JP2021072332A - Inspection device, method, and program - Google Patents

Abstract

To provide an inspection technology in which an inspection target is not limited to a pattern defect.SOLUTION: An inspection device that inspects an object includes an acquisition unit that acquires a plurality of pieces of measurement data from measurement data at a predetermined position of a predetermined pattern of the object, and a distribution analysis unit that analyzes the distribution between predetermined patterns.SELECTED DRAWING: Figure 12

Description

An embodiment of the present invention relates to a technique for inspecting a pattern formed on an object.

In order to process a semiconductor device such as a logic element or a storage element, a large number of semiconductor wafer processing steps are required to form various characteristics of the semiconductor device and a plurality of layers. For example, lithography is a processing step of transferring a pattern from a mask to a resist placed on a semiconductor wafer. Other processing steps include chemical-mechanical polishing (CMP), etching, film formation, ion implantation and the like. A plurality of semiconductor devices can be processed on a single semiconductor wafer and can be further divided into individual semiconductor devices. The divided object is called a die.

In order to confirm whether an object such as a wafer processed in this way is normally processed, it is necessary to inspect a pattern of a semiconductor circuit or the like formed on the object. Generally, the object is illuminated, and the image signal of the reflected light is used to confirm the pattern formed on the object.

Patent Document 1 discloses a technique of using a standard reference die comparison inspection device for inspecting a wafer. Then, in order to generate a standard reference die for use in the standard reference die comparison inspection, an image of the standard reference die is generated using eight dies around the central die. Eight dies are used because MDAT (Multi-Die Auto Threat; automatic thresholding of multiple dies) is used.

Special Table 2010-534408 Gazette

However, there is a problem that the detection target is only a defect. In addition, it is necessary to set the defect determination condition to be larger than the pattern variation normally assumed. If it is made smaller than the normal pattern variation, there is a problem that a large number of pseudo defects are detected.

Defect detection is detected by the magnitude of the difference between the peripheral die and the standard die generated from the peripheral die at the same position, and the determination condition is calculated from the peripheral difference value. Therefore, there is a problem that the detection sensitivity is lowered due to the influence of the surrounding pattern.

In order to reduce the influence of the surrounding pattern, it is necessary to limit the inspection target area to a minute area to reduce the influence of the surrounding pattern. Further, due to the above, the entire surface of the die cannot be inspected with high sensitivity. That is, there is a problem that only a region assumed in advance can be inspected with high sensitivity.

The present invention has been intensively researched and completed by paying attention to such a problem, and an object of the present invention is to provide an inspection technique in which an inspection target is not limited to a pattern defect.

In order to solve the above problems, the first invention is an inspection device for inspecting an object, and an acquisition unit that acquires a plurality of measurement data from measurement data at a predetermined position of a predetermined pattern of the object. An inspection device including a distribution analysis unit that analyzes the predetermined inter-pattern distribution.

The second invention is an inspection method for inspecting an object, in which a plurality of measurement data are acquired from measurement data at a predetermined position of a predetermined pattern of the object and the distribution between the predetermined patterns is analyzed. The method.

A third invention is an inspection program for inspecting an object, in which a step of acquiring a plurality of measurement data from measurement data of a predetermined position of a predetermined pattern of the object and analysis of the distribution between the predetermined patterns are analyzed. It is an inspection program that causes a computer to execute the steps to be performed.

According to the present invention, it is possible to provide an inspection technique in which the inspection target is not limited to the defect of the pattern.

It is a schematic block diagram of the inspection apparatus which concerns on this embodiment. It is a figure for demonstrating the image acquisition method (the 1 in the case of a line sensor) which concerns on this embodiment. It is a figure for demonstrating the image acquisition method (2 in the case of a line sensor) which concerns on this Embodiment. It is a figure for demonstrating the image acquisition method (in the case of an area sensor) which concerns on this embodiment. It is a figure for demonstrating the image division method which concerns on this Embodiment. It is a schematic block diagram of the image processing apparatus (image distribution type) which concerns on this embodiment. It is a schematic block diagram of the image processing apparatus (image duplication type) which concerns on this embodiment. It is a schematic block diagram of the image processing apparatus (image concatenation transfer type) which concerns on this embodiment. It is a figure for demonstrating the method of generating the inter-image statistic which concerns on this embodiment. This is a part of the flowchart of the image processing method according to the present embodiment. This is another part of the flowchart of the image processing method according to the present embodiment. It is a figure which shows an example of the output of the image processing apparatus which concerns on this embodiment.

Embodiments of the present invention will be described with reference to the drawings. Here, the same reference numerals are given to common parts in each figure, and duplicate description will be omitted.

[A. Outline of this embodiment]
FIG. 1 is a schematic configuration diagram of an inspection device according to the present embodiment. The inspection device 100 includes a stage 110, a light source 130, a collimating lens 140, a BS (beam splitter) 150, an objective lens 160, an imaging lens 170, a camera 180, and an image processing device 190.

The stage 110 carries an object to be inspected (object) 120. The object 120 is a semiconductor wafer or the like. Further, the stage 110 can move in the X, Y, Z, and Θ directions.

The light source 130 is composed of an LPP (Laser Produced Plasma), an HgXe lamp light source, a laser light source, and the like. The collimating lens 140 makes this ray parallel. The BS (beam splitter) 150 splits the light beam from the collimating lens 140 in half and reflects it downward. The objective lens 160 collects the light rays from the BS 150 on the object 120.

The light rays reflected from the surface of the object 120 pass through the BS 150, and the imaging lens 170 forms the reflected light rays. The camera 180 is a line camera such as a TDI (Time Delivery Integration) camera or an area camera, and the image sensor of the camera 180 captures an optical image on the surface of the object 120. The image processing device 190 performs image processing on the image signal captured by the camera 180, and inspects the object 120 by a process described later. Here, among the apparatus configurations of the inspection apparatus 100, the apparatus configuration other than the image processing apparatus 190 is the same as that of a general optical semiconductor wafer inspection apparatus, and detailed description thereof will be omitted. Although the description has been given here with an optical microscope, an image of an object acquired by a microscope other than the optical microscope such as an electron microscope, an X-ray microscope, or a terahertz microscope may be used. Further, although the object is described as a wafer of a semiconductor device, the object is not limited to the wafer alone, and the object is not limited to the wafer, but is a mask (photomask, EUV mask, etc.), FPD (Flat Panel Display), interposer, TSV (Through Silicon Via). It can also be applied to printed circuit boards and the like.

