WO2024023955A1 - 測長システム、モデル作成システム及び測長方法 - Google Patents
測長システム、モデル作成システム及び測長方法 Download PDFInfo
- Publication number
- WO2024023955A1 WO2024023955A1 PCT/JP2022/028873 JP2022028873W WO2024023955A1 WO 2024023955 A1 WO2024023955 A1 WO 2024023955A1 JP 2022028873 W JP2022028873 W JP 2022028873W WO 2024023955 A1 WO2024023955 A1 WO 2024023955A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- image
- measurement
- unit
- learning
- length measurement
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Ceased
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
- G01B15/00—Measuring arrangements characterised by the use of electromagnetic waves or particle radiation, e.g. by the use of microwaves, X-rays, gamma rays or electrons
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N23/00—Investigating or analysing materials by the use of wave or particle radiation, e.g. X-rays or neutrons, not covered by groups G01N3/00 – G01N17/00, G01N21/00 or G01N22/00
- G01N23/22—Investigating or analysing materials by the use of wave or particle radiation, e.g. X-rays or neutrons, not covered by groups G01N3/00 – G01N17/00, G01N21/00 or G01N22/00 by measuring secondary emission from the material
- G01N23/225—Investigating or analysing materials by the use of wave or particle radiation, e.g. X-rays or neutrons, not covered by groups G01N3/00 – G01N17/00, G01N21/00 or G01N22/00 by measuring secondary emission from the material using electron or ion
- G01N23/2251—Investigating or analysing materials by the use of wave or particle radiation, e.g. X-rays or neutrons, not covered by groups G01N3/00 – G01N17/00, G01N21/00 or G01N22/00 by measuring secondary emission from the material using electron or ion using incident electron beams, e.g. scanning electron microscopy [SEM]
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
- G01B2210/00—Aspects not specifically covered by any group under G01B, e.g. of wheel alignment, caliper-like sensors
- G01B2210/56—Measuring geometric parameters of semiconductor structures, e.g. profile, critical dimensions or trench depth
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N2223/00—Investigating materials by wave or particle radiation
- G01N2223/40—Imaging
- G01N2223/401—Imaging image processing
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N2223/00—Investigating materials by wave or particle radiation
- G01N2223/40—Imaging
- G01N2223/418—Imaging electron microscope
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N2223/00—Investigating materials by wave or particle radiation
- G01N2223/60—Specific applications or type of materials
- G01N2223/611—Specific applications or type of materials patterned objects; electronic devices
- G01N2223/6116—Specific applications or type of materials patterned objects; electronic devices semiconductor wafer
Definitions
- the present disclosure relates to a length measurement system, a model creation system, and a length measurement method.
- Patent Document 1 discloses an image processing method that optimizes the brightness and gradation of a specific target area within an image. Therefore, a specific area (for example, a person's face) is detected, a tone curve for image enhancement is calculated using the histogram of the entire image and a histogram of the specific area, and contrast correction is performed using the calculated tone curve. conduct.
- a specific area for example, a person's face
- a tone curve for image enhancement is calculated using the histogram of the entire image and a histogram of the specific area
- contrast correction is performed using the calculated tone curve. conduct.
- Patent Document 2 discloses a learning model that can selectively convert partial images, such as specific patterns included in an image or edges of other structures, with high precision. By changing the degree of learning depending on the importance of each part of the image, it is possible to generate high-quality images for important parts, while reducing the processing required for learning for unimportant parts, improving learning efficiency. It is something.
- the left column shows photographed images (schematic diagrams).
- the photographed image 100 is, for example, a SEM image of a pattern formed on a semiconductor wafer.
- a second pattern that is not to be measured is provided on the semiconductor wafer in close proximity to the first pattern that is to be measured.
- the right column shows a brightness profile 111 along the line 103 on the photographed image 100.
- a brightness profile 112 obtained when the second pattern does not exist is also shown.
- the first pattern is an isolated pattern, the contrast with the surroundings is large as shown in the brightness profile 112, and it is expected that edge extraction from the first pattern image (measurement pattern image) 101 for measurement will be performed with high precision. can.
- the brightness is isolated in the boundary area between the first pattern and the second pattern due to the influence of the second pattern image (non-measurement pattern image) 102.
- the luminance of the first pattern image 101 does not decrease as in the case of the pattern, and is connected to the luminance of the second pattern image 102 without completely decreasing, resulting in a luminance distribution like the luminance profile 111.
- the accuracy of edge extraction from the first pattern image 101 for measurement decreases, and as a result, the measurement accuracy also decreases.
- Patent Document 1 If the technique disclosed in Patent Document 1 is applied to the photographed image shown in FIG. 1A, by making the first pattern image 101 obvious, the second pattern image 102 will also be made obvious, and the above-mentioned problem will occur. You can't solve problems. Further, even if conversion is performed with different accuracy between important and non-important parts according to Patent Document 2, the above problem cannot be solved because processing to increase the difference in contrast is not performed.
- the photographed image 100 is converted into a length measurement image 120 for dimension measurement, as schematically shown in FIG. 1B.
- the length measurement image 120 is an image in which the contrast of the first pattern image 121, which is a measurement pattern image, with respect to the second pattern image 122, which is a non-measurement pattern image, is improved compared to the original photographed image 100.
- the length measurement image 120 may be an image in which the second pattern image 122, which is a non-measurement pattern image, is deleted.
- the contrast of the second pattern image 122 and the background part of the length measurement image 120 (referring to the area other than the first pattern image 121 and the second pattern image 122 in the length measurement image 120) is It may also be a reduced image.
