Charged particle beam device

Abstract

Provided is a charged particle beam device with which it is possible to prevent image quality deterioration associated with a captured image. A charged particle beam device includes an imaging device that irradiates a sample with a charged particle beam and images information of the sample, and a computer. A computer stores each image (scanned image) obtained by scanning the same region a plurality of times, classifies each image into an image including a degraded image and an image not including the degraded image, and stores a target image obtained by image integration from the image not including the degraded image. A charged particle beam device is provided with a database that stores data such as information obtained from an imaging device that includes a scanned image, a classification result, and a target image. The charged particle beam device obtains a target image having high S / N and sharpness in which random noise is suppressed and image quality degradation is prevented by performing image integration on images that do not include the selected degraded image among the scanned images.

Description

[0001] This application is a divisional application of a patent application with application number 201980095455.7 and invention name “Charged Particle Beam Device”, which entered the national stage on October 15, 2021. Technical Field

[0002] The present invention relates to the technology of charged particle beam device and photography technology. Background Art

[0003] Charged particle beam devices include scanning electron microscopes (SEMs), focused ion beam (FIBs), and the like. Charged particle beam devices observe and analyze samples by irradiating a sample with a charged particle beam. For example, a scanning electron microscope is a device that uses an electron beam as a charged particle beam and images detection signals such as secondary electrons and reflected electrons generated by scanning a narrowed electron beam (probe) on a sample.

[0004] The detection signal used in the imaging of the above-mentioned charged particle beam device (for example, a signal after collecting and amplifying secondary electrons released from the irradiated surface of the sample) has more noise components due to its characteristics. Therefore, it is known that there is a technology for accumulating multiple images of the same area to suppress irregularly generated noise (random noise). As an example of the prior art, patent document 1 can be cited. The following is recorded in patent document 1: In order to eliminate the sense of disharmony in the accumulation process of suppressing random noise, the shading value of the accumulated image is normalized, and the normalized accumulation is performed to maintain the brightness of the observed field of view.

[0005] Prior art literature

[0006] Patent Literature

[0007] Patent Document 1: International Publication No. 2017 / 090204

[0008] Patent Document 2: Japanese Patent No. 4301385 Summary of the invention

[0009] Problems to be solved by the invention

[0010] Usually, in the shooting of SEM (recording images on a recording medium), each image (sometimes recorded as a scanned image) obtained by scanning the same area multiple times is accumulated, thereby obtaining a target image with suppressed random noise. The time of these multiple scans helps to detect the signal amount, so the dynamic range is wide, requiring tens of milliseconds to hundreds of seconds. During the generation of the accumulated image, if the SEM and the sample as the object cannot maintain a stable state with high precision, the sharpness in the target image will deteriorate.

[0011] However, both the SEM and the sample being the object are subject to variation due to the sample's state transition caused by beam irradiation and the interference in the installation environment. For example, in the case of sample state transition caused by beam irradiation, electrons and gas molecules aggregate due to charging (charge) and contamination, hindering beam irradiation. In the case of interference in the installation environment, interference (vibration of the installation environment including sound waves and electromagnetic waves) is transmitted to the lens barrel and the workbench, causing the position of the probe scanning on the sample to shift. In addition, due to thermal drift and workbench drift, the field of view position moves with each scan. Due to the reduction in probe current, the signal amount of secondary electrons decreases. Due to the reduction in vacuum, the probe collides with the gas present in the atmosphere and scatters. The various phenomena mentioned above may occur.

[0012] In order to obtain a target image with good image quality, the user as an operator searches for imaging conditions of the charged particle beam device that are less likely to cause image quality degradation by trial and error. This trial and error may require more operation time than expected, causing damage to the sample and reducing the throughput.

[0013] Although the accumulated image obtained by accumulating the scanned images including the images affected by these phenomena suppresses random noise, it produces image quality degradation in terms of sharpness. Patent document 2 describes calculating the positions that match between a certain image and the next image obtained sequentially from the same field of view, and accumulating at the calculated positions. This technology is effective in preventing the image quality degradation of the accumulated image caused by the drift of each image. However, this technology is not effective in preventing the image quality degradation of the accumulated image caused by the slight drift generated in one image.