The image processing device 190 is a functional block, and is not limited to being implemented in hardware, and may be implemented in a computer as software such as a program, and its implementation form is not limited. For example, it may be installed and implemented on a dedicated server connected to a client terminal such as a personal computer and a wired or wireless communication line (Internet line, etc.), or it may be implemented using a so-called cloud service. Good.

[B. Example of image processing device]
A specific embodiment of the image processing apparatus 190 will be described. In this embodiment, an image of an object is used, but the measurement data may be data other than an image such as a reflected light spectrum. In the case of data other than images, each measurement data may be treated in the same manner as the pixel value of this embodiment. For example, in the case of the reflected light spectrum, the measured value of each wavelength can be replaced with the pixel value of this embodiment and processed in the same manner.

(B1. Defect detection)
This embodiment comprises four processing steps. The first processing step is an image acquisition (or input) step. Acquire a plurality of images of a predetermined position of a predetermined pattern on an object. It is assumed that the input image has completed the filter processing, the rotation / inversion processing, the difference processing with the reference image, and the alignment processing as preprocessing. Here, the plurality of images at a predetermined position in a predetermined pattern may be from within the same object, from within the plurality of objects, or a combination thereof. Here, "acquiring a plurality of images of a predetermined position of a predetermined pattern on an object" means acquiring a plurality of images of the pattern intended to form the same pattern on the object. Say.

The second processing step is a pixel statistic calculation step. The pixel statistic is calculated from the pixel values in a predetermined area of each image. The "pixel statistic" is a value calculated from pixel values in a predetermined area such as an average value, an intermediate value, a maximum value, a minimum value, and a dispersion value of pixel values in a predetermined area. There may be a plurality of pixel statistics. Here, it is the average value of the pixel values in a predetermined area. Further, the size of the predetermined region is not particularly limited and may be one pixel or more. In the case of one pixel, the pixel value may be extracted as a pixel statistic.

The third processing step is an inter-image statistic calculation step. The distribution analysis of the pixel statistics of each region obtained in the second processing step is performed, and the inter-image statistics are calculated. The "inter-image statistic" is a value calculated from the pixel statistic such as the average value, the intermediate value, the maximum value, the minimum value, and the variance value of the pixel statistic. There may be a plurality of inter-image statistics. Here, the average value and the variance value of the pixel statistics are used.

The fourth processing step is a defect detection step. The defect determination threshold value is set from the inter-image statistics obtained in the third processing step, and the region where the pixel value statistics are larger (or smaller) than the defect determination threshold value is output as a defect. The output data includes the position in the object (die index and coordinates in the die) of the output area and additional information (pixel statistic and inter-image statistic of the output area).

The above four processing steps are performed on the inspection area of the object. If necessary, the images may be divided into groups, inter-image statistics may be calculated for each group, and defect detection may be performed. Here, group division refers to the position in the die where the image was acquired, the position in the wafer plane (radius, angle, x-coordinate, y-coordinate, etc.), the position in the shot of lithography, the scanning direction of lithography, and the pattern direction (rotation, inversion). Etc.) to divide into multiple groups. Further, a plurality of types of group division may be performed, and the defect determination may be performed by reflecting the inter-image statistics obtained by the plurality of group division.

(B2. Hot spot extraction)
This embodiment comprises five processing steps. The first processing step is an image acquisition (or input) step. Acquire a plurality of images of a predetermined position of a predetermined pattern on an object.

The second processing step is a pixel statistic calculation step. The pixel statistic is calculated from the pixel values in a predetermined area of each image.

The third processing step is an inter-image statistic calculation step. The distribution analysis of the pixel statistic of the predetermined region obtained in the second processing step is performed, and the inter-image statistic is calculated. Here, the average value and the variance value of the pixel statistics in a predetermined area are used.

The first to third processing steps are performed on a plurality of predetermined areas. At this time, the predetermined area may be another area (which may overlap) in the same image or an area in another image. The in-plane distribution of the pixel statistic and the inter-image statistic may be output as a two-dimensional image or digitized. For example, the numerical values of the pixel statistic and the inter-image statistic are converted into colors and gray levels and the in-plane distribution is output as a two-dimensional image, the wire frame or contour lines are output, and the cross-sectional shape at a predetermined position is output. You may output it. As the digitization, the approximate plane may be obtained from the in-plane distribution of the pixel statistic and the inter-image statistic, and the normal vector and slope of the approximate plane may be output.

The fourth processing step is the in-plane distribution statistic calculation step. The in-plane distribution statistic is calculated by performing the in-plane distribution analysis of the inter-image statistic obtained in a plurality of predetermined regions. The "in-plane distribution statistic" is calculated from the inter-image statistic of a plurality of predetermined regions such as the average value, the intermediate value, the maximum value, the minimum value, and the variance value of the inter-image statistic of the plurality of predetermined regions. It is a statistic. Here, the average value and the variance value of the inter-image statistics of a plurality of predetermined regions are used.

The fifth processing step is a hotspot extraction step. The hotspot determination threshold is set from the in-plane distribution statistic obtained in the fourth processing step, and the region where the inter-image statistic is larger (or smaller) than the hotspot determination threshold is output as a hotspot. The output data includes the position in the object of the area and additional information. Further, without setting the hotspot determination threshold value, a region sampled based on the order of large (or small) inter-image statistics or a region having a predetermined cumulative probability may be output as a hotspot.

Extracting a region with a large inter-image statistic using the average value of the pixel values as the pixel statistic and the variance value of the pixel statistic as the inter-image statistic is a region with a large variance value and a large variation from a plurality of regions. Means to extract. Since hot spots have a small processing margin and are sensitive to processing conditions, it is considered that there is a large variation. Therefore, hotspots can be extracted by extracting regions with large variations.

(B3. Processing monitor point selection)
This embodiment comprises three processing steps. The first processing step is an image acquisition (or input) step. Here, in an image including processing monitor points (for example, points measured by a CD-SEM), a plurality of points may be taken from the same die, points within the same die may be taken from a plurality of dies, or a combination thereof. .. Further, an image including a region extracted by the hot spot extraction described above may be used.

The second processing step is a pixel statistic calculation step. The pixel statistic is calculated from the pixel values in a predetermined area of each image.

Here, the predetermined area is an area including the processing monitor point. The region may be the region extracted by the hot spot extraction described above. In the case of the region extracted by the hotspot extraction, the pixel statistic obtained by the hotspot extraction may be used.