- a length measurement system that is an embodiment of the present disclosure is a length measurement system that measures the dimensions of a pattern formed on a sample, and includes a computer system that includes a length measurement image conversion section and a length measurement section.
- the length measurement image conversion unit converts the measurement target captured image of the sample into the measurement target length measurement image using the length measurement image conversion model, and the length measurement unit converts the measurement target image that is included in the measurement target length measurement image using the length measurement image conversion model.
- the length measurement image conversion model uses training data that is a combination of a learning image of the sample and a learning length measurement image corresponding to the learning image.
- the contrast between the measurement pattern image in the learning length measurement image and the non-measurement pattern image that exists near the measurement pattern image is the same as the contrast between the measurement pattern image and the non-measurement pattern image in the learning captured image.
- the contrast is improved or the non-measurement pattern image in the learning captured image is deleted.
- FIG. 2 is a diagram for explaining a problem of the present disclosure.
- FIG. 2 is a diagram for explaining a problem of the present disclosure.
- This is an example of the configuration of a length measurement system.
- This is an example of the configuration of a model creation system.
- FIG. 3 is a functional block diagram of an image generation model creation section. This is an example of a generative model confirmation screen.
- FIG. 3 is a functional block diagram of an area dividing section.
- FIG. 3 is a functional block diagram of a length measurement edge detection section.
- FIG. 3 is a diagram showing how edges are detected by applying interactive machine learning.
- FIG. 3 is a functional block diagram of an image conversion model creation section.
- FIG. 3 is a diagram for explaining an area specifying method. This is an example of a label image.
- FIG. 3 is a functional block diagram of a length measurement image conversion section.
- FIG. 2 is a diagram showing a configuration example of a length measurement system.
- the length measurement system shown in FIG. 2 is a system that performs length measurement processing of a pattern formed on a semiconductor wafer using a charged particle beam device such as a scanning electron microscope (SEM).
- a length measurement system acquires an image of a fine pattern formed on a semiconductor wafer and performs dimension measurement.
- the length measurement system includes an SEM 11, a control device 12 that controls the SEM 11, a computer system 1 that executes length measurement processing, an input device 13 that inputs necessary information to the computer system 1, an information input screen, and measurement results. It is equipped with a display 14 for displaying the following information.
- the computer system 1 includes an arithmetic processing unit 2 including one or more CPUs (Central Processing Units) and a storage device 3.
- the storage device 3 includes programs related to length measurement processing, other programs related to SEM control, and data used by these programs for processing, such as a length measurement image conversion model 31 and a length measurement recipe 32, which will be described later. Processing results of those programs are stored.
- the computer system 1 may be configured integrally with the control device 12.
- the length measurement process is executed by the arithmetic processing unit 2 executing a program related to the length measurement process. That is, the computer system 1 functions as a length measurement processing device. In other words, this program causes the computer system to function as a length measurement processing device. Programs and their functions executed by the computer system 1 and the like are referred to as "functions", "units", etc.
- the arithmetic processing section 2 includes a photographed image input section 21 which is an interface into which images photographed by the SEM 11 are input, an image conversion section 22 for length measurement that converts the photographed image inputted by the photographed image input section 21 into a length measurement image, and a length measurement image conversion section 22 that converts the photographed image inputted by the photographed image input section 21 into a length measurement image.
- the length measurement unit 23 functions as a length measurement unit 23 that measures dimensions of the length measurement image created by the length image conversion unit 22 according to the length measurement recipe 32 .
- a length measurement image conversion model 31 is used.
- FIG. 3 is a diagram showing a configuration example of a model creation system that creates a length measurement image conversion model 31 used by the length measurement system.
- the computer system 5 includes a photographed image input section 61 that is an interface through which photographed images are input, and an image generation model creation section 62 that creates an image generation model 30 using the photographed images inputted by the photographed image input section 61 as training data. , functions as an image conversion model creation section 63 that creates a length measurement image conversion model 31 using the image generation model 30 using the photographed image input by the photographed image input section 61 as training data.
- the image storage medium 15 stores captured images of a semiconductor wafer, which is a measurement target, captured by the SEM 11 of the length measurement system.
- the hardware configuration of the computer system 5 is the same as the hardware configuration of the computer system 1, and redundant explanation will be omitted.
- An input device 16 and a display 17 are connected to the computer system 5.
- FIG. 3 shows an example in which photographed images used as teacher data are stored in the image storage medium 15, the image storage medium 15 is connected to the computer system 1, and the photographed images stored in the image storage medium 15 are stored in the image storage medium 15. It is also possible to use a length measurement system that measures the dimensions of.
- the model creation system creates a length measurement image conversion model that converts a captured image as shown in FIG. 1A into a length measurement image as shown in FIG. 1B. For this reason, teacher data that is a combination of the captured image and the image for length measurement is created.
- FIG. 4A shows a functional block diagram of the image generation model creation section 62.
- the image generation model creation section 62 is a functional block that creates the image generation model 30.
- the image generation model 30 is a model that generates, from a label image indicating the shape of a pattern area, an image that has the shape shown in the label image and is likened to a real image (hereinafter sometimes referred to as a pseudo image). . As will be described later, the image generation model 30 is used in the process of creating a length measurement image to be used as training data.
- the image generation model creation section 62 includes a region division section 71, an image enhancement processing section 72, and a first learning section 73.
- the area dividing unit 71 divides the captured image into a measurement pattern area and other background areas.
- the captured image is divided into a measurement pattern area corresponding to the first pattern image (measurement pattern image) 101 and other background areas.
