially equivalent predetermined properties are continuous in a first direction; acquiring a difference image based on a comparison between an image of the sample and a reference image; and determining a defect in a predetermined processing region within the region corresponding to the continuous region in the difference image, based on parameters of the central region based on the difference image, parameters of a first peripheral region on one side of the central region in the first direction based on the difference image, and parameters of a second peripheral region on the other side of the central region in the first direction based on the difference image.

[0021] In the above inspection method, in the determining step, a defect candidate region including pixels having parameters in a predetermined range in the difference image may be specified, and a region including the defect candidate region may be set as the processing region.

[0022] In the above inspection method, in the determining step, the parameters may be converted into converted parameters by applying a conversion operator that converts the parameter of each pixel of the difference image to the processing region, and the defect may be determined based on the converted converted parameters.

[0023] In the above inspection method, the conversion operator is a kernel having a central part including one or more cells, a first peripheral part including two or more cells on the one side of the central part, and a second peripheral part including two or more cells on the other side of the central part, and in the determining step, a first defect signal is calculated from the parameter of a target pixel that is a pixel of the difference image at a position corresponding to the cell in the central part, and a first statistical value of the parameters of a first peripheral region pixel group consisting of pixels of the difference image at positions corresponding to the cells in the first peripheral part, a second defect signal is calculated from the parameter of the target pixel, and a second statistical value of the parameters of a second peripheral region pixel group consisting of pixels of the difference image at positions corresponding to the cells in the second peripheral part, and the parameter of the target pixel may be converted into the converted parameter based on a condition satisfied by the first defect signal and the second defect signal.

[0024] In the above inspection method, the first statistical value and the second statistical value may include at least one of an average value and a standard deviation.

[0025] In the above inspection method, in the determining step, the conditions may include a first condition that the sign of the first defect signal and the sign of the second defect signal are equal, and a second condition that the absolute value of the first defect signal is larger than a value obtained by dividing the absolute value of the second defect signal by a number greater than 1, and the absolute value of the second defect signal is larger than a value obtained by dividing the absolute value of the first defect signal by a number greater than 1.

[0026] In the above inspection method, in the determining step, when both the first condition and the second condition are satisfied, the parameter of the target pixel may be converted into a defect signal sum that is the sum of the first defect signal and the second defect signal, and when at least one of the first condition and the second condition is not satisfied, the parameter of the target pixel may be converted into a specific value.

[0027] In the above inspection method, in the determining step, the defect in the central region may be determined based on a first defect signal obtained by correcting the parameter of the pixel of the difference image in the central region based on a first statistical value of the parameters of the pixels of the difference image in the first peripheral region, and a second defect signal obtained by correcting the parameter of the pixel of the difference image in the central region based on a second statistical value of the parameters of the pixels of the difference image in the second peripheral region.

[0028] In the above inspection method, in the determining step, the defect in the central region may be determined based on a first condition including that the sign of the first defect signal and the sign of the second defect signal have a predetermined relationship.

[0029] In the above inspection method, in the determining step, the defect in the central region may be determined based on a second condition including that both the first defect signal and the second defect signal are values that are large to the same extent.

[0030] In the above inspection method, in the determining step, the defect in the central region may be determined when both of the first condition including that the sign of the first defect signal and the sign of the second defect signal have a predetermined relationship and the second condition including that both the first defect signal and the second defect signal are values that are large to the same extent are satisfied.

[0031] In the above inspection method, the sample has multiple pairs of two patterns extending in a second direction intersecting the first direction, or pairs of two patterns extending in a second direction intersecting the first direction, where at least one of them is adjacent in the second direction with a gap between them, and these pairs are spaced apart in the first direction. The continuous region may include multiple gaps aligned in the first direction.

[0032] In the above inspection method, the sample has multiple sets of circular patterns arranged in a second direction intersecting the first direction with gaps in between, spaced apart in the first direction, and the continuous region may include multiple gaps arranged in the first direction.

[0033] In the above inspection method, the sample may contain the defect in the gap portion. [Effects of the Invention]

[0034] This disclosure provides an inspection apparatus and an inspection method that can improve the accuracy of defect detection. [Brief explanation of the drawing]

[0035] [Figure 1] This is a block diagram illustrating an inspection device according to Embodiment 1. [Figure 2] This figure illustrates an image of a sample captured in the inspection apparatus according to Embodiment 1. [Figure 3] This figure illustrates a reference image of a sample in the inspection apparatus according to Embodiment 1. [Figure 4] This figure illustrates the difference image of a sample in the inspection apparatus according to Embodiment 1. [Figure 5] This figure illustrates a kernel used by the determination unit in the inspection device according to Embodiment 1. [Figure 6] This figure illustrates the converted image of a sample in the inspection apparatus according to Embodiment 1. [Figure 7] This figure illustrates another kernel used by the determination unit in the inspection apparatus according to Embodiment 1. [Figure 8] This is a block diagram illustrating an information processing device according to Embodiment 1. [Figure 9] This is a flowchart illustrating an example of the inspection method according to Embodiment 1. [Figure 10] This figure illustrates an image of a sample captured in the inspection apparatus according to Embodiment 2. [Figure 11] This figure illustrates a reference image of a sample in the inspection apparatus according to Embodiment 2. [Figure 12] This figure illustrates the difference image of a sample in the inspection apparatus according to Embodiment 2. [Figure 13] This figure illustrates an image of a sample captured by the inspection apparatus according to Embodiment 2. [Figure 14] This figure illustrates a reference image of a sample in the inspection apparatus according to Embodiment 2. [Figure 15] This figure illustrates the difference image of a sample in the inspection apparatus according to Embodiment 2. [Figure 16] This figure illustrates an image of a sample captured in the inspection apparatus according to Embodiment 2. [Figure 17] This figure illustrates a reference image of a sample in the inspection apparatus according to Embodiment 2. [Figure 18] This figure illustrates the difference image of a sample in the inspection apparatus according to Embodiment 2. [Figure 19] This is a block diagram illustrating an inspection device according to Embodiment 3. [Figure 20] This is a flowchart illustrating an example of the inspection method according to Embodiment 3. [Figure 21] This is a diagram illustrating an inspection device according to Embodiment 4. [Modes for carrying out the invention]

[0036] The specific configuration of this embodiment will be described below with reference to the drawings. The following description illustrates preferred embodiments of the present disclosure, and the scope of the present disclosure is not limited to the following embodiments. In the following description, the same reference numerals indicate substantially the same thing. Some reference numerals and hatching have been omitted to avoid cluttering the drawings.

[0037] <Embodiment 1> The inspection apparatus according to Embodiment 1 will now be described. Figure 1 is a block diagram illustrating the inspection apparatus 10 according to Embodiment 1. As shown in Figure 1, the inspection apparatus 10 of this embodiment includes a identification unit 11, an acquisition unit 12, and a determination unit 13.

[0038] (Specific part) The identification unit 11 identifies a continuous region in the patterned sample in which substantially equivalent predetermined properties are continuous in a first direction. The identification unit 11 may identify a continuous region in which substantially equivalent predetermined properties are continuous in a first direction in an image of the patterned sample, or it may identify a continuous region in which substantially equivalent predetermined properties are continuous in a first direction in a reference image of the patterned sample. As an example, the identification unit 11 may define a first direction and identify a region consisting of a specific number of pixels arranged in an arbitrary direction (vertical, horizontal, diagonal) in the image plane of the imaged image G10 or the reference image G20 described later as a continuous region if the difference between the maximum and minimum values ​​of the parameters (e.g., brightness) of a specific number of pixels arranged in an arbitrary direction (vertical, horizontal, diagonal) remains within a predetermined range.

[0039] Furthermore, the identification unit 11 may identify a continuous region in which substantially equivalent predetermined properties are continuous in a first direction, based on the design information of the pattern formed on the sample. In this case, the design information may include identification information indicating the continuous region.

[0040] Figure 2 is a diagram illustrating an image G10 of a sample 50 in the inspection apparatus 10 according to Embodiment 1. Figure 3 is a diagram illustrating a reference image G20 of a sample 50 in the inspection apparatus 10 according to Embodiment 1. As shown in Figures 2 and 3, a line and space pattern 51 may be formed on the sample 50. The pattern 51 may include, for example, a first material that reflects illumination light. The space between the patterns 52 between the patterns 51 may include a second material that absorbs illumination light. Alternatively, the pattern 51 may include a first material that transmits illumination light. The space between the patterns 52 may include a second material that absorbs or reflects illumination light. As a result, in the image G10 and reference image G20, the portion of the pattern 51 includes a white portion. Also, in the image G10 and reference image G20, the portion between the patterns 52 includes a black portion. In the following explanation, the pattern 51 portion will be assumed to include a white portion, and the portion between the patterns 52 will be assumed to include a black portion. However, the pattern 51 portion may be assumed to include a black portion, and the portion between the patterns 52 may be assumed to include a white portion.