[0014] Hereinafter, such a cause of degradation of the image quality of the accumulated image is referred to as an image quality degradation cause. In addition, an image affected by the image quality degradation cause is referred to as a degraded image.

[0015] The object of the present invention is to suppress random noise by accumulation for a target image captured by a charged particle beam device, and to prevent image quality degradation related to sharpness caused by image quality degradation. In addition, the object of the present invention is to capture a target image with high S / N and sharpness. The above-mentioned topics, structures, effects, etc. are clarified in the description of the method for implementing the invention.

[0016] Means for solving problems

[0017] A representative embodiment of the present invention has the structure shown below. A charged particle beam device of one embodiment comprises: a photographing device, which irradiates a sample with a charged particle beam and images the information of the sample; and a computer, which includes a memory for storing each image (scanned image) obtained by scanning the same area multiple times using the photographing device and a processor for performing image processing, wherein the computer performs the following processing: for each image obtained by scanning the same area multiple times, classifying into a first image including a degraded image and a second image not including a degraded image; performing image accumulation on the second image not including the degraded image; and saving a third image obtained by the image accumulation as a target image.

[0018] Furthermore, the charged particle beam device includes a database (DB) storing data including the images, information (including environmental information) obtained from the imaging device along with the images, selection states related to classification of the images, and the third image.

[0019] The computer performs selection processing including the classification, storage of data in the database, etc. Alternatively, the database may be distributed to an external device connected to the computer.

[0020] The user as an operator can correct the result of the selection process. The user particularly needs to confirm the result of the selection process when the cause of image quality degradation occurs. Among the causes of image quality degradation, there are causes that can be determined based on the transition of environmental information. Therefore, the computer associates the scanned image, the graphic of the environmental information, and the result of the selection process (selection state) and displays them on the screen. The items displayed in the graphic are preferably items that the user can judge the degraded image. The items can be, for example, shooting timing, probe current, and vacuum degree.

[0021] The user can confirm whether the image quality of the target image captured by the charged particle beam device is degraded. Since it is difficult to visually determine the image quality degradation, the computer calculates the image quality evaluation index value of the target image, and arranges and displays the target image and the image quality evaluation index value on the screen. The image evaluation value can also be displayed in a graph. Hereinafter, the value obtained based on the objective evaluation method used to evaluate the quality of the image is referred to as the image quality evaluation index value.

[0022] Effects of the Invention

[0023] According to a representative embodiment of the present invention, for a target image captured by a charged particle beam device, random noise can be suppressed by accumulation, and prevention of image quality degradation or improvement of image quality related to sharpness caused by image quality degradation can be achieved. In addition, according to a representative embodiment of the present invention, a target image of S / N and sharpness can be captured. Furthermore, according to a representative embodiment of the present invention, it is possible to help suppress damage to the sample and reduction in processing volume caused by the operator's search for shooting conditions during observation. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 It is a diagram showing the configuration of a charged particle beam device according to Embodiment 1 of the present invention.

[0025] Figure 2 This is a diagram showing a processing flow of an imaging function for preventing image quality degradation including a selection process in the charged particle beam device according to the first embodiment.

[0026] Figure 3 This is a diagram showing an example of a screen displayed on a display device as a user interface regarding settings before scanning in the charged particle beam device according to the first embodiment.

[0027] Figure 4 This is a diagram showing an example of a screen displayed on a display device as a user interface regarding display of a scan image in the charged particle beam device according to the first embodiment.

[0028] Figure 5 This is a diagram showing an example of a screen displayed on a display device as a user interface regarding display of an integrated image in the charged particle beam device according to the first embodiment.

[0029] Figure 6 This is a diagram showing a structural example of a database in Implementation Example 1.

[0030] Figure 7 This is a diagram showing the data flow of an imaging function for preventing image quality degradation in the charged particle beam device according to the first embodiment.

[0031] Figure 8This is a diagram showing an example of classification based on feature amounts and machine learning in the selection process in the charged particle beam device according to the first embodiment.

[0032] Fig. 9 This is a diagram showing the data flow of an observation function for preventing image quality degradation in the charged particle beam device according to the second embodiment of the present invention.