- region segmentation may be performed manually or by image processing, or any combination of these may be used, in which the user verifies and corrects the region segmentation performed by image processing. obtain.
- Known methods for region segmentation through image processing include methods using image features (luminance clustering using the k-means method) and methods using machine learning (supervised and unsupervised segmentation). These can be applied.
- a region division process using a length measurement recipe for measuring the dimensions of the measurement pattern image will be described.
- the length measurement recipe 32 is data representing the length measurement conditions of the measurement pattern image, including the shape of the measurement pattern (for example, line, hole, ellipse, circle, etc.), the length measurement cursor (for example, the coordinates of the upper left vertex, width, length, etc.). length, number, etc.), measurement algorithms, length measurement parameters, etc. are registered.
- shape of the measurement pattern for example, line, hole, ellipse, circle, etc.
- the length measurement cursor for example, the coordinates of the upper left vertex, width, length, etc.
- length, number, etc. measurement algorithms, length measurement parameters, etc. are registered.
- FIG. 5 shows a functional block diagram of the area dividing section 71 that performs area dividing processing using the length measurement recipe 32.
- the region dividing section 71 includes a length measurement edge detecting section 74, a region specifying section 75, and a region extracting section 76.
- Schematic diagrams 74a, 75a, and 76a are schematic diagrams showing processing results in each of these functional blocks.
- the length measurement edge detection unit 74 uses the length measurement recipe 32 to detect an edge 74c of the photographed image.
- an edge 74c is displayed as a white circle.
- the edge 74c can be extracted as a change point in the brightness profile along the line 74b.
- FIG. 5 shows a functional block diagram of the area dividing section 71 that performs area dividing processing using the length measurement recipe 32.
- the region dividing section 71 includes a length measurement edge detecting section 74, a region specifying section 75, and a region extracting section 76.
- the area identifying unit 75 identifies a pattern boundary 75b by interpolating and connecting adjacent edges 74c, as shown in a processing result 75a.
- the area extracting unit 76 specifies a measurement pattern area based on the pattern boundary 75b specified by the area specifying unit 75.
- a region 76b corresponding to the region surrounded by the pattern boundary 75b is the measurement pattern region.
- a label image indicating a measurement pattern area corresponding to the measurement pattern image included in the photographed image is obtained. Note that when a plurality of measurement pattern areas exist in the photographed image, it is preferable to extract the measurement pattern area for each measurement pattern area.
- FIG. 6A shows a functional block diagram of the length measurement edge detection unit 74 that detects edges from the captured image.
- the length measurement edge detection unit 74 includes a length measurement edge learning unit 77 and a length measurement edge inference unit 78, and enables accurate edge detection using an interactive machine learning method.
- the length measurement edge learning unit 77 uses the teacher data to learn a learning model for inferring edges, and the length measurement edge inference unit 78 uses the learning model learned by the length measurement edge learning unit 77 to perform edge inference.
- FIG. 6B shows how the length measurement edge detection unit 74 applies interactive machine learning to detect edges.
- a white frame 79 in the figure is a length measurement cursor, and a change point in the brightness profile in this area is detected as an edge. Details of the detection process are defined in the length measurement recipe 32.
- white circles are edges that are teacher data
- white triangles are edges that are inferred using the learning model.
- the first teacher data 77a is provided by the user.
- the inference edge 78a is the result of inference using the learning model that learned the first teacher data 77a.
- the inference edge 78a includes inference results located on the pattern boundary and inference results deviated from the pattern boundary. Therefore, the user deletes the inference results that deviate from the pattern boundaries, and sets the inference results located on the pattern boundaries and the initially given teacher data as new teacher data 77b.
- an inference edge 78b is obtained as a result of inference using a learning model that has learned the teacher data 77b. This operation is repeatedly performed, and finally, when the result of inference using the learning model converges on the pattern boundary (in this case, the third inference edge 78c), the learning of the learning model is terminated.
- the region dividing unit 71 can generate a label image indicating the region corresponding to the measurement pattern image from the photographed image.
- the first learning unit 73 performs learning of the image generation model 30 using the combination of the label image indicating the measurement pattern area generated by the area dividing unit 71 and the measurement pattern image of the photographed image as training data.
- the error between the pseudo measurement pattern image generated by the image generation model 30 and the measurement pattern image of the original photographed image converges to a certain value or less, the learning of the image generation model 30 is finished.
- the shape of the pseudo measurement pattern image generated by the image generation model 30 follows the shape of the pattern area of the label image, while the image corresponds to the original captured image.
- an image may be output in which the brightness of the generated pseudo measurement pattern image is improved compared to the original photographed image.
- the teacher data used by the first learning section 73 an image obtained by performing image processing on the photographed image so that the measurement pattern area is emphasized by the image enhancement processing section 72 may be used.
- the first learning unit 73 learns a learning model that outputs a label image indicating the area corresponding to the measurement pattern image from the photographed image, and the image generation model 30 uses the learning model that outputs the label image.
- a pseudo image may be output based on the label output for the photographed image and the brightness of pixels at locations corresponding to the output label in the photographed image. Furthermore, in the first learning section 73, for reasons such as consolidating the number of image generation models 30 stored in the storage device 7, a plurality of types of measurement pattern images may be learned using one image generation model. . For example, in the example of FIG. 1A, the first pattern image 101 and the second pattern image 102 may be learned together using one image generation model 30. The image generation model 30 that has been trained is stored in the storage device 7.
- the image generation model creation unit 62 preferably includes a GUI (Graphical User Interface) for checking the performance of the created image generation model 30.