The third processing step is a processing monitor point selection step. The distribution analysis of the pixel statistic of the predetermined region obtained in the second processing step is performed, and the region where the pixel statistic has the maximum, minimum, and predetermined cumulative frequency is selected as the processing monitor point. It may be selected except for the region deviating from the predetermined statistical distribution. For example, by confirming the areas where the pixel statistics are the maximum and the minimum, it is possible to confirm the range of processing variation more efficiently than the fixed processing monitor points and the randomly sampled processing monitor points.

(B4. Processing variation analysis)
This embodiment consists of seven processing steps. The first processing step is an image acquisition (or input) step. Acquire a plurality of images of a predetermined position of a predetermined pattern on an object.

The second processing step is a pixel statistic calculation step. The pixel statistic is calculated from the pixel values in a predetermined area of each image. Here, it is the average value of the pixel values in a predetermined area.

The third processing step is a grouping step. The pixel statistic of each image obtained in the second processing step is divided into predetermined groups. Group division is the position in the die or wafer plane (radius, angle, x-coordinate, y-coordinate, etc.) from which the image was acquired, the position in the shot of lithography, the scanning direction and pattern direction of lithography (rotation, inversion, etc.), etc. It is to divide into multiple groups based on. Group division may be performed by a plurality of types. Here, the grouping is performed after the pixel statistic is calculated, but the grouping may be performed before the image acquisition or after the image acquisition.

The fourth processing step is an inter-image statistic calculation step. The distribution analysis of the pixel statistics in a predetermined region is performed for each group divided in the third processing step, and the inter-image statistics are calculated. Here, the average value and the variance value of the pixel statistics in a predetermined area are used.

The fifth processing step is an intergroup statistic calculation step. The distribution analysis of the inter-image statistic for each group obtained in the fourth processing step is performed, and the inter-group statistic is calculated. The "inter-group statistic" is a statistic calculated from the inter-image statistic for each group such as the average value, the intermediate value, the maximum value, the minimum value, and the variance value of the inter-image statistic for each group.

The first to fifth processing steps are performed on a plurality of predetermined areas. At this time, the predetermined area may be another area (which may overlap) in the same image or an area in another image. The in-plane distribution of the pixel statistic, the inter-image statistic, and the inter-group statistic may be output as a two-dimensional image or digitized. If the in-plane distribution is not obtained, it is not necessary to perform the first to fifth processing steps for a plurality of predetermined regions.

The sixth processing step is the intergroup distribution statistic calculation step. The in-plane distribution analysis of the inter-group statistics obtained in the fifth processing step is performed, and the in-group distribution statistics are calculated. "Inter-group distribution statistic" is calculated from inter-group statistics of a plurality of predetermined regions such as an average value, an intermediate value, a maximum value, a minimum value, and a variance value of inter-group statistics of a plurality of predetermined regions. It is the statistic that is done.

The seventh processing step is a processing variation analysis step. The distribution of processing variations is analyzed by analyzing the inter-image statistic, the inter-group statistic, the inter-group in-plane distribution statistic, etc. obtained in the fourth to sixth processing steps. For example, when the inter-image statistics differ between groups, it can be interpreted that a predetermined area may have a tendency to differ between groups. The effectiveness of the group division type can be judged from the intergroup statistics. From the in-plane distribution statistics between groups, it is possible to determine where in the in-plane the variation occurs.

(B5. Processing monitor spot extraction)
This embodiment comprises four processing steps. The first processing step is an image acquisition (or input) step. Acquire a plurality of images of a predetermined position of a predetermined pattern on an object. Further, the image may be acquired from an object having different processing conditions.

The second processing step is a pixel statistic calculation step. The pixel statistic is calculated from the pixel values in a predetermined area of each image. Here, it is the average value of the pixel values in a predetermined area.

The third processing step is an inter-image statistic calculation step. The distribution analysis of the pixel statistic of the predetermined region obtained in the second processing step is performed, and the inter-image statistic is calculated. Here, it is a variance value of the pixel statistic in a predetermined area.

The first to third processing steps are executed for a plurality of predetermined areas. At this time, the predetermined area may be another area (may overlap) in the same image or an area in another image.

The fourth processing step is a processing monitor spot extraction step. The distribution analysis of the inter-image statistic obtained in a predetermined region is performed, and the region where the inter-image statistic is maximized is extracted as a processing monitor spot. Regions outside the predetermined statistical distribution may be excluded. Further, a plurality of areas may be extracted as processing monitor spots by sampling from the one with the larger inter-image statistic.

Further, when a plurality of predetermined regions are acquired from an object having the same processing conditions under a plurality of processing conditions, the dispersion value of the pixel statistic for each predetermined region is small within the same processing conditions, and the processing conditions are different. And a region having a large dispersion value of the pixel statistic for each predetermined region may be extracted. The in-plane distribution of the inter-image statistics of a plurality of predetermined regions may be output as a two-dimensional image or a numerical value.

(B6. Processing window analysis point selection)
This embodiment comprises three processing steps. The first processing step is an image acquisition (or input) step. A plurality of images of a predetermined position of a predetermined pattern are acquired from each of the objects produced under a plurality of processing conditions. Here, it is desirable that there are processing-sensitive points in the image.

The second processing step is a pixel statistic calculation step. The pixel statistic is calculated from the pixel values in a predetermined area of each image. Here, it is the average value of the pixel values in a predetermined area. Here, it is desirable that the predetermined region is a region including a process-sensitive point.

The third processing step is a processing window analysis point selection step. The distribution analysis of the pixel statistic calculated from the predetermined area of each image of the object created under the same processing condition is performed for each different processing condition, and the area where the pixel statistic has the maximum, minimum, and predetermined cumulative frequency is processed. Select as a window analysis point. It may be selected except for the region deviating from the predetermined statistical distribution.

The selected processing window analysis point is confirmed by CD-SEM or the like, and the processing window (optimal range of processing conditions) is obtained. In the past, since confirmation was performed at the same position in the die, an error occurred in the processing window obtained due to processing variation. According to this embodiment, it is possible to efficiently perform confirmation including processing variation, and it is possible to obtain a processing window with higher accuracy than before.

(B7. Processing change monitor)
This embodiment comprises three processing steps. The first processing step is an image acquisition (or input) step. A plurality of images of a predetermined position of a predetermined pattern are acquired from an object.