[0033] Fig.10 This is a diagram showing an example of a screen displayed on a display device as a user interface regarding error correction in the charged particle beam device according to the second embodiment. DETAILED DESCRIPTION

[0034] Hereinafter, embodiments of the present invention will be described in detail based on the drawings. In addition, in principle, the same reference numerals are given to the same parts in all the drawings, and repeated descriptions are omitted.

[0035] (Implementation Method 1)

[0036] use Figures 1 to 8 , a charged particle beam device according to Embodiment 1 of the present invention is described.

[0037] [Charged particle beam device]

[0038] Figure 1 The structure of the charged particle beam device 1 of Embodiment 1 is shown. The charged particle beam device 1 shows the case of SEM. The charged particle beam device 1 is composed of an imaging device 10, a computer 20, a display device 21, an operating device 22, etc. The operator as a user operates the charged particle beam device 1 through the display device 21 and the operating device 22. The user can observe the shape and structure of the surface of the sample 110, for example, on the screen of the display device 21.

[0039] The imaging device 10 is a SEM body, and a structure based on known technology can be applied. Figure 1 In the configuration example, the imaging device 10 includes an electron gun 101, a focusing lens 102, a deflection coil 103, an objective lens 104, a detector 105, a workbench 106 in a sample chamber, a vacuum pump 107, an environmental sensor 108, and the like.

[0040] The sample chamber is made into a vacuum state by a vacuum pump 107. The environmental sensor 108 is a sensor group that measures the state of the sample chamber such as vacuum degree, temperature, vibration, electromagnetic waves, etc., and is used to confirm the reached vacuum degree in the sample chamber. The sample 110 is arranged on the workbench 106. The workbench 106 can move in at least two directions (X direction and Y direction shown in the figure) in a plane perpendicular to the charged particle beam L1, and can move the field of view position. The electron gun 101 generates a charged particle beam L1 and irradiates it in the vertical direction (Z direction shown in the figure). The focusing lens 102 and the objective lens 104 focus the charged particle beam L1. The deflection coil 103 deflects the charged particle beam L1 in the X direction and Y direction shown in the figure, thereby realizing the scanning of the surface of the sample 110 by the charged particle beam L1. The detector 105 detects the signal of the secondary electron L2 from the sample 110, amplifies it through the signal amplifier, and transmits it to the computer 20.

[0041] The computer 20 is a control device of the SEM, and performs control such as driving of each unit of the imaging device 10. The computer 20 includes a processor 201, a memory 202, an input / output interface unit 203, etc., and these components are connected to a bus.

[0042] The input / output interface unit 203 is connected to the display device 21, the operating device 22, etc. The computer 20 displays a graphical user interface (GUI) for utilizing the functions of the SEM on the screen of the display device 21. The charged particle beam device 1 receives input operations from the user through the operating device 22. The operating device 22 includes a keyboard, a mouse, etc.

[0043] Whenever a signal is transmitted from the photographing device 10 to the input / output interface unit 203, the processor 201 forms a 2D image in the memory 202. The charged particle beam device 1 has an observation function of displaying the latest 2D image (sometimes recorded as an observation image) on the GUI. In addition, the processor 201 sets a condition suitable for storing the target image to the photographing device 10, performs image processing, information addition, etc. The charged particle beam device 1 has a photographing function of storing the target image thus obtained in the memory 202 (storage medium).

[0044] In addition, the configuration is not limited to the above, and an external storage device, server, or other device may be connected to the input / output interface 203 via communication, and various computer resources may be used to perform calculations, control, data storage, etc. In addition, multiple users and multiple display devices 21 may exist.

[0045] [Processing Flow]

[0046] Figure 2The processing flow for realizing the imaging function for preventing image quality degradation including the selection process of the computer 20 of the charged particle beam device 1 according to the first embodiment includes steps S1 to S9. The following will be described in order of the steps.

[0047] In step S1, the computer 20 receives the pre-scanning settings based on the user's operation. The settings are, for example, Figure 3 GUI. The setting may also utilize the default setting value. The setting includes the setting of the number of scans N for obtaining the scanned image. The charged particle beam device 1 starts scanning according to the user's instructions. The computer 20 controls the scanning of the imaging device 10 according to the settings before scanning. In addition, in this step S1, the items that must be set for the imaging device 10 before scanning are set. In this setting, for example, by setting the skip of steps S5 and S7 described later, it is possible to specify that the user as the operator does not intervene from the start of scanning to the saving of the target image.