- FIG. 4B shows an example of the generated model confirmation screen 130.
- the label image/reference image load button 131 on the generation model confirmation screen 130 the captured image and the label image created by dividing the captured image into regions are loaded and are used as the reference image 135 and label image 134, respectively. Display.
- two types of measurement pattern images exist in the reference image 135.
- the AI model load button 132 the corresponding image generation model is loaded. For example, a first image generation model for the first measurement pattern area 137 and a second image generation model for the second measurement pattern area 138 are loaded.
- the pseudo measurement pattern image 137b is generated from the label image indicating the first measurement pattern area 137 using the first image generation model, and the pseudo measurement pattern image 137b is generated using the second image generation model.
- a pseudo measurement pattern image 138b is generated from the label image indicating the second measurement pattern area 138 using the label image.
- the generated pseudo image 136 is displayed alongside the reference image 135.
- FIG. 7 shows a functional block diagram of the image conversion model creation section 63.
- the image conversion model creation unit 63 is a functional block that creates the image conversion model 31 for length measurement.
- the image conversion model creating section 63 includes a region specifying section 81 , a region dividing section 82 , an image generating section 83 , and a second learning section 84 .
- the region specifying section 81, the region dividing section 82, and the image generating section 83 create a length measurement image that becomes learning data for the length measurement image conversion model 31.
- the area specifying unit 81 specifies an area including the measurement pattern image from the photographed image.
- the method of specifying the area is arbitrary.
- the user may manually specify the area using the specification box 81a shown in FIG.
- the area may be specified using the coordinate information of the length measurement cursor defined in the length measurement recipe.
- the area may be specified by expanding the layout of the corresponding pattern using a layout design drawing of the device.
- FIG. 10A shows an example of a GUI for specifying an area.
- a captured image to be subjected to area specification processing is selected from the image load button 91 on the area specification screen 90 and displayed as a captured image 94.
- a designation box 95 indicating the area to be designated is displayed on the photographed image 94, and the coordinates and size of the designation box 95 are determined.
- the user presses the manual specification button 92 for manual specification, and the automatic specification button 93 for automatic specification based on the length measurement recipe or other information.
- a manual designation screen 96 as shown in FIG. 10B is displayed.
- the shape, coordinates, and size of the specification box 95 can be specified.
- an automatic designation screen 97 as shown in FIG. 10C is displayed. For example, if multiple automatic designation methods are available, the user can select the automatic designation method to apply.
- the area specification on the area specification screen 90 is not limited to the rectangular shape shown as the specification box 95, but may take other shapes such as an ellipse, or may take other forms of specification such as the center coordinates of the area. .
- the area specification screen 90 includes a GUI according to the area specification form.
- a label image 134 (see FIG. 4B) is obtained from the photographed image and displayed to the user, and the user can select, for example, the first measurement pattern in the label image 134 by operating the GUI.
- the area 137 may also be specified.
- the region dividing unit 82 divides the region in the photographed image designated as such into a measurement pattern region and other regions.
- the processing of the region dividing section 82 is similar to the processing of the region dividing section 71 described in FIG. 4A, so a redundant explanation will be omitted.
- the captured image may be directly divided into regions without specifying the region by the region specifying unit 81.
- a label image 85 shown in FIG. 9A is obtained.
- the label image 85 is an image obtained by region-dividing the captured image 100 shown in FIG. 8, and includes two measurement pattern regions 85a.
- the image generation unit 83 uses the label image 85 generated by the area division unit 82 and the image generation model 30 to generate a pseudo captured image corresponding to the length measurement image corresponding to the captured image.
- FIG. 9B shows an example of the length measurement image 86 corresponding to the photographed image 100, which is generated from the label image 85.
- the measurement pattern image 86a is a pseudo measurement pattern image generated by the image generation model 30 for the measurement pattern region 85a, and for example, the background 86b is a uniform monochrome image.
- the present invention is not limited to this, and the brightness of an area other than the measurement pattern area 85a of the original captured image 100 may be reduced and combined with the pseudo measurement pattern image. That is, it is sufficient that the contrast of the pseudo measurement pattern image in the length measurement image 86 is improved compared to the contrast of the measurement pattern image 101 in the original captured image 100.
- the image generation unit 83 performs an adjustment to reduce the brightness of the area other than the measurement pattern area 85a of the original captured image 100, and to increase the brightness within the measurement pattern area 85a. may be synthesized.
- the contrast of the measurement pattern image 86a in the length measurement image 86 is significantly improved over the contrast of the measurement pattern image 101 in the original captured image 100.
- Adjustments to increase the brightness within the measurement pattern area 85a may be made so that the user can use a GUI such as the area designation screen 90 to perform operations such as how much to increase the brightness.
- the second learning unit 84 trains the length measurement image conversion model 31 using the combination of the photographed image and the length measurement image generated by the image generation unit 83 as training data.
- the error between the length measurement image generated by the length measurement image conversion model 31 and the length measurement image generated by the image generation unit 83 converges to a certain level or less, the learning of the length measurement image conversion model 31 is completed.
- FIG. 11 shows an example of the conversion model confirmation screen 140.
- the captured image load button 141 on the conversion model confirmation screen 140 the captured image is loaded and displayed as an input image 144.
- the input image 144 includes a measurement pattern image 146 and a non-measurement pattern image 147 adjacent thereto.
- the AI model load button 142 the length measurement image conversion model is loaded.