The second processing step is a pixel statistic calculation step. The pixel statistic is calculated from the pixel value of a predetermined area of each image and the reference value (including the image) of the pixel statistic stored in advance. Here, it is the average value of the difference values between the pixel value and the reference value in the predetermined area. The reference value of the pixel statistic may be a pixel unit or a predetermined area unit for which the pixel statistic is calculated. Further, after calculating the pixel statistic, the calculation with the reference value may be performed. There may be a plurality of reference values for one predetermined region (for example, an average value and a variance value).

The third processing step is an inter-image statistic analysis step. The distribution analysis of the pixel statistic of the predetermined region obtained in the second processing step is performed, and the inter-image statistic is calculated. The in-plane distribution of the pixel statistics may be output as a two-dimensional image or digitized.

The first to third processing steps may be performed on a plurality of regions, and the in-plane distribution of the pixel statistics and the inter-image statistics may be converted into a two-dimensional image or digitized and output.

(B8. Processing change analysis)
This embodiment consists of seven processing steps. The first processing step is an image acquisition (or input) step. A plurality of images of a predetermined position of a predetermined pattern are acquired from an object.

The second processing step is a pixel statistic calculation step. The pixel statistic is calculated from the pixel value of a predetermined area of each image and the reference value (including the image) of the pixel statistic stored in advance. Here, it is the average value of the difference values between the pixel value and the reference value in the predetermined area.

The third processing step is a grouping step. The pixel statistics of each region obtained in the second processing step are divided into predetermined groups.

The fourth processing step is an inter-image statistic calculation step. The distribution analysis of the pixel statistics of each region is performed for each group divided in the third processing step, and the inter-image statistics are calculated. Here, the average value and the variance value of the pixel statistics of the area are used. In addition, an area, a group, a group division type, or the like in which the inter-image statistics differ between the groups may be output. Areas, groups, group division types, etc. may be output in comparison with the overall inter-image statistics that are not divided into groups.

The fifth processing step is an intergroup statistic calculation step. The inter-image statistic for each group obtained in the fourth processing step is analyzed between the groups, and the inter-group statistic is calculated. The inter-group statistic is a statistic calculated from the inter-image statistic for each group such as the average value, the intermediate value, the maximum value, the minimum value, and the variance value of the inter-image statistic for each group. Here, a predetermined area, group, group division type, or the like may be output according to the inter-group statistics.

The first to fifth processing steps are executed for a plurality of predetermined areas. At this time, the predetermined area may be another area (which may overlap) in the same image or an area in another image. Further, the in-plane distribution of the pixel statistic, the inter-image statistic, and the inter-group statistic may be digitized and output as a two-dimensional image.

The sixth processing step is the intergroup distribution statistic calculation step. The in-plane distribution analysis of the inter-group statistics obtained in the fifth processing step is performed, and the in-group distribution statistics are calculated. The intergroup in-plane distribution statistic is calculated from the intergroup statistics of a plurality of predetermined regions such as the average value, the intermediate value, the maximum value, the minimum value, and the variance value of the intergroup statistics of a plurality of predetermined regions. It is a statistic.

The seventh processing step is a processing change analysis step. The distribution of processing changes is analyzed by analyzing the inter-image statistic, the inter-group statistic, the inter-group in-plane distribution statistic, etc. obtained in the fourth to sixth processing steps. For example, when the inter-image statistics differ between the groups, it can be interpreted that there is a possibility that a predetermined area has different processing changes between the groups. The effectiveness of the group division type can be judged from the intergroup statistics. From the intergroup distribution statistics, it is possible to determine where in the plane the processing change occurs.

(B9. Image acquisition method, in the case of line camera, part 1)
FIG. 2 is a diagram for explaining an image acquisition method (No. 1 in the case of a line camera) according to the present embodiment. In order to obtain the same position of the die or shot to be inspected, it is necessary to invert the image depending on the scanning direction.

One device area is called a die, and one exposure area by lithography is called a shot. A normal die is smaller than a shot and contains multiple dies in the shot. Here, a case where 2 × 2 dies 220 are arranged in one shot 210 will be described (the explanations of other image acquisition methods described later are also arranged in the same manner).

A plurality of shots are arranged on the wafer 200. In the inspection, images are acquired by sequentially scanning from the edge of the inspection target area. Here, the case of starting from the upper left will be described.

Scanning shall be performed in the left-right direction. The area corresponding to one scan is defined as one die row. An image of one die row is taken from the top and then scanned at the same position on the adjacent die on the bottom. Repeat this until the bottom edge of the wafer (see solid arrow).

As a result, all the same positions of the dies on the inspection area are scanned and an image is acquired. In this case, since the scanning directions are opposite in the odd-numbered scanning and the even-numbered scanning, it is necessary to reverse the image direction before the subsequent image processing or to perform the processing separately in the scanning direction.

Next, the adjacent inspection area in the die is scanned from the bottom to the top in order to acquire an image (see the dotted arrow). This is repeated until the image acquisition of the entire inspection area is completed. Here, the images of the same positions of all the inspection dies are continuously acquired, but the processing may be divided and processed due to the limitation of the processing capacity of the image processing apparatus 190 and the like.

(B10. Image acquisition method, in the case of line camera, part 2)
FIG. 3 is a diagram for explaining an image acquisition method (No. 2 in the case of a line camera) according to the present embodiment. Double the image memory is required to obtain the same position of the die or shot to be inspected, but image inversion depending on the scanning direction is not necessary.

In this example, different inspection areas in the die are scanned in each of the left and right scanning directions. Therefore, since the same position of all inspection dies can be acquired in the same scanning direction, it is not necessary to flip the images horizontally and align them. However, twice the image storage capacity is required as compared with the method of FIG.

(B11. Image acquisition method, in the case of area camera)
FIG. 4 is a diagram for explaining an image acquisition method (in the case of an area camera) according to the present embodiment. This example is when an area camera is used. The image is acquired by sequentially moving the same position in the die of the inspection die.

(B12. Image division)
FIG. 5 is a diagram for explaining an image division method according to the present embodiment. The image processing device 190 is composed of a plurality of image processing units (in this figure, two image processing units 1 and 2). The acquired image is divided in the scanning direction and distributed to each image processing unit. Here, too, the case where 2 × 2 (4) dies are arranged in one shot will be described.