[0048] In step S2 , the computer 20 stores the image and environmental information obtained through one scan. Figure 1 The imaging device 10 scans the surface of the sample 110 with the charged particle beam L1 according to the control from the computer 20, and transmits the signal obtained by the detector 105 to the computer 20. The processor 201 forms a two-dimensional image each time a signal is transmitted, and stores it in the memory 202. For the scanned image, the processor 201 associates the management information such as ID and the environmental information obtained from the imaging device 10 and sets it as a database (described later). Figure 6 The scanned image DB) is stored in the memory 202.

[0049] In step S3, the computer 20 determines the end of the scan of the charged particle beam L1 based on the number of scans N. The computer 20 determines whether the number of implementations of step S2 has reached N times. If it is determined that it has reached N times (yes), the process proceeds to step S4 to end the scan of the imaging device 10. If it is determined that it has not reached N times (no), the process returns to step S2 to continue the scan of the imaging device 10.

[0050] In step S4, the computer 20 performs a selection process to classify the scanned images into images containing degraded images and images not containing degraded images. The processor 201 refers to the scanned image DB and performs a selection process on the scanned images obtained in step S2. The result of the selection process is associated with the management information such as ID as a database (described later). Figure 6 The selection state DB) is stored in the memory 202. A specific embodiment of the selection process is described in [Data flow in the shooting function] described later.

[0051] In step S5, the computer 20 displays the results obtained in steps S2 to S4 on the display device 21. The results here mainly refer to the scanned image and the environmental information during the scan, and the selection status as the output of the selection process, which are obtained by referring to the scanned image DB and the selection status DB. In this display, for example, the following is used Figure 4 GUI.

[0052] If there is an inappropriate item in the selection state determined by the selection process in step S4, the user as an operator can input a correction to the GUI through the operating device 22. The computer 20 updates the selection state DB according to the correction input.

[0053] In step S6, the computer 20 performs the accumulation process to obtain an accumulated image with suppressed random noise. The processor 201 refers to the selection state DB and performs the accumulation process on the image whose selection state is selected. The result of the accumulation process is associated with the management information such as ID and the image quality evaluation index value for evaluating the accumulated image as a database (described later). Figure 6 The accumulated images DB are stored in the memory 202. The image quality evaluation index values ​​are, for example, line profile, sharpness, and granularity. The processor 201 performs image processing to obtain each accumulated image. The specific embodiment of the accumulation processing is described in the [Data flow in the shooting function] described later.

[0054] In step S7, the computer 20 displays the result obtained in step S6 on the display device 21. The result here mainly refers to the cumulative image as the output of the accumulation process and the image quality evaluation index value for evaluating the cumulative image, which is obtained by referring to the cumulative image DB. In this display, for example, the following is used Figure 5 GUI.

[0055] In step S8, the computer 20 receives input from the user as an operator as to whether the result obtained in step S6 is satisfactory. In other words, step S8 is to confirm whether the user needs to make a reselection. The result here is displayed on the display device 21 in step S7. If it is satisfactory (yes), it proceeds to step S8. If it is not satisfactory (no), it returns to step S5 and can correct the selection state again. In this display, for example, using the Figure 5 GUI.

[0056] In step S9, the computer 20 records the accumulated image obtained in step S6 as a target image to a recording medium, and ends a series of processing procedures. The target image can be saved in a place and name arbitrarily specified by the user. As a recording medium, for example, there is a memory 202 built into the computer 20. Alternatively, it can be recorded to an external storage device (DVD, flash memory, etc.) or a storage device on the network through the input and output interface unit 203. In addition, the charged particle beam device 1 adds information associated with the accumulated image DB to the recorded data so as to ensure traceability (if necessary, the information also includes information of the scanned image DB and the selected state DB).