- the execution button 143 a length measurement image 145 obtained by converting the input image 144 using the length measurement image conversion model is displayed. By comparing the input image 144 and the converted length measurement image 145, the user can confirm that an appropriate length measurement image has been generated.
- the storage device 3 of the computer system 1 of the length measurement system stores the trained length measurement image conversion model 31 and length measurement recipe 32 (see FIG. 2).
- the control device 12 uses the information of the length measurement recipe 32 read out from the computer system 1 to obtain an image of, for example, a semiconductor wafer using the SEM 11 .
- a photographed image taken by the SEM 11 is input into the computer system 1 from the control device 12 by the photographed image input section 21 .
- FIG. 12 shows a functional block diagram of the length measurement image conversion section 22.
- the length measurement image conversion section 22 includes an image conversion section 150.
- the image conversion unit 150 uses the length measurement image conversion model 31 to convert a captured image as shown in FIG. 1A into a length measurement image as shown in FIG. 1B.
- the length measurement unit 23 performs dimension measurement on the length measurement image converted by the length measurement image conversion unit 22 in accordance with the content defined in the length measurement recipe 32, so that the size of the image in the vicinity of the measurement pattern image is measured. Even if there is a non-measurement pattern image that obstructs the length measurement, the contrast of the measurement pattern image with respect to the non-measurement pattern image is improved in the length measurement image compared to the original photographed image. , erroneous measurements can be suppressed.
- a length measurement image conversion model that has undergone machine learning for length measurement images, high-speed and automatic image conversion is possible.
- the present disclosure is not limited to the embodiments described above, and includes various modifications.
- the embodiments described above are described in detail to explain the present disclosure in an easy-to-understand manner, and the embodiments are not necessarily limited to those having all the configurations described.
- it is possible to replace a part of the configuration of one embodiment with the configuration of another embodiment and it is also possible to add the configuration of another embodiment to the configuration of one embodiment.
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Electromagnetism (AREA)
- Computational Linguistics (AREA)
- Biomedical Technology (AREA)
- Data Mining & Analysis (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Biophysics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Artificial Intelligence (AREA)
- Image Analysis (AREA)
- Length-Measuring Devices Using Wave Or Particle Radiation (AREA)
- Image Processing (AREA)
Abstract
Description
Claims (19)
- 試料上に形成されたパターンの寸法計測を行う測長システムであって、
測長用画像変換部と測長部とを備えるコンピュータシステムを有し、
前記測長用画像変換部は、前記試料を撮影した計測対象撮影画像を、測長用画像変換モデルにより計測対象測長用画像に変換し、
前記測長部は、前記計測対象測長用画像に含まれる計測パターン像の寸法を計測し、
前記測長用画像変換モデルは、前記試料を撮影した学習用撮影画像と当該学習用撮影画像に対応する学習用測長用画像との組み合わせである教師データを用いて学習されており、
前記学習用測長用画像における前記計測パターン像と前記計測パターン像に近接して存在する非計測パターン像とのコントラストは、前記学習用撮影画像における前記計測パターン像と前記非計測パターン像とのコントラストよりも向上されている、または前記学習用測長用画像では、前記学習用撮影画像における前記非計測パターン像が削除されていることを特徴とする測長システム。 - 請求項1において、