[0041] Pattern 51 and the space between patterns 52 have portions that extend in one direction. The direction in which Pattern 51 and the space between patterns 52 extend is called the first direction. In this case, the sample 50 has a continuous region 53 in which substantially equivalent predetermined properties are continuous in the first direction. The substantially equivalent predetermined property may be luminance. The continuous region 53 in the captured image G10 corresponds to the continuous region 53 of the sample 50. The continuous region 53 in the captured image G10 may include a space between patterns 52 in which black portions with a luminance smaller than the predetermined luminance are continuous in the first direction. Also, the continuous region 53 in the captured image G10 may include a pattern 51 in which white portions with a luminance greater than the predetermined luminance are continuous in the first direction. Note that the substantially equivalent predetermined property is not limited to luminance, but may be other properties such as wavelength due to the color of reflected light. Also, the substantially equivalent predetermined property may be the first material or the second material in the design information. The identification unit 11 identifies the continuous region 53 including the space between patterns 52 as an example.

[0042] Reference image G20 may be an image of sample 50. For example, reference image G20 may be an image of a good sample 50 that does not contain any defects. Also, reference image G20 may be an image of a part of sample 50 that is different from the part captured in captured image G10. Furthermore, reference image G20 may be an image generated from design information. The continuous region 53 in reference image G20 corresponds to the continuous region 53 of sample 50. The continuous region 53 in reference image G20 may include pattern 51 and may also include the space between patterns 52, just as in captured image G10.

[0043] (Acquisition Department) Figure 4 is a diagram illustrating a difference image G30 of a sample 50 in an inspection apparatus 10 according to Embodiment 1. As shown in Figure 4, the acquisition unit 12 acquires a difference image G30 based on a comparison between the captured image G10 and the reference image G20 of the sample 50. The acquisition unit 12 acquires the difference image G30 by taking the difference between the parameters of each pixel in the captured image G10 and the parameters of each pixel in the reference image G20. When taking the difference, the acquisition unit 12 subtracts each pixel in the captured image G10 from each pixel in the reference image G20 that corresponds to each of those pixels. The difference image G30 includes an image region obtained by subtracting the pixels that captured a continuous region 53 in the captured image G10 from the pixels in the reference image G20 that correspond to the same portion of the continuous region 53 captured by the captured image G10. This image region may be referred to as "the region corresponding to the continuous region 53 in the difference image G30," or simply as the continuous region 53 of the difference image 30 (the continuous region 53 in the difference image G30). The acquisition unit 12 may subtract the pixels that captured the continuous region 53 in the captured image G10 from the pixels in the reference image G20 that correspond to the same portion as the continuous region 53 captured in the captured image G10. The acquisition unit 12 only needs to acquire an image based on a comparison between the captured image G10 and the reference image G20 of the sample 50, and the acquisition unit 12 may acquire the difference image G30 by taking the ratio of the parameters of each pixel in the captured image G10 to the parameters of each pixel in the reference image G20.

[0044] The parameter for each pixel in the difference image G30 is, for example, luminance (luminance difference value). The difference image G30 may be a virtual one for the defect detection process described later, and it is not essential that it be generated and displayed as an image. Therefore, the acquisition unit 12 may acquire comparison parameters for multiple pixels by comparing the pixel parameter (luminance) in the captured image G10 of the sample 50 with the pixel parameter (luminance) in the corresponding reference image G20. This configuration of the acquisition unit 12 is also within the scope of the technical concept of this disclosure. Furthermore, in the following description, the pixel parameter of the difference image G30 may be read as the comparison parameter corresponding to the pixel in the captured image G10 or the reference image G20 (a comparison parameter obtained by comparing the pixel parameter (luminance) in the captured image G10 of the sample 50 with the pixel parameter (luminance) in the corresponding reference image G20). This configuration is also within the scope of the technical concept of this disclosure.

[0045] (Judgment Department) The determination unit 13 determines defects based on the pixel parameters of the difference image G30. Specifically, the determination unit 13 determines defects in a predetermined processing area within the region corresponding to the continuous region 53 in the difference image G13, based on the parameters of the difference image G30 of the central region, the parameters of the difference image G30 of the first peripheral region on one side in the first direction from the central region, and the parameters of the difference image G30 of the second peripheral region on the other side in the first direction from the central region.

[0046] (kernel) Figure 5 is a diagram illustrating a kernel W used by the determination unit 13 in the inspection apparatus 10 according to Embodiment 1. Figure 5 also shows the distribution KP of parameters (luminance) on pixels of the difference image G30 at positions corresponding to each cell of the kernel W. As shown in Figure 5, the kernel W includes a plurality of cells arranged in a first direction in which the continuous region 53 extends. The size of each cell may virtually correspond to the size of pixels in the difference image G30. The kernel W has a central part W0 including the cells in the central part, a first peripheral part W1 including the cells on one side of the central part W0 in the first direction, and a second peripheral part W2 including the cells on the other side of the central part W0 in the first direction.

[0047] The central region W0 contains one or more cells. The first peripheral region W1 contains two or more cells. The second peripheral region W2 contains two or more cells. In the figure, as an example, a kernel W is shown in which the central region W0 contains one cell, and the first peripheral region W1 and the second peripheral region W2 each contain four cells. The cell in the central region W0 is called the cell of interest CC0. The pixel in the difference image G30 at the position corresponding to the cell of interest CC0 is called the pixel of interest CP0. There may be a gap between the cell of interest CC0 and the cell on the other side of the first peripheral region W1. There may also be a gap between the cell of interest CC0 and the cell on one side of the second peripheral region W2. The size of these gaps may be changed depending on the size of the detected defect. The kernel W functions as a transformation operator that transforms the parameters of each pixel in the difference image G30.

[0048] The region to which kernel W is applied to the difference image G30 is called the processing region. The contiguous region 53 in the difference image G30 may have a wider range than the processing region. The determination unit 13 applies kernel W to the entire contiguous region 53 of the difference image G30 in a sweeping manner, thereby sequentially making the entire contiguous region 53 the processing region. The determination unit 13 converts the pixel parameters of the difference image G30 into conversion parameters by applying kernel W as a conversion operator to the processing region in the contiguous region 53 of the difference image G30. If the pixel parameters of the difference image G30 include brightness, the determination unit 13 converts the brightness of the difference image G30 into conversion brightness.

[0049] The following describes the process performed by the determination unit 13 to convert the pixel parameters of the difference image G30, determined by the kernel W, into conversion parameters, divided into processes A, B, and C. The process of converting the pixel parameters of the difference image G30 into conversion parameters can also be described as the process of converting the difference image G30 into the conversion image G40 described later.

[0050] (Process A) The determination unit 13 applies the kernel W to the processing area of ​​the difference image G13 to calculate statistical values ​​for the parameters (luminance) of the pixels in the difference image G13 at positions corresponding to each cell in the first peripheral area W1 (referred to as the first peripheral area pixel group). Here, the average value is used as the statistical value for explanation. The statistical value may include at least one of the average value and the standard deviation. The case in which the statistical value includes the standard deviation will be described later. As an example, as shown in equation (1) below, the determination unit 13 calculates the average value for the parameters (luminance) of the first peripheral area pixel group as the first statistical value. The determination unit 13 also applies the kernel W to the processing area of ​​the difference image G13 to calculate statistical values ​​for the parameters (luminance) of the pixels in the difference image G13 at positions corresponding to each cell in the second peripheral area W2 (referred to as the second peripheral area pixel group). As an example, as shown in equation (2) below, the determination unit 13 calculates the average value for the parameters (luminance) of the second peripheral area pixel group as the second statistical value. Note that in equations (1) and (2), the first statistical value is denoted by TV_1 and the second statistical value by TV_2. Also, in this example, pixels 2 to 5 belong to the first peripheral region pixel group, and pixels 6 to 9 belong to the second peripheral region pixel group.