[0057] [Shooting function GUI]

[0058] Figure 3 This is an example of a screen displayed on the display device 21 as a user interface for the pre-scanning settings of the charged particle beam device 1 of Embodiment 1. The "pre-scanning settings" screen has a number setting unit 302, an interval setting unit 303, a resolution setting unit 304, and a pixel dwell time setting unit 305 as items of "scanning settings" (corresponding to each setting unit).

[0059] The scan start button 301 is a button that the user can use to instruct the start of scanning. By pressing the scan start button 301, the charged particle beam device 1 controls the imaging device 10 so as to achieve the settings before scanning, and moves from step S1 to step S2. In addition, when step S4 is completed, the charged particle beam device 1 displays Figure 4 screen.

[0060] The number setting unit 302 is an item that can specify the number of scans for obtaining a scanned image. N set by the number setting unit 302 is N used for determination in step S3. The interval setting unit 303 is an item that can specify the interval between each scan for multiple scans. The charged particle beam device 1 controls so that the charged particle beam L1 does not irradiate the image after completing one scan. Figure 1 The charged particle beam device 1 scans the surface of the sample 110 and stops at the interval set by the interval setting unit 303. After that, the charged particle beam device 1 scans again, repeats scanning and pausing, and obtains a scanned image. The resolution setting unit 304 is an item that can specify the resolution of the image obtained by scanning. The pixel dwell time setting unit 305 is an item that can specify the time that each pixel is irradiated with the charged particle beam L1 for the image obtained by scanning. The charged particle beam device 1 sets the control signal of the deflection coil 103 so that it can scan for the time set by the pixel dwell time setting unit 305.

[0061] Figure 3 The setting parts of are described as combo boxes as an example of GUI, but they can also be replaced by sliders, preset buttons, etc.

[0062] Figure 4 The present invention is a display of a scanned image of the charged particle beam device 1 of the first embodiment, and is an example of a screen displayed on the display device 21 as a user interface. As a feature of the screen, the scanned image is displayed in association with a graph of environmental information. And, as a feature, the display can be performed frame by frame in order to confirm the changes in the field of view. As described above, according to the charged particle beam device 1, it is possible to monitor image quality degradation factors.

[0063] Figure 4 The "Confirmation of Selection Status" screen has an image display unit 101, a frame-by-frame forward button 402, a selection unit 403, a number display unit 404, a thumbnail display unit 405, a chart display unit 406, a scroll bar 407, a selection number label 408, a non-selection number label 409, a beginning setting unit 410, an end setting unit 411, a full selection button 412, a full release button 413, an automatic selection button 414, and an image accumulation button 415.

[0064] The image display unit 401 is a part that displays an image in a scanned image at a size that can confirm the details. In the illustrated example, the image display unit 401 displays the image #4, but it is not limited to this, and information that can determine the image can also be added. The frame-by-frame forward button 402 is a button for performing frame-by-frame forward display (in addition, frame by frame corresponds to the image). The frame-by-frame forward button 402 includes a left arrow button and a right arrow button. The charged particle beam device 1 updates the image of the image display unit 401 so that if the user presses the left arrow button, it returns to 1 frame, and if the right arrow button is pressed, it advances 1 frame. In the case of photographing a sample that is easily contaminated, if a certain number of scans is reached, the contrast of the image is reduced. In order to obtain a target image without the influence of contamination, the user updates the image of the image display unit 401 by advancing frame by frame through the frame-by-frame forward button 402, and confirms the timing of the contrast reduction. Thus, the user can select an image that is not affected by sample contamination and can obtain the desired target image.

[0065] The selection unit 403 indicates the selection state related to the classification of the image, and is a control (GUI component) for correcting the selection state determined by the selection process of step S4. For example, for each image, the image with a check is set to "selected", and the image without a check is set to "non-selected". The user presses the selection unit 403, thereby reversing the checked state of the selection unit 403. In the example shown in the figure, the selection state of #2 and #3 is "selected", and the selection state of #1 and #4 is "non-selected". The number display unit 404 is a part that indicates the order in which the images are acquired (for example, indicated by #1, etc.). The thumbnail display unit 405 is a part that displays the thumbnails of the scanned images in the order in which they are acquired.