前記試料を撮影する荷電粒子線装置を備え、
前記コンピュータシステムは、前記荷電粒子線装置が撮影した撮影画像を取り込む撮影画像入力部を備えることを特徴とする測長システム。 - パターンが形成された試料を撮影した撮影画像を測長用画像に変換する測長用画像変換モデルを作成するモデル作成システムであって、
画像変換モデル作成部を備えるコンピュータシステムを有し、
前記画像変換モデル作成部は領域分割部、画像生成部及び学習部を備え、
前記画像変換モデル作成部の前記領域分割部は、前記撮影画像に含まれる計測パターン像の領域である計測パターン領域と前記計測パターン領域以外の背景領域とに分割したラベル画像を生成し、
前記画像変換モデル作成部の前記画像生成部は、前記ラベル画像の前記計測パターン領域に擬似計測パターン像を生成した擬似撮影画像を作成し、
前記画像変換モデル作成部の前記学習部は、前記撮影画像と前記擬似撮影画像との組み合わせである教師データを用いて、前記測長用画像変換モデルの学習を行い、
前記擬似撮影画像における前記擬似計測パターン像と、前記背景領域に存在し、前記擬似計測パターン像に近接して存在する非計測パターン像とのコントラストは、前記撮影画像における前記計測パターン像と前記非計測パターン像とのコントラストよりも向上されている、または前記擬似撮影画像では、前記撮影画像における前記非計測パターン像が削除されていることを特徴とするモデル作成システム。 - 請求項3において、
前記画像変換モデル作成部の前記領域分割部は、測長エッジ検出部、領域特定部及び領域抽出部を備え、
前記測長エッジ検出部は、前記計測パターン像のエッジを検出し、
前記領域特定部は、検出された前記エッジを補間して前記計測パターン像のパターン境界を特定し、
前記領域抽出部は、特定された前記パターン境界に囲まれた領域を前記計測パターン領域として抽出することを特徴とするモデル作成システム。 - 請求項4において、
前記コンピュータシステムは、前記計測パターン像の測長条件を表すデータである測長レシピを記憶しており、
前記測長エッジ検出部は測長エッジ学習部及び測長エッジ推論部を備え、インタラクティブ機械学習により前記測長レシピに定義された測長カーソル内における前記計測パターン像のエッジを検出し、
前記測長エッジ学習部は、前記計測パターン像のエッジを示す教師データにより学習モデルの学習を行い、
前記測長エッジ推論部は、前記学習モデルにより前記測長カーソル内における前記計測パターン像のエッジの推論を行い、
前記学習モデルにより推論された前記計測パターン像のエッジのうち、前記計測パターン像の境界から乖離した前記計測パターン像のエッジを削除して新たな教師データとして、前記測長エッジ学習部による前記学習モデルの学習を行うことを特徴とするモデル作成システム。 - 請求項3において、
前記画像変換モデル作成部はさらに前記撮影画像に含まれる前記計測パターン像が存在する領域を指定する領域指定部を備えることを特徴とするモデル作成システム。 - 請求項3において、
前記コンピュータシステムはさらに画像生成モデル作成部を備え、
前記画像変換モデル作成部の前記画像生成部は、前記画像生成モデル作成部が作成した画像生成モデルにより、前記ラベル画像の前記計測パターン領域に前記擬似計測パターン像を生成することを特徴とするモデル作成システム。 - 請求項7において、
前記画像生成モデル作成部は、領域分割部及び学習部を備え、
前記画像生成モデル作成部の前記領域分割部は、前記撮影画像に含まれる前記計測パターン像の領域である計測パターン領域と前記計測パターン領域以外の背景領域とに分割したラベル画像を生成し、
前記画像生成モデル作成部の前記学習部は、前記ラベル画像と前記撮影画像または前記撮影画像に前記計測パターン像を強調する画像処理を行った画像との組み合わせである教師データを用いて、前記画像生成モデルの学習を行うことを特徴とするモデル作成システム。 - 請求項8において、
前記画像生成モデル作成部の前記領域分割部は、測長エッジ検出部、領域特定部及び領域抽出部を備え、
前記測長エッジ検出部は、前記計測パターン像のエッジを検出し、
前記領域特定部は、検出された前記エッジを補間して前記計測パターン像のパターン境界を特定し、
前記領域抽出部は、特定された前記パターン境界に囲まれた領域を前記計測パターン領域として抽出することを特徴とするモデル作成システム。 - 請求項9において、
前記コンピュータシステムは、前記計測パターン像の測長条件を表すデータである測長レシピを記憶しており、
前記測長エッジ検出部は測長エッジ学習部及び測長エッジ推論部を備え、インタラクティブ機械学習により前記測長レシピに定義された測長カーソル内における前記計測パターン像のエッジを検出し、
前記測長エッジ学習部は、前記計測パターン像のエッジを示す教師データにより学習モデルの学習を行い、
前記測長エッジ推論部は、前記学習モデルにより前記測長カーソル内における前記計測パターン像のエッジの推論を行い、
前記学習モデルにより推論された前記計測パターン像のエッジのうち、前記計測パターン像の境界から乖離した前記計測パターン像のエッジを削除して新たな教師データとして、前記測長エッジ学習部による前記学習モデルの学習を行うことを特徴とするモデル作成システム。 - コンピュータシステムを備えた測長システムにより、試料上に形成されたパターンの寸法計測を行う測長方法であって、
前記コンピュータシステムは、測長用画像変換部と測長部とを備え、
前記測長用画像変換部は、前記試料を撮影した計測対象撮影画像を、測長用画像変換モデルにより計測対象測長用画像に変換し、
前記測長部は、前記計測対象測長用画像に含まれる計測パターン像の寸法を計測し、
前記測長用画像変換モデルは、前記試料を撮影した学習用撮影画像と当該学習用撮影画像に対応する学習用測長用画像との組み合わせである教師データを用いて学習されており、
前記学習用測長用画像における前記計測パターン像と前記計測パターン像に近接して存在する非計測パターン像とのコントラストは、前記学習用撮影画像における前記計測パターン像と前記非計測パターン像とのコントラストよりも向上されている、または前記学習用測長用画像では、前記学習用撮影画像における前記非計測パターン像が削除されていることを特徴とする測長方法。 - 請求項11において、
前記コンピュータシステムはさらに前記測長用画像変換モデルを作成する画像変換モデル作成部を備え、
前記画像変換モデル作成部は領域分割部、画像生成部及び学習部を備え、
前記画像変換モデル作成部の前記領域分割部は、前記学習用撮影画像に含まれる前記計測パターン像の領域である計測パターン領域と前記計測パターン領域以外の背景領域とに分割したラベル画像を生成し、
前記画像変換モデル作成部の前記画像生成部は、前記学習用測長用画像として、前記ラベル画像の前記計測パターン領域に擬似計測パターン像を生成した擬似撮影画像を作成し、
前記画像変換モデル作成部の前記学習部は、前記学習用撮影画像と前記擬似撮影画像との組み合わせである教師データを用いて、前記測長用画像変換モデルの学習を行うことを特徴とする測長方法。 - 請求項12において、
前記画像変換モデル作成部の前記領域分割部は、測長エッジ検出部、領域特定部及び領域抽出部を備え、
前記測長エッジ検出部は、前記計測パターン像のエッジを検出し、
前記領域特定部は、検出された前記エッジを補間して前記計測パターン像のパターン境界を特定し、
前記領域抽出部は、特定された前記パターン境界に囲まれた領域を前記計測パターン領域として抽出することを特徴とする測長方法。 - 請求項13において、
前記コンピュータシステムは、前記計測パターン像の測長条件を表すデータである測長レシピを記憶しており、