The acquired image is distributed to each image processing unit so that the same position image in the die can be processed by one image processing unit. Due to restrictions such as the processing capacity of the image processing unit, an image at the same position in the die may be processed by a plurality of image processing units (especially effective when the number of dies is large). Alternatively, the image size processed by one image processing unit may be adjusted according to the number of inspection dies. Further, the acquired image may be divided in the scanning direction as shown in FIG. 5 and distributed to a plurality of image processing units.

(B13. Configuration of image processing device)
Three configuration examples of the image processing device 190 will be introduced. FIG. 6 is a schematic configuration diagram of an image distribution type image processing device 190. FIG. 7 is a schematic configuration diagram of an image duplication type image processing device 190. FIG. 8 is a schematic configuration diagram of a result composite output type image concatenated transfer type image processing device 190. Here, the image concatenated transfer type is also referred to as a daisy chain type.

In FIG. 6, the image data acquired by the camera 180 is transferred to the image distribution units 1 to M, and the image distribution units 1 to M distribute the transferred image data to the image processing units 1 to N via a network. The reason why there are a plurality of image distribution units is that the image data acquired by the camera 180 has a large capacity and cannot be processed by one image distribution unit. Therefore, if the processing capacity of the image distribution unit is sufficient, one image distribution unit may be used.

Here, the image distribution units 1 to M may be preprocessed so that the image processing units 1 to N can easily process the images. The preprocessing includes distortion correction of the optical system, pixel value correction for each pixel, pixel rearrangement, extraction of an area corresponding to an inspection area, and the like. Further, the alignment process of the same position image of the die may be performed here. Further, the reference image may be acquired from the image server to perform alignment processing and difference image creation.

The image processing units 1 to N process and inspect the image data transferred from the image distribution units 1 to M. There may be a plurality of cameras 180. When a plurality of cameras acquire images at the same position of an object, image data corresponding to the same position may be connected to each image distribution unit. By doing so, processing between images at the same position acquired by a plurality of cameras becomes easy.

In FIG. 7, the image data acquired by the camera 180 is duplicated by the image duplication unit and transferred to the image processing units 1 to M. Each image processing unit acquires necessary image data according to the processing area allocated to each image processing unit and performs inspection. The inspection process (including preprocessing such as pixel value conversion and alignment process) may be distributed to the image processing units M + 1 to N which are not connected to the image duplication unit, if necessary.

In FIG. 8, the image data acquired by the camera 180 is connected to the image processing unit 1, and each image processing unit transfers a copy of the acquired image data to the next image processing unit. Each image processing unit acquires necessary image data according to the processing area allocated to each image processing unit and performs inspection. The inspection process (including preprocessing such as pixel value conversion and alignment process) may be distributed to the image processing units M + 1 to N which are not connected to the image duplication unit, if necessary.

(B14. Explanation of inter-image statistic generation)
FIG. 9 is a diagram for explaining a method of generating from the image to the inter-image statistic according to the present embodiment. Here, the images from 1 to n are arranged. Here, it is assumed that the images from 1 to n have completed the filter processing, the rotation / inversion processing, the difference processing with the reference image, the alignment processing, and the like as preprocessing. The inter-image statistic generation method is composed of three processing steps.

The first processing step calculates a statistic (pixel statistic) of a region (here, a region of 5 pixels × 5 pixels) in the region image. For example, the average value, the dispersion value, the maximum value, the minimum value, and the like of the pixel values in the area are calculated. Further, the difference value from the reference image may be calculated, and the average value, the dispersion value, the maximum value, the minimum value, and the like of the difference value may be calculated. This is done for the same area of all 1 to n images.

In the second processing step, the distribution analysis of the pixel statistics of the 1-n images obtained in the first processing step is performed, and the statistics (inter-image statistics) are calculated. For example, the average value, the variance value, the maximum value, the minimum value, and the like of the pixel statistics of the 1 to n images are calculated. Here, the difference value from the reference image may be obtained.

In the third processing step, the first and second processing steps are performed on the area set in the image. The area may be set individually for each area, the image may be automatically set in the area of 5 pixels × 5 pixels as shown in FIG. 9, or a combination thereof may be used.

Here, the areas for which the pixel statistics are calculated do not overlap, but the areas may be set so as to overlap. For example, when dividing into an area of 4 pixels × 4 pixels, another set of area divisions may be performed in which the areas are shifted by 2 pixels in the vertical and horizontal directions. This has the disadvantage of increasing the amount of calculation, but has the advantage of reducing the risk that the defect will be divided into a plurality of regions and the defect signal will become small and undetectable.

If there is additional information such as the type and direction (reversal / rotation) of the pattern of the region, grouping may be performed based on the additional information, and the first and second processes may be performed.

By this process, an image in which each statistic (pixel statistic) is used as a pixel value is created. Here, the number of pixels in one image is the number of predetermined regions.

Next, a method of calculating various statistics such as a statistic between images 1 and n (inter-image statistic) using this pixel statistic will be described.

(B15. Processing calculation method 1: In the case of the same processing conditions)
A specific processing method will be described using a mathematical formula. It consists of 12 processing steps. The first processing step is an image acquisition (or input) step. Let G (Die, Area, x, y) be a pixel value when a plurality of images of the same position of the die in the wafer are acquired. Here, Die indicates the die index in the Wafer, Area indicates the index in the die of the acquired image, and x and y indicate the pixel positions in a predetermined area. It is assumed that the input image has been preprocessed such as filter processing, rotation / inversion processing, difference processing, and alignment.

ここで、ｗは所定の領域のｘ方向画素数、ｈは所定の領域のｙ方向画素数、ｉはダイ内の所定の領域のインデックスである。所定の領域の位置はレシピ等で指定された位置、または所定の領域の大きさ等を指定して自動的に配置された位置、その組み合わせでもよい。所定の領域の大きさには特に制限はなく１画素以上であればよい。また、複数の領域が設定されている場合、領域間でオーバラップがあってもよい。ここでは、矩形領域として説明したが、所定の領域は矩形とは限らず任意形状でよい。 The second processing step is a pixel statistic calculation step. The statistic is calculated from the pixel values of a predetermined area in the image. For example, the average pixel value of the area is expressed by the equation (1).

Here, w is the number of pixels in the x-direction of the predetermined region, h is the number of pixels in the y-direction of the predetermined region, and i is the index of the predetermined region in the die. The position of the predetermined area may be a position specified in the recipe or the like, a position automatically arranged by designating the size of the predetermined area, or a combination thereof. The size of the predetermined area is not particularly limited and may be one pixel or more. Further, when a plurality of areas are set, there may be overlap between the areas. Although described here as a rectangular area, the predetermined area is not limited to a rectangle and may have an arbitrary shape.