[0051] TV_1 = Average value of the parameters (luminance) of pixels 2 to 5 (1) TV_2 = Average value of parameters (brightness) for pixels 6 to 9 (2)

[0052] The determination unit 13 applies the kernel W to the processing area of ​​the difference image G13 to obtain the parameter (luminance) of the pixel (target pixel CP0) of the difference image G13 at the position corresponding to the target cell CC0 in the central part W0 of the kernel W. In the figure, the target pixel CP0 is shown as pixel 1. The parameter (luminance) of the target pixel CP0 is called the target value. Even if luminance is the parameter, it is also called the target value. If the central part W0 contains multiple cells, the determination unit 13 may calculate the statistical value of the parameter (luminance) of each pixel in the difference image G13 at the position corresponding to each target cell CC0 in the central part W0, that is, the statistical value of the parameter (luminance) of multiple target pixels CP0.

[0053] (Process B) The determination unit 13 calculates a first defect signal from the value of interest and the first statistical value. Specifically, as shown in equation (3) below, the determination unit 13 calculates the first defect signal from the difference between the value of interest and the first statistical value (mean value). The determination unit 13 also calculates a second defect signal from the value of interest and the second statistical value. Specifically, as shown in equation (4) below, the determination unit 13 calculates the second defect signal from the difference between the value of interest and the second statistical value (mean value). In equations (3) and (4), the first defect signal is denoted by KS_1, the second defect signal by KS_2, and the value of interest by CV.

[0054] KS_1 = CV - TV_1 (3) KS_2 = CV - TV_2 (4)

[0055] (Process C) The determination unit 13 converts the value of interest in the central region of the difference image G30 based on the conditions satisfied by the first and second defect signals. For example, the determination unit 13 converts the brightness (value of interest) of the pixel of interest CP0 to converted brightness based on the conditions satisfied by the first and second defect signals. Specifically, the determination unit 13 converts the parameter (brightness) of the pixel of interest CP0 to the defect signal sum when both condition I and condition II are satisfied. Condition II uses equations (5) and (6). The defect signal sum is the sum of the first defect signal and the second defect signal, as calculated by equation (7). Equation (7) shows the defect signal sum in terms of KSW.

[0056] The determination unit 13 replaces the parameter (luminance) of the pixel of interest CP0 in the central region of the difference image G30 with a specific value if at least one of conditions I and II is not met. The specific value may be a predetermined luminance, for example, 0.

[0057] [Condition I] The sign of the first defect signal (KS_1) is equal to the sign of the second defect signal (KS_2).

[0058] [Condition II] Both the first and second defect signals must be of roughly equal magnitude. That is, as shown in equation (5), the absolute value of the first defect signal is greater than the absolute value of the second defect signal divided by a number L greater than 1, and as shown in equation (6), the absolute value of the second defect signal is greater than the absolute value of the first defect signal divided by a number L greater than 1. In equations (5) and (6), L is a number greater than 1. For example, L is 4. Also, |KS_1| and |KS_2| represent the absolute values ​​of the first defect signal (KS_1) and the second defect signal (KS_2), respectively.

[0059] |KS_1|>(|KS_2|÷L) (5) |KS_2|>(|KS_1|÷L) (6)

[0060] KSW = KS_1 + KS_2 (7)

[0061] The determination unit 13 applies kernel W to each pixel of the continuous region 53 in the difference image G30 and outputs the pixel value after processing. The determination unit 13 sweeps the entire continuous region 53 with kernel W by sequentially changing the pixel of the difference image G13 at the position corresponding to the cell of interest CC0 of kernel W (the pixel of interest CP0). As a result, the determination unit 13 converts the pixel parameters of the difference image G30 to conversion parameters (conversion luminance) for the entire continuous region 53. More specifically, the determination unit 13 converts the pixel parameters of the continuous region 53 in the difference image G30 to the defect signal sum KSW or a specific value (for example, 0).

[0062] In the above explanation, the determination unit 13 converts the parameter (luminance) of the pixel CP0 of interest into a defect signal sum when both conditions I and II are met, and replaces the parameter (luminance) of the pixel CP0 of interest with a specific value when at least one of conditions I and II is not met. However, it is not limited to this. That is, the determination unit 13 may convert the parameter (luminance) of the pixel CP0 of interest into a defect signal sum when condition I is met, and replace the parameter (luminance) of the pixel CP0 of interest with a specific value when condition I is not met. Alternatively, the determination unit 13 may convert the parameter (luminance) of the pixel CP0 of interest into a defect signal sum when condition II is met, and replace the parameter (luminance) of the pixel CP0 of interest with a specific value when condition II is not met. Alternatively, the determination unit 13 may convert the parameter (luminance) of the pixel CP0 of interest into a defect signal sum when at least one of conditions I and II is met, and replace the parameter (luminance) of the pixel CP0 of interest with a specific value when neither condition I nor condition II is met.

[0063] The determination unit 13 may apply kernel W to the difference image G30 to convert the difference image G30 into a converted image. The converted image includes converted parameters (converted brightness) for each pixel, which are obtained by converting the parameters (brightness) of each pixel in the continuous region 53 of the difference image G30. Figure 6 is a diagram illustrating the converted image G40 of the sample 50 in the inspection apparatus 10 according to Embodiment 1. As shown in Figure 6, the converted image G40 is an image in which, similar to the difference image G30, for example, M pixels are arranged in the first direction and for example, N pixels are arranged in the second direction. Each pixel (p') of the converted image G40 m、n ) parameter I' m、n By applying kernel W, each pixel (p) of the difference image G30 is processed. m、n ) parameter I m、n It is changed from the defective signal sum KSW to a specific value (e.g., 0).

[0064] The determination unit 13 determines defects based on the conversion parameters. For example, the determination unit 13 determines that a location in the converted image G40 where the conversion parameter (conversion brightness) of a pixel is within a predetermined range (e.g., above a threshold) is a defect. In this way, the determination unit 13 determines defects in a predetermined processing area within the area corresponding to the continuous area 53 based on parameters based on the difference image G30 of the central area, parameters based on the difference image G30 of the first peripheral area, and parameters based on the difference image G30 of the second peripheral area.

[0065] Based on the above explanation, it can be said that the determination unit 13 determines defects in the central region based on a first defect signal (KS_1) obtained by correcting the pixel parameters of the difference image G30 in the central region based on a first statistical value of the pixel parameters of the difference image G30 in the first peripheral region, and a second defect signal (KS_2) obtained by correcting the pixel parameters of the difference image G30 in the central region based on a second statistical value of the pixel parameters of the difference image G30 in the second peripheral region. Furthermore, it can be said that the determination unit 13 determines defects in the central region based on conditions relating to the first defect signal (KS_1) and the second defect signal (KS_2). Specifically, the determination unit 13 determines defects in the central region based on converted parameters obtained by converting the pixel parameters of the difference image G30 in the central region based on conditions relating to the first defect signal (KS_1) and the second defect signal (KS_2). Note that the conditions relating to the 1-defect signal (KS_1) and the 2-defect signal (KS_1) may include at least one of the above conditions I and II.

[0066] (Statistical values) As mentioned above, the first and second statistical values ​​may be the average values ​​of the parameters (luminance) of the first peripheral pixel group (pixels 2 to 5) and the second peripheral pixel group (pixels 6 to 9), respectively. However, when the statistical values ​​are average values, depending on the variation in parameters (luminance), it may not be possible to adequately represent the parameter (luminance) profiles of the first peripheral region and the second peripheral region. For example, if there are singular points that are abnormally high or low in luminance compared to other parts, such as spike noise, the average value will be strongly influenced by the values ​​of such singular points. Therefore, the first and second defective signals will also be strongly influenced by the values ​​of such singular points.

[0067] Therefore, in this embodiment, the determination unit 13 may calculate the standard deviation in addition to the mean as a statistical value. Furthermore, when calculating the first defect signal and the second defect signal, the determination unit 13 may use the standard deviation as shown in equations (8) and (9).

[0068] KS_1=(CV-TV_1) / σ_1 (8) KS_2 = (CV - TV_2) / σ_2 (9)

[0069] Here, σ_1 and σ_2 represent the standard deviation of the parameters (luminance) of the first peripheral region pixel group and the standard deviation of the parameters (luminance) of the second peripheral region pixel group, respectively. As a result, the first and second defect signals appropriately reflect the luminance distribution of the first and second peripheral regions when singularities or other issues are present.