[0066] The graph display unit 406 is a part that displays the environmental information obtained from the imaging device 10 as a graph associated with the image. In the example shown in the figure, the graph display unit 406 plots the emission current in a square and the vacuum degree in a circle. In the thumbnail display unit 405, the images are arranged in the order in which they were obtained, so the horizontal axis of the graph display unit 406 represents time, and the vertical axis represents the numerical value of each environmental information.

[0067] Apart from the cause of image quality degradation, it is difficult for the SEM and the sample to maintain a stable state with high accuracy, and in reality, the environmental information of the imaging device 10 is sometimes not fixed. By displaying the environmental information in a graph as described above, the user as an operator can optimize the selection state by referring to the change of the environmental information.

[0068] The scroll bar 407 is used to display the thumbnail display unit 405 and the chart display unit 406 as needed. Figure 4 The position of the knob in the scroll bar 407 is linked to the display positions of the thumbnail display unit 405 and the graph display unit 406. The thumbnail display unit 405 and the graph display unit 406 are arranged in the order obtained, and when the user moves the knob to the left, the old information is displayed, and when the user moves the knob to the right, the new information is displayed.

[0069] The selection number label 408 indicates the number of sheets whose selection state is "selected" in the scanned image. The number of images used in the accumulation process plays a role in suppressing random noise. Therefore, by referring to the selection number label 408, the user can imagine the random noise that will appear in the image when it becomes an accumulated image before the accumulation process. The non-selection number label 409 indicates the number of sheets whose selection state is "non-selected" in the scanned image. The user can use the non-selection number label 409 to predict the remaining number of sheets that can change the selection state from "non-selected" to "selected".

[0070] The beginning setting unit 410 is an item that can specify the beginning of the range to be processed cumulatively. For example, the field of view may drift in the period before a certain number of scans is reached. In order to exclude the degraded images before a certain number of scans are reached from the objects of cumulative processing, the user can specify the beginning of the range. The end setting unit 411 is an item that can specify the end of the range to be processed cumulatively. For example, when a certain number of scans is reached, whitening may occur due to charging. Or, sometimes, the contrast is reduced due to contamination. In order to exclude the degraded images after a certain number of scans are reached from the objects of cumulative processing, the user can specify the end of the range.

[0071] The all-select button 412 is a button that can update the selection status at once. When the user presses the all-select button 412, the charged particle beam device 1 sets all the selection statuses displayed in the selection section 403 to "selected". The all-release button 413 is a button that can update the selection status at once. If the user presses the all-release button 413, the charged particle beam device 1 sets all the selection statuses displayed in the selection section 403 to "non-selected". The automatic selection button 414 is a button that can update the selection status at once. When the user presses the automatic selection button 414, the charged particle beam device 1 sets the selection status of the image containing the degraded image to "non-selected" and sets the selection status of the image not containing the degraded image to "selected" through selection processing. This processing is the same as Figure 2 The same as step S4, used when returning to the selection state.

[0072] The image accumulation button 415 is a button that can perform image accumulation. When the user presses the image accumulation button 415, the charged particle beam device 1 newly obtains the selection state ID (information for managing the selection state recorded in the selection state DB described later) and updates it with the selection state of the selection unit 403. Among them, the range from the image at the beginning of the start setting unit 410 to the image at the end of the end setting unit 411 is set to "non-selected". The charged particle beam device 1 moves from step S5 to step S6, and performs the accumulation process with reference to the selection state DB. In addition, when the accumulation process is completed, the charged particle beam device 1 turns off the Figure 4 screen, open Figure 5 screen.

[0073] Figure 5 The display of the accumulated image of the charged particle beam device 1 according to the first embodiment is shown as an example of a screen displayed on the display device 21 as a user interface. The features of the screen include: the image quality evaluation index value is displayed in a graph as an index for judging the image quality, so that it can be evaluated quantitatively; and the reference data is recorded and compared, so that it can be evaluated relatively. The reference data refers to the accumulated image generated by changing the selection state from the same scanned image, and the image quality evaluation index value of the accumulated image, which is reference data for judging the image quality. The image quality evaluation index value is explained, for example, when line profile, sharpness, and granularity are used.