It is also possible to process a plurality of the same pattern regions (including rotation / inversion) in the die as the same predetermined region by performing rotation / inversion processing as preprocessing. In this case, it is preferable to save additional information (rotation / inversion and position in the die) so that it can be used for grouping.

ここで、ｎは画像を取得したＤｉｅの個数である。ダイ内の複数の同一パターン領域全体の画像間統計量を算出する場合は、nは画像を取得したダイの個数とダイ内の同一パターン領域数の積となる。 The third processing step is an inter-image statistic calculation step. Calculate the inter-image statistics for a given area. For example, the inter-image averaging and inter-image variance of the pixel value averaging in a predetermined region are represented by the equations (2) and (3), respectively.

Here, n is the number of Die that acquired the image. When calculating the inter-image statistics for a plurality of identical pattern regions in a die, n is the product of the number of dies for which images have been acquired and the number of identical pattern regions in the dies.

または、

ここで、Ｔｈ_ａ ^＋、Ｔｈ_ｂ ^＋、Ｔｈ_ａ ⁻、Ｔｈ_ｂ ⁻はレシピ等で指定した固定値である。これにより抽出されたダイと所定の領域の位置を欠陥として出力する。 The fourth processing step is a defect detection step. Check for defects in a given area. For example, defect detection is performed using equations (4) and (5) as defect determination conditions.

Or

Here, Th _a ⁺ , Th _b ⁺ , Th _a ⁻ , and Th _b ⁻ are fixed values specified in the recipe or the like. The position of the die extracted by this and the predetermined area is output as a defect.

In the conventional method, the difference from the reference image is taken as preprocessing, and the defect determination condition is set by the statistic with the peripheral area in the same image instead of the inter-image statistic. Therefore, it was not possible to eliminate the influence outside the predetermined region on the defect determination in the predetermined region. In this embodiment, since the defect determination condition is set only in the predetermined area, the influence other than the predetermined area can be eliminated.

When the third and fourth processing steps are performed for each group (for example, the central portion and the outer peripheral portion of the wafer), the dispersion between images becomes smaller and inspection with high sensitivity becomes possible. A plurality of groups may be performed, the optimum grouping may be selected, or defects may be determined by a combination.

The fifth processing step is an observation die detection step. When the predetermined region includes a pattern for observing the pattern shape by CD-SEM, sampling may be performed in consideration of the pattern variation distribution. For example, when confirming the entire pattern variation, the maximum / minimum area of the pixel statistic excluding the defect obtained in the fourth processing step is output. If the average pattern shape is to be confirmed, the area where the pixel statistic has a cumulative probability of around 50% is output. In addition to this, a region of a specific cumulative probability may be output depending on the purpose. Instead of sorting, a histogram of pixel statistics may be created and regions near a predetermined cumulative probability may be extracted.

ここで、ｍは所定の領域の個数である。 The sixth processing step is an in-plane distribution statistic calculation step. The in-plane distribution of inter-image statistics for a plurality of predetermined regions is calculated. For example, the in-plane mean and the in-plane variance of the inter-image dispersion of the pixel value averages in a predetermined region are represented by the equations (6) and (7).

Here, m is the number of predetermined regions.

ここで、Ｄはレシピ等で指定した固定値である。 Since the inter-image variance of the pixel value average in a predetermined region is affected by the inter-image average, it may be corrected as in the equation (8) to obtain the corrected inter-image variance.

Here, D is a fixed value specified in the recipe or the like.

ここで、ＧＴｈ_ａ ^＋、ＧＴｈ_ｂ ^＋はレシピ等で指定した固定値である。これにより検出された領域位置を異常領域（ホットスポット）として出力する。 The seventh processing step is an abnormal region detection step. Detects an abnormal region with large variations in pixel values. For example, the abnormal region is detected by using the equation (9) as an abnormality determination condition.

Here, GTh _a ⁺ and GTh _b ⁺ are fixed values specified in a recipe or the like. The area position detected by this is output as an abnormal area (hot spot).

The eighth processing step is a confirmation die extraction step. The confirmation die is extracted from the abnormal region detected in the sixth processing step. Corresponding to the abnormal region statistic _{S ave (Die, i a)} (i a is the index corresponding to the abnormal region) rearranges the extracted outputs a die with a statistic of a predetermined cumulative probability. At this time, the die having the maximum / minimum statistic excluding defects may be extracted. Instead of sorting, a histogram of the statistics may be created and dies near a predetermined cumulative probability may be extracted.

The ninth processing step is an abnormal region grouping step. In order to analyze the cause of variation in the abnormal region, the pixel statistics of the abnormal region are grouped. Here, a case where grouping is performed according to the die position in the shot will be described. S _ave (Die, _{i a)} to grouping in shot die position.

ここで、ｇはショット内ダイ位置を示すインデックスである。 The tenth processing step is an intra-group statistic calculation step. Calculate inter-image statistics for each group. For example, the intra-group average and the intra-group variance of the pixel value average of the region are represented by the equations (10) and (11).

Here, g is an index indicating the die position in the shot.

The eleventh processing step is an intergroup analysis step. Find out if intragroup statistics differ between groups. For example, find out if the intra-group variance differs depending on the in-shot die position. If they are different, it is expected that the variation is caused by the mask.

The twelfth processing step is a variation cause analysis step. The 8th, 9th, and 10th processing steps are performed in various groups (for example, the scanning direction of lithography, the position in the wafer (radius, angle, x, y), etc.), and the group has different dispersion within the group. Extract the division. The processing can be further stabilized by estimating the causative process from the grouping extracted here and taking countermeasures.

(B16. Processing calculation method 2: In the case of multiple processing conditions)
The case of extracting a processing-sensitive region (processing monitor spot) will be described here by taking an FEM (Focus Exposure Matrix) wafer performed in confirmation of lithography conditions as an example. This embodiment comprises four processing steps. The first processing step is an image acquisition (or input) step. Let G (F, E, Area, x, y) be a pixel value when an image of the same position of the die in the FEM wafer is acquired for the die created under a plurality of exposure conditions. Here, F is the focal position, E is the exposure amount, Area is the index in the die of the image, and x and y are the pixel positions in the image.

The second processing step is a pixel statistic calculation step. The statistic is calculated from the pixel values of a predetermined area in the image. For example, the average pixel value in a predetermined area is represented by the equation (12).