[0070] (Another kernel) Figure 7 is a diagram illustrating another kernel WABC used by the determination unit 13 in the inspection apparatus 10 according to Embodiment 1. As shown in Figure 7, kernel WABC may be a configuration in which multiple cells of kernel W are arranged in a direction orthogonal to the first direction. For example, as shown in Figure 7, kernel WABC may be a configuration in which three columns of kernel W are arranged in a direction orthogonal to the first direction. Kernel WABC includes subkernels WA, WB, and WC. Subkernels WA, WB, and WC are each equivalent to kernel W.

[0071] The determination unit 13 calculates the first statistical values ​​(TVA_1, TVB_1, and TVC_1) and the second statistical values ​​(TVA_2, TVB_2, and TVC_2) for each of the subkernels WA, WB, and WC in the kernel WABC, using equations (1) and (2).

[0072] The determination unit 13 calculates the first defect signals (KSA_1, KSB_1, and KSC_1) and the second defect signals (KSA_2, KSB_2, and KSC_2) in each of the subkernels WA, WB, and WC of the kernel WABC, using equations (3) and (4).

[0073] The determination unit 13 transforms the parameters of the target pixel CP0 (which may be the central pixel of the central subkernel (subkernel WB) when multiple subkernels are arranged in a direction orthogonal to the first direction) corresponding to the central part of kernel WABC in each of subkernels WA, WB, and WC in kernel WABC, based on [Condition I] and [Condition II]. For example, if both [Condition I] and [Condition II] are satisfied in each of subkernels WA, WB, and WC in kernel WABC, the determination unit 13 transforms the target pixel CP0 of the difference image G30 at the position corresponding to the target cell CC0 in the central part of kernel WB into a defect signal sum (KSWABC) as shown in equation (10) below.

[0074] KSWABC=KSA_1+KSA_2+KSB_1+KSB_2 +KSC_1+KSC_2+ (10)

[0075] The determination unit 13 converts the pixel CP0 of interest to a specific value (e.g., 0) if at least one of the subkernels WA, WB, and WC in kernel WABC does not satisfy at least one of [Condition I] and [Condition II].

[0076] (Information processing device) Figure 8 is a block diagram illustrating an information processing device IP according to Embodiment 1. As shown in Figure 8, the inspection device 10 may include an information processing device IP that includes, for example, a computer such as a microcomputer, server, or personal computer.

[0077] The information processing device IP may include a processor PRC, memory MMR, storage device STR, and a user interface UI. The storage device STR stores programs for the processes to be performed by each component of the inspection device 10. The processor PRC loads the programs from the storage device STR into memory MMR and executes them. In this way, the processor PRC realizes the functions of each component in the inspection device 10. The user interface UI may include input devices such as a keyboard, mouse, and imaging device, and output devices such as a display, printer, and speaker.

[0078] The information processing device IP comprises at least one memory MMR for storing instruction sets and at least one processor PRC. The processor PRC executes the instruction sets to identify a continuous region 53 in the sample 50 on which the pattern 51 is formed, where substantially equivalent predetermined properties are continuous in a first direction. The processor PRC also executes the instruction sets to obtain a difference image G30 based on a comparison between the captured image G10 and the reference image G20. Furthermore, the processor PRC executes the instruction sets to determine defects in a predetermined processing area within the continuous region 53 based on parameters based on the difference image G30 of the central region, parameters based on the difference image G30 of the first peripheral region, and parameters based on the difference image G30 of the second peripheral region. The processor PRC may be configured to execute such instruction sets.

[0079] Each component of the information processing device IP may be implemented with dedicated hardware. Furthermore, some or all of each component may be implemented by general-purpose or dedicated circuits, processors (PRCs), or combinations thereof. These may be implemented by a single chip or by multiple chips connected via a bus. Some or all of each component may be implemented by a combination of the aforementioned circuits and programs. Additionally, a CPU, GPU, FPGA (Field-programmable Gate Array), quantum processor (quantum computer control chip), etc., can be used as the processor (PRC).

[0080] Furthermore, if some or all of the components of an information processing device IP are implemented by multiple information processing device IPs or circuits, these multiple IPs or circuits may be centrally located or distributed. For example, the information processing device IPs or circuits may be implemented in a form where each is connected via a communication network, such as a client-server system or a cloud computing system. In addition, the functions of the information processing device IP may be provided in a SaaS (Software as a Service) format.

[0081] (Testing method) Next, the inspection method according to this embodiment will be described. Figure 9 is a flowchart illustrating the inspection method according to Embodiment 1. As shown in Figure 9, the inspection method of this embodiment comprises a step S11 for identification, a step S12 for acquisition, and a step S13 for determination. The acquisition step S12 may be performed before the identification step S11, or it may be performed in parallel with the identification step S11.

[0082] In step S11, the identification unit 11 identifies a continuous region 53 in the sample 50 on which the pattern 51 is formed, in which substantially equivalent predetermined properties are continuous in a first direction.

[0083] In step S12, the acquisition unit 12 acquires a difference image G30 based on a comparison between the captured image G10 of the sample 50 and the reference image G20.

[0084] In step S13, the determination unit 13 determines a defect in a predetermined processing area within the continuous region 53 of the difference image G30, based on parameters based on the difference image G30 of the central region, parameters based on the difference image G30 of the first peripheral region, and parameters based on the difference image G30 of the second peripheral region.

[0085] Specifically, in step S13, the determination unit 13 first generates a first defect signal from the parameter (luminance) of the pixel of interest CP0 and the first statistical value of the first peripheral region pixel group. The determination unit 13 also generates a second defect signal from the luminance of the pixel of interest CP0 and the second statistical value of the second peripheral region pixel group. Then, based on conditions I and II satisfied by the first and second defect signals, the determination unit 13 converts the parameter (luminance) of the pixel of interest CP0 into a conversion parameter (converted luminance). Based on this, the determination unit 13 determines whether a defect exists. The determination unit 13 may also determine whether a defect exists if the conversion parameter in the converted image G40 is within a predetermined range (for example, a value greater than or equal to a threshold).

[0086] Next, the effects of this embodiment will be explained. The inspection device 10 of this embodiment determines defects based on parameters based on the difference image G30 of the central region, parameters based on the difference image G30 of the first peripheral region, and parameters based on the difference image G30 of the second peripheral region. Therefore, it is possible to differentiate afterimages and the like that appear in the captured image G10 and the reference image G20, thereby improving the accuracy of defect detection. The inspection device 10 of this embodiment, under the configuration described above, can be said to determine defects not only from the image appearing in the difference image G30 (or instead), but also from the relationship between the candidate defect location in the difference image G30 and the surrounding region. In particular, the inspection device 10 places importance on the balance between the parameters on one side and the parameters on the other side of the candidate defect location in a continuous region 53 in which substantially equivalent predetermined properties are continuous in the first direction. That is, for defects in the continuous region 53, defects are determined based on conditions I and II that the parameters on one side and the parameters on the other side satisfy. This improves the accuracy of defect detection.

[0087] <Embodiment 2> Next, an inspection apparatus according to Embodiment 2 will be described. In the inspection apparatus of this embodiment, the identification unit 11 identifies a candidate defect region based on the difference image G30. The determination unit 13 processes the continuous region 53 containing the candidate defect region. The pixel parameter will be described as brightness, but the parameter may be something other than brightness.

[0088] Figure 10 is an example of the captured image G10 of the sample 50 in the inspection apparatus 10 according to Embodiment 2. Figure 11 is an example of the reference image G20 of the sample 50 in the inspection apparatus 10 according to Embodiment 2. Figure 12 is an example of the difference image G30 of the sample 50 in the inspection apparatus 10 according to Embodiment 2. Each figure shows the luminance profile KP in the continuous region 53.

[0089] As shown in Figure 10, the captured image G10 has a parameter distribution in the continuous region 53 in which the parameters change along the first direction. In the following, the parameter distribution will be described using a luminance profile KP in which the luminance changes along the first direction. For example, the luminance profile KP in the captured image G10 has a high-luminance portion that includes pixels with a luminance greater than a predetermined threshold. On the other hand, as shown in Figure 11, in the luminance profile KP of the reference image G10, the portion corresponding to the high-luminance portion in the captured image G10 has a flat luminance profile KP. Therefore, as shown in Figure 12, the luminance profile KP of the difference image G0 has a high-luminance portion in the portion corresponding to the high-luminance portion in the captured image G10.