前記測長エッジ検出部は測長エッジ学習部及び測長エッジ推論部を備え、インタラクティブ機械学習により前記測長レシピに定義された測長カーソル内における前記計測パターン像のエッジを検出し、
前記測長エッジ学習部は、前記計測パターン像のエッジを示す教師データにより学習モデルの学習を行い、
前記測長エッジ推論部は、前記学習モデルにより前記測長カーソル内における前記計測パターン像のエッジの推論を行い、
前記学習モデルにより推論された前記計測パターン像のエッジのうち、前記計測パターン像の境界から乖離した前記計測パターン像のエッジを削除して新たな教師データとして、前記測長エッジ学習部による前記学習モデルの学習を行うことを特徴とする測長方法。 - 請求項12において、
前記画像変換モデル作成部はさらに前記学習用撮影画像に含まれる前記計測パターン像が存在する領域を指定する領域指定部を備えることを特徴とする測長方法。 - 請求項12において、
前記コンピュータシステムはさらに画像生成モデル作成部を備え、
前記画像変換モデル作成部の前記画像生成部は、前記画像生成モデル作成部が作成した画像生成モデルにより、前記ラベル画像の前記計測パターン領域に擬似計測パターン像を生成することを特徴とする測長方法。 - 請求項16において、
前記画像生成モデル作成部は、領域分割部及び学習部を備え、
前記画像生成モデル作成部の前記領域分割部は、前記学習用撮影画像に含まれる前記計測パターン像の領域である計測パターン領域と前記計測パターン領域以外の背景領域とに分割したラベル画像を生成し、
前記画像生成モデル作成部の前記学習部は、前記ラベル画像と前記学習用撮影画像または前記学習用撮影画像に前記計測パターン像を強調する画像処理を行った画像との組み合わせである教師データを用いて、前記画像生成モデルの学習を行うことを特徴とする測長方法。 - 請求項17において、
前記画像生成モデル作成部の前記領域分割部は、測長エッジ検出部、領域特定部及び領域抽出部を備え、
前記測長エッジ検出部は、前記計測パターン像のエッジを検出し、
前記領域特定部は、検出された前記エッジを補間して前記計測パターン像のパターン境界を特定し、
前記領域抽出部は、特定された前記パターン境界に囲まれた領域を前記計測パターン領域として抽出することを特徴とする測長方法。 - 請求項18において、
前記コンピュータシステムは、前記計測パターン像の測長条件を表すデータである測長レシピを記憶しており、
前記測長エッジ検出部は測長エッジ学習部及び測長エッジ推論部を備え、インタラクティブ機械学習により前記測長レシピに定義された測長カーソル内における前記計測パターン像のエッジを検出し、
前記測長エッジ学習部は、前記計測パターン像のエッジを示す教師データにより学習モデルの学習を行い、
前記測長エッジ推論部は、前記学習モデルにより前記測長カーソル内における前記計測パターン像のエッジの推論を行い、
前記学習モデルにより推論された前記計測パターン像のエッジのうち、前記計測パターン像の境界から乖離した前記計測パターン像のエッジを削除して新たな教師データとして、前記測長エッジ学習部による前記学習モデルの学習を行うことを特徴とする測長方法。
Priority Applications (4)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202280097740.4A CN119546923A (zh) | 2022-07-27 | 2022-07-27 | 测长系统、模型制作系统以及测长方法 |
| KR1020247042960A KR20250016301A (ko) | 2022-07-27 | 2022-07-27 | 측장 시스템, 모델 제작 시스템 및 측장 방법 |
| PCT/JP2022/028873 WO2024023955A1 (ja) | 2022-07-27 | 2022-07-27 | 測長システム、モデル作成システム及び測長方法 |
| TW112126663A TWI868806B (zh) | 2022-07-27 | 2023-07-18 | 測長系統、模型作成系統及測長方法 |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| PCT/JP2022/028873 WO2024023955A1 (ja) | 2022-07-27 | 2022-07-27 | 測長システム、モデル作成システム及び測長方法 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2024023955A1 true WO2024023955A1 (ja) | 2024-02-01 |
Family
ID=89705652
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/JP2022/028873 Ceased WO2024023955A1 (ja) | 2022-07-27 | 2022-07-27 | 測長システム、モデル作成システム及び測長方法 |
Country Status (4)
| Country | Link |
|---|---|
| KR (1) | KR20250016301A (ja) |
| CN (1) | CN119546923A (ja) |
| TW (1) | TWI868806B (ja) |
| WO (1) | WO2024023955A1 (ja) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH0721372A (ja) * | 1993-06-15 | 1995-01-24 | Nikon Corp | 画像検出装置 |
| JP2011186678A (ja) * | 2010-03-05 | 2011-09-22 | Toshiba Teli Corp | 画像処理装置および画像処理プログラム |
| JP2020186959A (ja) * | 2019-05-13 | 2020-11-19 | 株式会社日立ハイテク | パターン評価システム及びパターン評価方法 |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2009200743A (ja) | 2008-02-20 | 2009-09-03 | Ricoh Co Ltd | 画像処理装置および画像処理方法および画像処理プログラムおよび撮像装置 |
| TWM489335U (en) * | 2011-08-22 | 2014-11-01 | Jiun-Yin Huang | Image identification system |
| KR102888847B1 (ko) | 2020-01-09 | 2025-11-19 | 주식회사 히타치하이테크 | 화상을 생성하는 시스템, 및 비일시적 컴퓨터 가독 매체 |
| JP2023056056A (ja) * | 2020-03-05 | 2023-04-19 | ソニーセミコンダクタソリューションズ株式会社 | データ生成方法、学習方法及び推定方法 |
| WO2022110877A1 (zh) * | 2020-11-24 | 2022-06-02 | 深圳市商汤科技有限公司 | 深度检测方法、装置、电子设备、存储介质及程序 |
| TW202228066A (zh) * | 2021-01-06 | 2022-07-16 | 富比庫股份有限公司 | 圖像物件分類方法、系統、電腦程式產品及電腦可讀取紀錄媒體 |
-
2022
- 2022-07-27 WO PCT/JP2022/028873 patent/WO2024023955A1/ja not_active Ceased