Here, w is the number of pixels in the x-direction of the predetermined region, h is the number of pixels in the y-direction of the predetermined region, and i is the index of the predetermined region in the die.

The third processing step is an inter-image statistic calculation step.

The average and variance of the pixel statistics of all groups are expressed by equations (13) and (14), respectively.

The average of the pixel statistics for each exposure amount group and the focal position variance of the total exposure amount are expressed by equations (15) and (16), respectively.

The average of the pixel statistics for each focal position group and the exposure amount variance of all focal positions are expressed by equations (17) and (18), respectively.

The correlation between the pixel statistic and the focal position is expressed by the equation (19).

The correlation between the pixel statistic and the exposure amount is expressed by the equation (20).

Here, n is the number of images, n _F is the number of conditions for the focal position, n _E is the number of conditions for the exposure amount, “F￣” is the average value of the focal position, and “E￣” is the average value of the exposure amount. Originally, the overline "￣" is written above "F" and "E", but due to the character code that can be used in the specification, "F", "E" and "￣""Is listed side by side. Here, the images under the same conditions are one image, n = n _F × n _E. Further, as a method of increasing the population, the same pattern image in the die or the same position image of the die in the same shot may be added.

The fourth processing step is a processing monitor spot extraction step. Extract the region that is highly dependent on the exposure conditions. For example, a region having a large exposure amount dispersion at all focal positions is extracted, and that position is output as a processing monitor spot. Extraction of a region that is greatly affected by either the focal position or the exposure amount is also useful for process control.

For example, in order to investigate the influence of the focal position, a region in which the influence of the focal position is large (the focal position dispersion of the total exposure amount is large) and the influence of the exposure amount is small (the exposure amount dispersion of the total exposure amount is small) is extracted. You may. When it has an extreme value such as the focal position, a region having a high correlation (the absolute value of the correlation with the focal position is large) may be extracted. An evaluation function that integrates the above may be set to extract the area. For example, the equation (21) may be used as the focal position evaluation value. Here, α, β, and γ are fixed values specified in the recipe or the like.

(B17. Flowchart of image processing)
An example of a flowchart of the image processing method according to the present embodiment will be described. Both FIGS. 10 and 11 show a flowchart of an image processing method according to the present embodiment.

In the image input step (S110), a plurality of images at predetermined positions of the plurality of dies are input. Here, a plurality of die-row dies are included. In addition, images of a plurality of die rows are saved in the image memory. As the inspection target die, four or more dies may be used. You may enter all the dies to be inspected. If the dies are small and the number of dies in one wafer is very large, they may be divided. When a plurality of identical pattern images in the die are included, inversion / rotation processing or the like may be performed as necessary.

The inside of one die is divided in the scanning direction and processed by a plurality of image processing units. At this time, it is preferable that the same portion in the die is distributed to the same image processing unit. For example, assuming that there are 70 dies and the image of one die is 4096 pixels × 4096 pixels, 70 images of 4096 pixels × 4096 pixels = 16M pixels are input images.

In the filter processing step (S120), the input image is filtered. As the filter processing, in addition to general filter processing such as a low-pass filter for noise reduction and a sobel filter for detecting edges, a filter for obtaining a difference from a reference image may be used. This process can be skipped as appropriate.

In the alignment step (S130), the position of the same location image in the die input in S110 is adjusted. Here, a characteristic pattern is extracted from the reference image. The reference image is an image of the same coordinates in the die saved in advance, one image of the input image (preferably an image near the center of the wafer), an image of the same pattern as the input image such as a simulation image. To use. The characteristic pattern may be decided in advance. There may be a plurality of characteristic patterns.

The correlation coefficient is calculated in the region containing the characteristic pattern while shifting one or both of each image or the reference image. The amount of shift having the largest correlation coefficient is used as the alignment value of the image with respect to the reference image. It is desirable to obtain the alignment value up to an accuracy of 1 pixel or less (0.0 or less). The correlation coefficient for a plurality of shift amounts may be obtained and complemented to obtain the shift amount corresponding to the maximum value of the correlation coefficient. An image shifted by the alignment value obtained in this way is created.

In the area division step (S140), each image is divided into a plurality of areas. For example, it is divided into 4 × 4 pixels. When the divided area is one pixel, S140 and S150 may be skipped. The divided areas may overlap or be spaced apart from each other. The size of the area may be determined and automatically divided, a predetermined area may be specified, or a combination thereof may be used. Moreover, the size of the region does not have to be the same.

In the pixel statistics processing step (S150), the pixel statistics are calculated for each region of each image. Here, the pixel statistic is an average value, a variance value, a maximum value, a minimum value, a range, an intermediate value, a mode value, or the like of pixel values in a region. An image of each statistic is created by this process.

For example, 70 images of 1024 pixels × 1024 pixels are created for each statistic. Hereinafter, the pixel average value in the region will be described as the pixel statistic.

In the pixel statistic distribution analysis step (S160), the inter-image statistic is calculated for each region of the pixel statistic obtained in S150. Here, the inter-image statistic is an average value, a variance value, a maximum value, a minimum value, a range, an intermediate value, a mode value, or the like of the pixel statistic of each region. The dies may be classified into a plurality of groups and the inter-image statistics may be calculated for each group. There may be a plurality of groove divisions. The group includes the die position in the shot of lithography, the position in the wafer (radius, angle, up / down, left / right, etc.), processing conditions, wafer, lot, scanning direction of lithography, and the like.

For each region of 1024 × 1024 = 1M, the average value, the variance value, the maximum value, the minimum value, the range, the intermediate value, the mode value, etc. are obtained for the pixel statistics for 70 dies. In the following, the mean value and the variance value are used. In this way, the inter-image statistics are calculated.

S170 and S180 described below relate to defect determination. If the defect is not judged, it may be skipped.

In the defect determination condition setting step (S170), the threshold value is set for each region based on the inter-image statistics obtained in S160. As an example of the threshold value, the average value ± (variance value × constant 1 + constant 2) may be used.

In the defect region extraction step (S180), the defect region (which region of which die) is extracted according to the threshold value set in S170. If there is a defective area in the vicinity of the same die, it may be merged and output.

When the in-shot die position is grouped, the inter-image statistic (dispersion value of pixel statistic) is about the same in each in-shot die position group, and the inter-image statistic (average value of pixel statistic) is compared with the variance value. If they are significantly different, a common defect (mask defect) is suspected. When grouped by the radial position in the wafer, it is possible to suppress the detection of pseudo defects due to the increase in variation due to the wafer edge.