[0090] The determination unit 13 acquires a luminance profile KP along a first direction in a continuous region 53 in the difference image G30. Then, the determination unit 13 identifies a candidate defect region 54 from the acquired luminance profile KP. For example, the determination unit 13 identifies a candidate defect region 54 in the difference image G30 that includes pixels with parameters (luminance) within a predetermined range (e.g., above a predetermined threshold). The candidate defect region 54 corresponds to a high-luminance portion in the captured image G10.

[0091] In this case, the determination unit 13 uses the region containing the defect candidate region 54 as the processing region. Specifically, the determination unit 13 may apply the kernel W in the processing region within the contiguous region 53 containing the defect candidate region 54.

[0092] According to this embodiment, the kernel W is applied only to the contiguous region 53 that includes the defect candidate region 54 identified based on the difference image G30, thus reducing processing time.

[0093] The determination unit 13 may perform processing other than applying kernel W in the processing area within the continuous region 53 that includes the defect candidate region 54. For example, the determination unit 13 identifies the defect candidate region 54 within the continuous region 53 from the brightness profile of the difference image G30. The determination unit 13 may then treat the defect candidate region 54 as the central region, the area to one side of the defect candidate region 54 in the first direction as the first peripheral region, and the area to the other side of the defect candidate region 54 in the first direction as the second peripheral region, and perform the above-mentioned processing (A), (B), and (C).

[0094] Figure 13 is a diagram illustrating the captured image G10 of the sample 50 in the inspection apparatus 10 according to Embodiment 2. Figure 14 is a diagram illustrating the reference image G20 of the sample 50 in the inspection apparatus 10 according to Embodiment 2. Figure 15 is a diagram illustrating the difference image G30 of the sample 50 in the inspection apparatus 10 according to Embodiment 2.

[0095] As shown in Figures 13 to 15, the sample 50 may have a plurality of patterns 51 extending in a second direction intersecting the first direction. The second direction may be perpendicular to the first direction. The plurality of patterns 51 may include portions arranged side by side in the second direction. For example, a pair of two patterns 51 arranged side by side in the second direction may have a gap 55 between the two patterns 51. The sample 50 may have multiple such pairs arranged at intervals in the first direction. If there are multiple pairs, the multiple gaps 55 are arranged at intervals in the first direction. The portion of the sample 50 without a pattern 51 is the space between patterns 52. Therefore, the space between patterns 52 includes the gaps 55. Thus, the space between patterns 52 that includes the region in which multiple gaps 55 are arranged at intervals in the first direction constitutes a continuous region 53. The sample 50 may contain defects in the gaps 55.

[0096] Note that pattern 51 of sample 50 may be replaced by the space between patterns 52. In that case, the space between patterns 52 is replaced by pattern 51. Therefore, pattern 51 includes gaps 55. A pattern 51 that includes a region in which multiple gaps 55 are arranged side by side with spacing in the first direction becomes a continuous region 53. Sample 50 may contain defects in the gaps 55.

[0097] Thus, the sample 50 may have one or more pairs of two patterns 51 extending in a second direction intersecting the first direction, or pairs of two patterns 52 extending in a second direction intersecting the first direction, with a gap 55 in between, adjacent in the second direction. If the sample 50 has multiple pairs, the multiple pairs may be spaced apart in the first direction. The continuous region 53 may include one or more gaps 55 aligned in the first direction.

[0098] The pixels in the gap portion 55 in the captured image G10 and the reference image G20 are different from the pixels in other parts. High brightness can occur. For example, due to the effect of afterimages during imaging, the brightness of pixels in the gap area 55 may become high. Therefore, if the high-brightness area in the captured image G10 is judged as a defect, such afterimages, which are not actually defects, will be mistakenly identified as defects.

[0099] Furthermore, the difference image G30, obtained by subtracting the captured image G10 from the reference image G20, has difficulty completely canceling out the effects of afterimages. For example, afterimages may remain due to differences in conditions such as light intensity, field of view, and magnification when capturing the captured image G10 and when capturing the reference image G20.

[0100] As described above, the inspection device 10 of this embodiment determines defects based on parameters based on the difference image G30 of the central region, parameters based on the difference image G30 of the first peripheral region, and parameters based on the difference image G30 of the second peripheral region. When a defect exists in the continuous region 53, the inspection device 10 determines the defect based on conditions I and II that are satisfied by the parameters on one side and the parameters on the other side. This improves the accuracy of defect detection.

[0101] Figure 16 is a diagram illustrating the captured image G10 of the sample 50 in the inspection apparatus 10 according to Embodiment 2. Figure 17 is a diagram illustrating the reference image G20 of the sample 50 in the inspection apparatus 10 according to Embodiment 2. Figure 18 is a diagram illustrating the difference image G30 of the sample 50 in the inspection apparatus 10 according to Embodiment 2.

[0102] As shown in Figures 16 to 18, the sample 50 may have a plurality of circular patterns 56 arranged in a second direction intersecting the first direction with a gap 55 in between. Here, "circular" includes not only strictly circular shapes but also circular shapes within a range that includes manufacturing tolerances for producing the pattern 51. Furthermore, "circular" also includes rectangles with no difference between the length in the first direction and the length in the second direction, and rounded rectangles in which the corners of the rectangle are rounded. Shapes with a difference between the length in the first direction and the length in the second direction are included in the aforementioned pattern 51 or the space between patterns 52.

[0103] The sample 50 may have one or more sets of multiple circular patterns 56 arranged in a second direction with gaps 55 in between. If the sample 50 has multiple sets, the multiple sets may be arranged in a first direction with gaps between them. If there are multiple sets, the multiple gaps 55 are arranged in a first direction with gaps between them. The parts of the sample 50 that do not have patterns 56 are the gaps between patterns 52. Therefore, the gaps between patterns 52 include the gaps 55. Thus, the gaps between patterns 52 that include the region in which multiple gaps 55 are arranged in a first direction with gaps between them becomes a continuous region 53. The sample 50 may contain defects in the gaps 55.

[0104] Note that the pattern 56 of sample 50 may be replaced by the space between patterns 52. In that case, the space between patterns 52 is replaced by pattern 56. Therefore, pattern 56 includes gaps 55. A pattern 56 that includes a region in which multiple gaps 55 are arranged side by side with spacing in the first direction becomes a continuous region 53. Sample 50 may contain defects in the gaps 55.

[0105] In the cases shown in Figures 16 to 18, as in Figures 13 to 15, the afterimage of the gap portion 55 becomes a problem. However, the inspection device 10 of this embodiment determines the defect based on conditions I and II, which are satisfied by the parameters of one side and the parameters of the other side. This makes it possible to improve the accuracy of defect detection.

[0106] <Embodiment 3> Next, an inspection apparatus according to Embodiment 3 will be described. The inspection apparatus of this embodiment does not require the use of a kernel W and performs defect determination. For example, the determination unit 13 may determine a defect in the central region based on a first defect signal obtained by correcting the pixel parameters of the difference image G30 of the central region based on a first statistical value of the pixel parameters of the difference image G30 of the first peripheral region, and a second defect signal obtained by correcting the pixel parameters of the difference image G30 of the central region based on a second statistical value of the pixel parameters of the difference image G30 of the second peripheral region. Here, the first defect signal obtained by correcting the pixel parameters of the difference image G30 of the central region based on a first statistical value of the pixel parameters of the difference image G30 of the first peripheral region is not limited to, for example, the first defect signal obtained by differentiating the pixel parameters of the difference image G30 of the central region by the first statistical value of the pixel parameters of the difference image G30 of the first peripheral region, as described above, but also includes a first defect signal obtained by performing processing other than difference, such as taking a ratio. The same applies to the second defect signal.

[0107] The determination unit 13 may determine a defect in the central region based on a first condition, which includes a predetermined relationship between the codes of the first defect signal and the second defect signal. Alternatively, the determination unit 13 may determine a defect in the central region based on a second condition, which includes that both the first defect signal and the second defect signal are of similarly large values. The determination unit 13 may determine a defect in the central region when both the first condition, which includes a predetermined relationship between the codes of the first defect signal and the second defect signal, and the second condition, which includes that both the first defect signal and the second defect signal are of similarly large values, are satisfied.

[0108] Furthermore, an example of identifying defect candidates and determining the truth value of defect candidates without requiring the use of kernel W will be explained with reference to the figure. Figure 19 is a block diagram illustrating an inspection device 10a according to Embodiment 3. As shown in Figure 19, the inspection device 10a includes a candidate identification unit 14 and a generation unit 15 in addition to the identification unit 11, acquisition unit 12 and determination unit 13. The functions of the identification unit 11 and acquisition unit 12 are the same as those of the identification unit 11 and acquisition unit 12 in the inspection device 10 described above. Note that the determination unit 13 may also have the functions of the candidate identification unit 14 and the generation unit 15.