- 2022-07-27 KR KR1020247042960A patent/KR20250016301A/ko active Pending
- 2022-07-27 CN CN202280097740.4A patent/CN119546923A/zh active Pending
-
2023
- 2023-07-18 TW TW112126663A patent/TWI868806B/zh active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH0721372A (ja) * | 1993-06-15 | 1995-01-24 | Nikon Corp | 画像検出装置 |
| JP2011186678A (ja) * | 2010-03-05 | 2011-09-22 | Toshiba Teli Corp | 画像処理装置および画像処理プログラム |
| JP2020186959A (ja) * | 2019-05-13 | 2020-11-19 | 株式会社日立ハイテク | パターン評価システム及びパターン評価方法 |
Also Published As
| Publication number | Publication date |
|---|---|
| TWI868806B (zh) | 2025-01-01 |
| CN119546923A (zh) | 2025-02-28 |
| KR20250016301A (ko) | 2025-02-03 |
| TW202422006A (zh) | 2024-06-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| TWI767108B (zh) | 用於檢查半導體試樣的方法與系統及在其上紀錄相關指令的電腦可讀取媒體 | |
| US11663713B2 (en) | Image generation system | |
| TWI766303B (zh) | 計測系統、產生在進行包含有特定之構造的半導體之畫像計測時所使用的學習模型之方法、以及儲存有用以使電腦實行產生在進行包含有特定之構造的半導體之畫像計測時所使用的學習模型之處理之程式的記錄媒體 | |
| US8655050B2 (en) | Pattern generating apparatus and pattern shape evaluating apparatus | |
| JP4585926B2 (ja) | パターンレイヤーデータ生成装置、それを用いたパターンレイヤーデータ生成システム、半導体パターン表示装置、パターンレイヤーデータ生成方法、及びコンピュータプログラム | |
| TW202142852A (zh) | 半導體試樣的基於深度學習之檢查的方法及其系統 | |
| JPWO2020121564A1 (ja) | 寸法計測装置、寸法計測プログラム及び半導体製造システム | |
| JPWO2021024402A1 (ja) | 寸法計測装置、寸法計測方法及び半導体製造システム | |
| JP2022139174A (ja) | 欠陥分類装置、方法およびプログラム | |
| CN112419208A (zh) | 一种基于施工图审查矢量图纸编译方法及系统 | |
| WO2020121739A1 (ja) | 画像マッチング方法、および画像マッチング処理を実行するための演算システム | |
| JP2022114331A (ja) | 検査システム、検査方法及び検査プログラム | |
| KR101402575B1 (ko) | 직접묘화장치용 화상표시장치 및 기록매체 | |
| WO2024023955A1 (ja) | 測長システム、モデル作成システム及び測長方法 | |
| CN118052992B (zh) | 一种图像识别模型的生成方法、系统、芯片及存储器 | |
| JP7392166B2 (ja) | 画像生成装置、画像生成方法およびプログラム | |
| WO2023248355A1 (ja) | 不適合検出装置、および、不適合検出方法 | |
| JP7451384B2 (ja) | パターンエッジ検出方法、パターンエッジ検出装置、パターンエッジ検出をコンピュータに実行させるためのプログラムが記録された記録媒体 | |
| CN114945801A (zh) | 图案边缘检测方法、图案边缘检测装置、记录有用于让计算机执行图案边缘检测的程序的记录介质 | |
| CN120375375B (zh) | 一种智能数据标注方法、系统、电子设备及存储介质 | |
| CN119359882B (zh) | 三维人物模型的纹理生成方法、系统及存储介质 | |
| JPH10320556A (ja) | 電子画像の領域判定方法、電子画像の領域判定システム、電子画像の領域判定システムを備えたsem装置、及び電子画像の領域判定プログラムを格納したコンピュータ読み取り可能な記録媒体 | |
| CN120411002A (zh) | 一种芯片外观缺陷检测方法 | |
| JP2020123064A (ja) | 画像マッチング判定方法、画像マッチング判定装置、および画像マッチング判定方法をコンピュータに実行させるためのプログラムを記録したコンピュータ読み取り可能な記録媒体 | |
| WO2021215006A1 (ja) | 解析装置及び解析プログラム |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| 121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 22953050 Country of ref document: EP Kind code of ref document: A1 |
|
| ENP | Entry into the national phase |
Ref document number: 20247042960 Country of ref document: KR Kind code of ref document: A |
|
| WWE | Wipo information: entry into national phase |
Ref document number: 1020247042960 Country of ref document: KR |
|
| WWE | Wipo information: entry into national phase |
Ref document number: 202280097740.4 Country of ref document: CN |
|
| WWP | Wipo information: published in national office |
Ref document number: 1020247042960 Country of ref document: KR |
|
| NENP | Non-entry into the national phase |
Ref country code: DE |
|
| WWP | Wipo information: published in national office |
Ref document number: 202280097740.4 Country of ref document: CN |
|
| 122 | Ep: pct application non-entry in european phase |
Ref document number: 22953050 Country of ref document: EP Kind code of ref document: A1 |
|
| NENP | Non-entry into the national phase |
Ref country code: JP |