In the inter-image statistic distribution analysis step (S190), the in-plane distribution statistic of the inter-image statistic (1024 × 1024) obtained in S160 is calculated. That is, the distribution of inter-image statistics is analyzed. The in-plane distribution statistic is an average value, a variance value, a maximum value, a minimum value, a range, an intermediate value, a mode value, etc. of the in-plane interimage statistic. In the case of grouping, inter-group statistics and inter-group in-plane distribution statistics may be further obtained. When grouping is performed based on processing conditions or the like, the correlation with the grouping conditions may be obtained. The variance value obtained in S160 may be used.

The variance value may be corrected by the average value. As a result, the influence of shot noise can be reduced.

In the monitor area determination condition setting step (S200), the monitor area determination condition is set according to the in-plane distribution statistic obtained in S190. The basic idea regarding the determination is the same as that of S170.

In the monitor area extraction step (S210), the monitor area is extracted according to the monitor area determination condition obtained in S200. A region having a large variance value obtained in S160 may be extracted.

In the review area extraction step (S220), the defect area is removed from the monitor area extracted in S210, and the maximum value / minimum value die and the area of a predetermined cumulative probability are extracted as review points.

(B17. Input, group division, output)
The input image requires a plurality of identical pattern images. For example, 1) an image obtained by selecting a plurality of the same parts of a die shot from within a wafer, within a lot, or between lots may be used. 2) An image obtained by selecting the same pattern (including inversion / rotation) from a plurality of points in the die shot wafer may be used. 3) An image obtained by selecting a plurality of the same part of the die shot from shot wafers having different processing conditions may be used. 4) An image saved in advance or an image created by simulation may be used. 5) The above combination may be used.

Grouping is 1) wafer position (radius, angle, X, Y, etc.), 2) die position in shot, 3) lithography scan direction, 4) processing conditions, 5) pattern direction, 6) wafer and lot, 7 ) The above combination may be used.

The output is 1) in-plane coordinates with large variations, 2) in-plane position that is highly correlated with processing conditions, etc. (in the case of multiple processing conditions, it is highly correlated with specific processing conditions and correlated with other processing conditions. In-plane position), 3) Large variation due to processing conditions (small variation when processing conditions are the same), 4) Die or shot index and die or shot coordinates in statistically out-of-range areas 5) Statistically the largest and smallest images within the assumed distribution, 6) Images with the specified cumulative probability, 7) Groups and grouping methods with large variability, 8) Statistical outlier groups, 9) Combinations of the above It may be.

(B18. An example of the output of the image processing device)
FIG. 12 is a diagram showing an example of the output of the image processing apparatus according to the present embodiment. In this figure, the probability plot is represented, the horizontal axis represents the statistic, and the vertical axis represents the cumulative probability.

The probability plot is such that when the statistic follows the assumed probability distribution, the plotted points ride on one straight line. Points that deviate from this straight line are abnormal points (Outlier Spots, indicated by dotted lines) that do not follow the assumed probability distribution, and are judged to be defects or the like. If the whole does not ride on a straight line, it indicates that the distribution of statistics does not follow the assumed probability distribution (the distribution is different).

The region where the statistic is maximum or minimum at the point on the straight line (Most Different Spots, indicated by the alternate long and short dash line) is considered to occur within the assumed variation distribution according to the assumed probability distribution. By confirming the shape of these regions with an electron microscope or the like and examining whether the shape is within the normal range, it is possible to efficiently determine whether or not the generated variation is allowed.

For a more accurate determination, several regions may be added according to the cumulative probability. For example, 1%, 50%, and 99% points are added as survey points.

[C. Action effect]
According to the present embodiment, it is possible to determine the defect or inspect whether or not the variation of the pattern is within the permissible range. That is, it is possible to provide an inspection technique in which the inspection target is not limited to the defects of the pattern. Further, by analyzing the in-plane distribution of the variation, it is possible to extract a region having a large variation. In addition, by grouping and analyzing between groups, it is possible to obtain a group on which the pattern variation depends and analyze the cause of the variation.

Although the embodiments of the present invention have been described above, two or more of these examples may be combined and carried out. Alternatively, one of these examples may be partially implemented.

Further, the present invention is not limited to the description of the embodiment of the above invention. Various modifications are also included in the present invention as long as those skilled in the art can easily conceive without departing from the description of the scope of claims.

100 Inspection device 190 Image processing device

Claims (12)

An inspection device that inspects an object
An acquisition unit that acquires a plurality of measurement data from measurement data at a predetermined position of a predetermined pattern of the object, and
A distribution analysis unit that analyzes the distribution between predetermined patterns, and
Inspection device equipped with. The inspection device according to claim 1, further comprising a pixel statistic calculation unit that calculates the statistic of the measurement data as a pixel statistic. The inspection device according to claim 1, wherein the distribution analysis unit outputs points deviating from a predetermined distribution. The inspection device according to claim 1, wherein the distribution analysis unit outputs a point having a predetermined cumulative probability. The distribution analysis unit outputs the inter-image distribution of the pixel statistic as an inter-image statistic.
The inspection device according to claim 1, further comprising an in-plane distribution analysis unit for analyzing the in-plane distribution of the inter-image statistics for the plurality of the predetermined patterns. The inspection device according to claim 5, wherein the in-plane distribution analysis unit extracts an in-plane region based on the inter-image statistics. The distribution analysis unit outputs the inter-image distribution of the pixel statistic as an inter-image statistic.
The inspection device according to claim 1, wherein a plurality of the objects are divided into groups, and the inter-image statistics are obtained for each group. The inspection device according to claim 7, further comprising an inter-group statistic calculation unit for calculating an inter-group statistic of the inter-image statistic obtained for each group. The inspection apparatus according to claim 8, further comprising an inter-group in-plane distribution statistic calculation unit for calculating the in-plane distribution statistic of the inter-group statistic. The inspection device according to claim 1, wherein the measurement data is an image. An inspection method that inspects an object
A plurality of measurement data are acquired from the measurement data of a predetermined position of a predetermined pattern of the object, and a plurality of measurement data are acquired.
An inspection method for analyzing the distribution between predetermined patterns. An inspection program that inspects an object
A step of acquiring a plurality of measurement data from the measurement data of a predetermined position of a predetermined pattern of the object, and
The step of analyzing the distribution between predetermined patterns and
A test program that lets your computer run.

Images