[0109] The candidate identification unit 14 identifies defect candidates on the continuous region 53 based on the difference image G30. The candidate identification unit 14 may also identify defect candidates as the aforementioned defect candidate region. For example, the candidate identification unit 14 identifies defect candidates that include pixels in the difference image G30 having parameters above a predetermined threshold. Defect candidates also correspond to high-brightness portions in the captured image G10.

[0110] The generation unit 15 generates a parameter distribution of the difference image G30 along the first direction for at least a portion of the continuous region 53 to which the defect candidate belongs. The parameter distribution may include the luminance profile KP described above. The generation unit 15 may also generate a parameter distribution for a peripheral region in the continuous region 53 that is different from the defect candidate, based on the statistical values ​​of the parameters of the difference image G30 for that peripheral region.

[0111] The peripheral region may include the first peripheral region and the second peripheral region described above. The statistical value may include the mean value. The statistical value may include at least one of the mean value and the standard deviation. The generation unit 15 may generate a parameter distribution for peripheral regions different from defect candidates in the continuous region 53 based on the statistical values ​​of the parameters of the difference image G30 of the peripheral region by performing the processing of equations (1) and (2) described above.

[0112] The generation unit 15 may generate a parameter distribution for a defect candidate region including a defect candidate based on the difference between the parameters of the difference image G30 of the defect candidate region and statistical values. For example, the parameter distribution for a defect candidate region including a defect candidate may include the first defect signal and the second defect signal described above. The generation unit 15 may generate a parameter distribution for a defect candidate region including a defect candidate based on the difference between the parameters of the difference image G30 of the defect candidate region and statistical values ​​by performing the processing of equations (3) and (4) described above.

[0113] The determination unit 13 determines the truth value of a defect candidate based on the parameter distribution. Specifically, the determination unit 13 determines whether there is a peak value that satisfies predetermined conditions based on the parameter distribution of the generated defect candidate region. Here, the predetermined conditions may include the condition that the difference between the parameters of the difference image G30 in the defect candidate region and the statistical value has a predetermined magnitude. The predetermined conditions may also include the condition that the first difference value, which is the difference between the parameters of the difference image G30 in the defect candidate region and the statistical value of the parameters of the difference image G30 for a first peripheral region that is on one side in the first direction relative to the defect candidate among the peripheral regions in the continuous region 53 that are different from the defect candidate, matches to a predetermined degree the second difference value, which is the difference between the parameters of the difference image G30 in the defect candidate region and the statistical value of the parameters of the difference image for a second peripheral region that is on the other side in the first direction relative to the defect candidate among the peripheral regions in the continuous region 53 that are different from the defect candidate. The first difference value and the second difference value may correspond to the first defect signal and the second defect signal described above. In other words, the determination unit 13 may determine the truth value of a defect candidate based on conditions I and II, using the first difference value and the second difference value as the aforementioned first defect signal and second defect signal. Furthermore, the predetermined conditions may further include the condition that the transformed parameters obtained by transforming the pixel parameters of the defect candidate (i.e., the difference image G30 of the central region) under the aforementioned equation (10) are within a predetermined range (for example, above a threshold).

[0114] Figure 20 is a flowchart illustrating an inspection method according to Embodiment 3. As shown in Figure 20, the inspection method of this embodiment includes, in addition to the identification step S21, acquisition step S22, and determination step S25, a step S23 for identifying defect candidates and a step S24 for generating a parameter distribution. The identification step S21 and acquisition step S22 are the same as the identification step S11 and acquisition step S12 described above.

[0115] In step S23, the candidate identification unit 14 identifies defect candidates on the continuous region 53 based on the difference image G30. For example, the candidate identification unit 14 identifies defect candidates that include pixels in the difference image G30 having parameters greater than or equal to a predetermined threshold.

[0116] In step S24, the generation unit 15 generates a parameter distribution of the difference image G30 along the first direction for at least a portion of the continuous region 53 to which the defect candidate belongs.

[0117] In step S25, the determination unit 13 determines the truth value of a defect candidate based on the parameter distribution. The determination unit 13 determines whether or not there is a peak value that satisfies predetermined conditions based on the parameter distribution. In this way, the inspection device 10a of this embodiment can determine defects without using a kernel W.

[0118] <Embodiment 4> Next, an optical apparatus according to Embodiment 4 will be described. The optical apparatus of this embodiment includes an inspection apparatus 10 or an inspection apparatus 10a. The inspection apparatus 10 and the inspection apparatus 10a are collectively referred to as the inspection apparatus 10. The optical apparatus includes, for example, an apparatus for inspecting a sample 50 using optical methods and the inspection apparatus 10. The optical apparatus may also include a review apparatus for reviewing the sample 50 and the inspection apparatus 10. In the following description, as an example of an optical apparatus, an apparatus including an apparatus for inspecting a sample 50 using optical methods and the inspection apparatus 10 will be referred to as the inspection apparatus 100.

[0119] Figure 21 is a diagram illustrating an inspection apparatus 100 according to Embodiment 4. The inspection apparatus 100 comprises an illumination optical system 110, an imaging optical system 120, and an inspection apparatus 10. The illumination optical system 110 illuminates the sample 50 using illumination light L111. The sample 50 is, for example, a mask having a pattern 51. Note that the sample 50 is not limited to a mask, but may also be a semiconductor substrate or a semiconductor chip, as long as it has a pattern 51. The illumination optical system 110 includes, for example, a light source 111, an ellipsoidal mirror 112, an ellipsoidal mirror 113, and a recessed mirror 114.

[0120] The imaging optical system 120 captures an image G10 of the sample 50 illuminated by illumination light L111. The imaging optical system 120 includes, for example, a perforated concave mirror 121, a convex mirror 122, and a detector 123. The perforated concave mirror 121 and the convex mirror 122 constitute a Schwarzschild magnifying optical system. The illumination optical system 110 and the imaging optical system 120 may further include optical components other than those described above, or any of the above optical components may be omitted.

[0121] Here, for the sake of convenience in explaining the inspection device 100, we introduce the XYZ Cartesian coordinate system. For example, the plane parallel to the stage surface of the stage 152 on which the sample 50 is placed is defined as the XY plane, and the direction perpendicular to the stage surface is defined as the Z axis direction. For convenience, the +Z axis direction is referred to as "up," and the -Z axis direction is referred to as "down." Note that "up" and "down" are for the sake of convenience in explaining the inspection device 100 and do not indicate the actual direction in which the inspection device 100 is positioned.

[0122] The light source 111 generates illumination light L111. The illumination light L111 contains, for example, EUV light of 13.5 nm, the same wavelength as the exposure wavelength of the EUV mask, which is sample 50. The illumination light L111 generated from the light source 111 is reflected by the ellipsoidal mirror 112. The illumination light L111 reflected by the ellipsoidal mirror 112 travels while being focused and is concentrated at the focal point IF1. The focal point IF1 is located in a position conjugate to the pattern surface 59 of sample 50.

[0123] The illumination light L111, after passing through the focal point IF1, spreads out as it travels and enters a reflecting mirror such as the ellipsoidal mirror 113. The illumination light L111 that enters the ellipsoidal mirror 113 is reflected by the ellipsoidal mirror 113, and as it travels, it is focused and enters the recessed mirror 114. In other words, the ellipsoidal mirror 113 causes the illumination light L111 to enter the recessed mirror 114 as a focused beam. The recessed mirror 114 is positioned above the sample 50. The illumination light L111 that enters the recessed mirror 114 and is reflected enters the sample 50. That is, the recessed mirror 114 causes the illumination light L111 to enter the sample 50.

[0124] The ellipsoidal mirror 113 focuses the illumination light L111 onto the sample 50. The illumination optical system 110 is positioned so that when the illumination light L111 illuminates the sample 50, the image of the light source 111 is projected onto the pattern surface 59 of the sample 50. Therefore, the illumination optical system 110 provides critical illumination. In this way, the illumination optical system 110 illuminates the sample 50 using critical illumination provided by the illumination light L111 generated by the light source 111.

[0125] The sample 50 is placed on the stage 152. The illumination light L111 is incident on the sample 50 from a direction inclined from the Z-axis direction. That is, the illumination light L111 is incident on the sample 50 at an oblique angle as oblique incidence illumination. The illumination light L111 may illuminate the sample 50 as such oblique incidence illumination.

[0126] Stage 152 is an XYZ driven stage. By moving Stage 152 in the X-axis and Y-axis directions, a desired area of ​​the sample 50 can be illuminated. Furthermore, by moving Stage 152 in the Z-axis direction, focus adjustment can be performed. Stage 152 may also be rotated using the X-axis, Y-axis, and Z-axis as rotation axes. Alternatively, instead of moving and rotating Stage 152 along the X-axis, Y-axis, and Z-axis directions, the illumination optical system 110 and the imaging optical system 120 may be moved and rotated.

[0127] Illumination light L111 from light source 111 illuminates the imaging area of ​​sample 50. Reflected light L112, which is incident from a direction inclined with respect to the Z axis and reflected by sample 50, is incident on perforated concave mirror 121. A hole 121a is provided in the center of the perforated concave mirror 121.

[0128] The reflected light L112 reflected by the perforated concave mirror 121 is incident on the convex mirror 122. The convex mirror 122 reflects the reflected light L112 incident from the perforated concave mirror 121 toward the hole 121a of the perforated concave mirror 121. The reflected light L112 that has passed through the hole 121a is detected by the detector 123. The detector 123 may be a detector 123 that includes a TDI (Time Delay Integration) sensor.

[0129] The detector 123 acquires image data of the sample 50. The detector 123 includes multiple image sensors arranged in a line in one direction. Linear image data captured by the multiple image sensors arranged in a line is called one-dimensional image data, or one frame. The detector 123 acquires multiple one-dimensional image data by scanning in a direction orthogonal to one direction. The image sensors are, for example, CCDs (Charge Coupled Devices). However, the image sensors are not limited to CCDs.

[0130] In this manner, the imaging optical system 120 collects the reflected light L112 from the sample 50 illuminated by the illumination light L111, and detects the collected reflected light L112 with the detector 123 to acquire image data of the sample 50. The image data is, for example, two-dimensional image data.

[0131] The reflected light L112 contains information such as defect images of the sample 50. The specularly reflected light of the illumination light L111 incident on the sample 50 from a direction tilted with respect to the Z axis is detected by the imaging optical system 120. If a defect exists in the sample 50, the defect image is observed as a dark image. This observation method is called bright-field observation. Multiple one-dimensional image data of the sample 50 acquired by the detector 123 are output to the inspection device 10 and processed into two-dimensional image data.

[0132] The inspection device 10 is connected to the imaging optical system 120 by a signal line or wirelessly. The inspection device 10 receives image data of the sample 50 from the detector 123 in the imaging optical system 120. The inspection device 10 processes the image data of the sample 50 received from the detector 123 as a two-dimensional image G10. Other functions and configurations of the inspection device 10 are as described in the previous embodiments.

[0133] Sample 50 is, for example, an EUV mask that corresponds to EUV light. However, sample 50 is not limited to an EUV mask. Sample 50 may also be a photomask that corresponds to illumination light L111 having a different wavelength. Furthermore, as mentioned above, sample 50 may also be a semiconductor substrate, a semiconductor chip, or a semiconductor device.

[0134] While embodiments of this disclosure have been described above, this disclosure includes appropriate modifications that do not impair its purpose and advantages, and is not limited by the embodiments described above. Furthermore, the components in the embodiments may be combined as appropriate. The inspection apparatus and inspection program described below also fall within the scope of the technical concept of the embodiments.

[0135] (Note A1) A identifying unit that identifies a continuous region in a patterned sample in which substantially equivalent predetermined properties are continuous in a first direction, An acquisition unit that acquires a difference image based on a comparison between the captured image of the sample and a reference image, A candidate identification unit identifies candidate defects on a continuous region based on the aforementioned difference image, A generation unit generates a parameter distribution of the difference image along the first direction for at least a portion of the continuous region to which the defect candidate belongs, A determination unit that determines the truth or falsity of the defect candidate based on the parameter distribution, An inspection device equipped with the following features. (Appendix A2) The generating unit is The parameter distribution for surrounding regions that are different from the defect candidate in the continuous region is generated based on the statistical values ​​of the parameters of the difference image of the said surrounding region. The inspection device described in Appendix A1. (Note A3) The generating unit is The parameter distribution for the defect candidate region including the aforementioned defect candidate is, The parameters of the difference image of the defect candidate region are generated based on the difference between the parameters and the statistical values. The inspection device described in Appendix A2. (Note A4) The determination unit, Based on the parameter distribution, it is determined whether or not there is a peak value that satisfies a predetermined condition. The aforementioned predetermined conditions are: The condition includes that the difference between the parameters of the difference image in the defect candidate region and the statistical value has a predetermined magnitude. The inspection device described in Appendix A3. (Note A5) The determination unit, Based on the parameter distribution, it is determined whether or not there is a peak value that satisfies a predetermined condition. The aforementioned predetermined conditions are: The first difference value is the difference between the parameters of the difference image in the defect candidate region and the statistical value of the parameters of the difference image for a first peripheral region that is on one side in the first direction relative to the defect candidate, among the peripheral regions in the continuous region that are different from the defect candidate. The condition includes the following: the parameters of the difference image in the defect candidate region and the second difference value, which is the difference between the parameters of the difference image in the second peripheral region, which is on the other side in the first direction relative to the defect candidate, among the peripheral regions in the continuous region that are different from the defect candidate, must match to a predetermined degree. The inspection device described in Appendix A3.

[0136] (Note B1) At least one memory to store the instruction set, Execute the aforementioned set of instructions, In a sample on which a pattern has been formed, identify a continuous region in which substantially equivalent predetermined properties are continuous in a first direction. A difference image is obtained based on a comparison between the captured image of the sample and a reference image. In a predetermined processing region within a continuous region of the difference image, a defect is determined based on parameters based on the difference image of the central region, parameters based on the difference image of a first peripheral region on one side of the central region in the first direction, and parameters based on the difference image of a second peripheral region on the other side of the central region in the first direction. A processor configured as follows: Having, Inspection device.

[0137] (Note C1) A step of identifying a continuous region in a patterned sample where substantially equivalent predetermined properties are continuous in a first direction, The steps include obtaining a difference image based on a comparison between the captured image of the sample and a reference image, A step of determining a defect in a predetermined processing area within a continuous region of the difference image, based on parameters based on the difference image of the central region, parameters based on the difference image of a first peripheral region on one side of the central region in the first direction, and parameters based on the difference image of a second peripheral region on the other side of the central region in the first direction. A test program that causes a computer to execute a command.

[0138] The test program described above, when loaded into a computer, includes a set of instructions (or software code) for causing the computer to perform one or more of the functions described in the embodiments. The program may be stored on a non-temporary computer-readable medium or a physical storage medium. Examples, but not limited to, include random-access memory (RAM), read-only memory (ROM), flash memory, solid-state drive (SSD) or other memory technologies, CD-ROM, digital versatile disc (DVD), Blu-ray® disc or other optical disc storage, magnetic cassette, magnetic tape, magnetic disk storage or other magnetic storage devices. The program may be transmitted over a temporary computer-readable medium or a communication medium. Examples, but not limited to, include temporary computer-readable medium or a communication medium that includes electrically, optically, acoustically or otherwise propagating signals. [Explanation of Symbols]

[0139] 10, 10a Inspection device 11 Specific section 12 Acquisition Department 13 Judgment section 14 Candidate identification part 15 Generation part 50 samples 51 patterns 52 patterns 53 Continuous Regions 54 Candidate defect regions 55 Gap 56 patterns 59 Patterned surfaces 100 Inspection device 110 Illumination optical system 111 Light source 112 Ellipsoidal mirror 113 Ellipsoidal mirror 114 Recessed Mirror 120 Imaging Optical System 121 Perforated concave mirror 121a Hole 122 Convex mirror 123 Detector 152 stages CC0 Featured Cell CP0 Featured Pixel G10 captured image G20 reference image G30 difference image IP Information Processing Device KP Brightness Profile KS_1 First Defect Signal KS_2 Second Defect Signal KSW Defective Signal Sum MMR memory PRC Processor STR storage TV_1 First Statistical Value TV_2 Second Statistical Value UI User Interface W, WA, WB, WC, WABC kernels W0 central part W1 First Peripheral Area W2 Second Peripheral Area

Claims

1. A identifying unit that identifies a continuous region in a patterned sample in which substantially equivalent predetermined properties are continuous in a first direction, An acquisition unit that acquires a difference image based on a comparison between the captured image of the sample and a reference image, A determination unit determines a defect in a predetermined processing area within a region corresponding to the continuous region in the difference image, based on parameters from the difference image of the central region, parameters from the difference image of a first peripheral region on one side of the central region in the first direction, and parameters from the difference image of a second peripheral region on the other side of the central region in the first direction. An inspection device equipped with the following features.

2. The determination unit, A candidate defect region is identified in the difference image, which includes pixels having parameters within a predetermined range. The region including the aforementioned candidate defect region is defined as the processing region. The inspection apparatus according to claim 1.

3. The determination unit, By applying a transformation operator that transforms the parameters of each pixel in the difference image to the processing area, the parameters are converted into transformation parameters. The defect is determined based on the converted conversion parameters. The inspection apparatus according to claim 1.

4. The conversion operator is a kernel having a central part containing one or more cells, a first peripheral part containing two or more cells on one side of the central part, and a second peripheral part containing two or more cells on the other side of the central part. The determination unit, A first defect signal is calculated from the parameters of the pixel of interest, which is a pixel of the difference image at a position corresponding to the central cell, and the first statistical value of the parameters of the first peripheral region pixel group, which consists of pixels of the difference image at a position corresponding to the first peripheral cell. A second defect signal is calculated from the parameters of the pixel of interest and the second statistical value of the parameters of the second peripheral region pixel group, which consists of pixels of the difference image at positions corresponding to the cells of the second peripheral region. Based on the conditions satisfied by the first defect signal and the second defect signal, the parameters of the pixel of interest are converted to the conversion parameters. The inspection apparatus according to claim 3.

5. The first and second statistical values ​​include at least one of the mean and the standard deviation. The inspection apparatus according to claim 4.

6. The aforementioned conditions are, The first condition is that the sign of the first defect signal and the sign of the second defect signal are equal, The second condition is that the absolute value of the first defect signal is greater than the value obtained by dividing the absolute value of the second defect signal by a number greater than 1, and the absolute value of the second defect signal is greater than the value obtained by dividing the absolute value of the first defect signal by a number greater than 1. including, The inspection apparatus according to claim 4.

7. The determination unit, If both the first and second conditions are met, the parameters of the pixel of interest are converted into a defect signal sum, which is the sum of the first defect signal and the second defect signal. If at least one of the first and second conditions is not met, the parameter of the pixel of interest is converted to a specific value. The inspection apparatus according to claim 6.

8. The determination unit, A first defect signal obtained by correcting the pixel parameters of the difference image in the central region based on a first statistical value of the pixel parameters of the difference image in the first peripheral region, Based on the second defect signal obtained by correcting the pixel parameters of the difference image in the central region based on the second statistical value of the pixel parameters of the difference image in the second peripheral region, Determining the defect in the central region, The inspection apparatus according to claim 1.

9. The determination unit determines the defect in the central region based on a first condition which includes that the codes of the first defect signal and the codes of the second defect signal have a predetermined relationship. The inspection apparatus according to claim 8.

10. The determination unit determines the defect in the central region based on a second condition, which includes the fact that both the first defect signal and the second defect signal are of similarly large values. The inspection apparatus according to claim 8.

11. The determination unit determines the defect in the central region when both of the following conditions are met: a first condition, which includes that the codes of the first defect signal and the second defect signal have a predetermined relationship; and a second condition, which includes that both the first defect signal and the second defect signal have values ​​of similar magnitude. The inspection apparatus according to claim 8.

12. The sample has multiple sets of two patterns extending in a second direction intersecting the first direction, or sets of two patterns extending in a second direction intersecting the first direction where at least one of them is adjacent in the second direction with a gap in between, spaced apart in the first direction. The continuous region includes a plurality of gap portions arranged in the first direction, The inspection apparatus according to claim 1.

13. The sample has multiple sets of circular patterns arranged in a second direction intersecting the first direction, with gaps in between, spaced apart in the first direction. The continuous region includes a plurality of gap portions arranged in the first direction, The inspection apparatus according to claim 1.

14. The sample includes the defect in the gap portion. The inspection apparatus according to claim 12 or 13.

15. A step of identifying a continuous region in a patterned sample where substantially equivalent predetermined properties are continuous in a first direction, The steps include obtaining a difference image based on a comparison between the captured image of the sample and a reference image, A step of determining a defect in a predetermined processing area within a region corresponding to the continuous region in the difference image, based on parameters based on the difference image of the central region, parameters based on the difference image of a first peripheral region on one side of the central region in the first direction, and parameters based on the difference image of a second peripheral region on the other side of the central region in the first direction. A testing method equipped with [a specific feature / feature].

16. In the determination step described above, A candidate defect region is identified in the difference image, which includes pixels having parameters within a predetermined range. The region including the aforementioned candidate defect region is defined as the processing region. The inspection method according to claim 15.

17. In the determination step described above, By applying a transformation operator that transforms the parameters of each pixel in the difference image to the processing area, the parameters are converted into transformation parameters. The defect is determined based on the converted conversion parameters. The inspection method according to claim 15.

18. The conversion operator is a kernel having a central part containing one or more cells, a first peripheral part containing two or more cells on one side of the central part, and a second peripheral part containing two or more cells on the other side of the central part. In the determination step described above, A first defect signal is calculated from the parameters of the pixel of interest, which is a pixel of the difference image at a position corresponding to the central cell, and the first statistical value of the parameters of the first peripheral region pixel group, which consists of pixels of the difference image at a position corresponding to the first peripheral cell. A second defect signal is calculated from the parameters of the pixel of interest and the second statistical value of the parameters of the second peripheral region pixel group, which consists of pixels of the difference image at positions corresponding to the cells of the second peripheral region. Based on the conditions satisfied by the first defect signal and the second defect signal, the parameters of the pixel of interest are converted to the conversion parameters. The inspection method according to claim 17.

19. The first and second statistical values ​​include at least one of the mean and the standard deviation. The inspection method according to claim 18.

20. In the determination step described above, The aforementioned conditions are, The first condition is that the sign of the first defect signal and the sign of the second defect signal are equal, The second condition is that the absolute value of the first defect signal is greater than the value obtained by dividing the absolute value of the second defect signal by a number greater than 1, and the absolute value of the second defect signal is greater than the value obtained by dividing the absolute value of the first defect signal by a number greater than 1. including, The inspection method according to claim 18.

21. In the determination step described above, If both the first and second conditions are met, the parameters of the pixel of interest are converted into a defect signal sum, which is the sum of the first defect signal and the second defect signal. If at least one of the first and second conditions is not met, the parameter of the pixel of interest is converted to a specific value. The inspection method according to claim 20.

22. In the determination step described above, A first defect signal obtained by correcting the pixel parameters of the difference image in the central region based on a first statistical value of the pixel parameters of the difference image in the first peripheral region, Based on the second defect signal obtained by correcting the pixel parameters of the difference image in the central region based on the second statistical value of the pixel parameters of the difference image in the second peripheral region, Determining the defect in the central region, The inspection method according to claim 15.

23. In the determination step described above, Based on a first condition including that the codes of the first defect signal and the second defect signal have a predetermined relationship, the defect in the central region is determined. The inspection method according to claim 22.

24. In the determination step described above, Based on a second condition, which includes that both the first defect signal and the second defect signal are of similarly large value, the defect in the central region is determined. The inspection method according to claim 22.

25. In the determination step described above, The defect in the central region is determined when both of the following conditions are met: a first condition, which includes that the sign of the first defect signal and the sign of the second defect signal have a predetermined relationship; and a second condition, which includes that both the first defect signal and the second defect signal have values ​​of similar magnitude. The inspection method according to claim 22.

26. The sample has multiple sets of two patterns extending in a second direction intersecting the first direction, or sets of two patterns extending in a second direction intersecting the first direction where at least one of them is adjacent in the second direction with a gap in between, spaced apart in the first direction. The continuous region includes a plurality of gap portions arranged in the first direction, The inspection method according to claim 15.

27. The sample has multiple sets of circular patterns arranged in a second direction intersecting the first direction, with gaps in between, spaced apart in the first direction. The continuous region includes a plurality of gap portions arranged in the first direction, The inspection method according to claim 15.

28. The sample includes the defect in the gap portion. The inspection method according to claim 26 or 27.

Images