[0074] Figure 5 The “Display of Cumulative Image” screen has reference data 501 , generated data 502 , a save button 503 , a cumulative image 504 , a contour position 505 , a line contour 506 , a sharpness 507 , a granularity 508 , an update button 509 , a reselect button 510 , and an end button 511 .

[0075] Reference data 501 is a display area for setting cumulative images with different selection states as reference data in advance and comparing them with generated data 502 described later for evaluation. The charged particle beam device 1 reads the data set for the reference data from the cumulative image DB described later and displays it in each control of the GUI. In addition, when the reference data is not set by the update button 509 described later, the charged particle beam device 1 may set the selection state of all images as "selected" and the range to be the object of the cumulative processing as all images, generate an image obtained by image accumulation, and set it as reference data.

[0076] The generated data 502 is a display area for checking whether the integrated image generated in the integration process of step S6 satisfies the user's request as a target image. The charged particle beam device 1 reads the data generated in the last integration process from the integrated image DB described later and displays it in each control of the GUI.

[0077] The save button 503 is a button that can save the accumulated image as a target image. The reference data 501 and the generated data 502 each have a save button 503, and can save the respective accumulated images. For example, the user sets an accumulated image with a high image quality evaluation in advance, so that even if the data generated in the final accumulation process is used as a target image and does not have a sufficient image quality, the accumulated image with a high image quality evaluation can be saved as the target image based on the previously set reference data 501.

[0078] The cumulative image 504 is an area for displaying the cumulative image. In the example shown in the figure, the cumulative image with the cumulative image ID C1 is displayed in the reference data 501, and the cumulative image with the cumulative image ID C2 is displayed in the generated data 502, and information (cumulative image ID) that can identify the image is added. In addition, the position where the information that can identify the image is drawn does not need to be on the cumulative image 504.

[0079] The contour position 505 is a control for specifying the position of a line contour 506 described later. In the example shown in the figure, the contour position 505 is specified by moving up and down in accordance with the scanning of the common raster method in the SEM. The method of specifying the position of the line contour 506 can also be applied to a specifying method based on mouse dragging or the like.

[0080] Line profile 506 is a graph in which the horizontal axis represents the position of the pixel and the vertical axis represents the density value of the pixel for the pixel located on the line drawn on cumulative image 504. In the example shown in the figure, the minimum value is shown in the shape of letter A (black), while the maximum value is shown in the background area (white). It can be judged that the sharpness of generated data with a small number of pixels from the maximum value to the minimum value is high. In addition, it can be judged that the contrast of generated data with a large difference between the maximum value and the minimum value is high.

[0081] The sharpness 507 displays the sharpness evaluation value calculated from the integrated image 504 in a bar form. For example, the DR method is used as a sharpness evaluation method. If the standard deviation of the intensity distribution of the probe in the SEM is σ, the sharpness R is given by To express.

[0082] The granularity 508 displays the evaluation value of the granularity calculated from the integrated image 504 in a bar form. For example, RMS granularity is used as the evaluation method of granularity. In this case, the granularity is expressed by the following formula 1.

[0083] [Formula 1]

[0084] The lightness and darkness value of each pixel: D i

[0085] Average value of lightness and darkness:

[0086] Granularity: RMS

[0087]

[0088] The update button 509 is a button that can update the reference data 501 . When the user presses the update button 509 , the charged particle beam device 1 sets the cumulative image ID described later associated with the generated data 502 to the cumulative image ID described later associated with the reference data 501 .

[0089] The reselect button 510 is a button that can return to step S5. When the user presses the reselect button 510, the charged particle beam device 1 is turned off. Figure 5 The screen is displayed again Figure 4 For example, when the user confirms that the image quality has deteriorated in the cumulative image generated by the cumulative processing of step S6, the reselect button 510 is used to change the selection state and generate the cumulative image again.

[0090] Here, the user is allowed to determine whether the image quality has deteriorated, but the processor 201 may determine whether the image quality has deteriorated without involving the user's determination. For example, when the image quality evaluation index value obtained by comparing the image obtained by selecting all of the multiple images (scanned images) and accumulating the images, that is, the reference accumulated image, with the current accumulated image is improved by more than 20%, the processor 201 considers that the accumulated image has no image quality deterioration and ends the processing flow.

[0091] In addition, for example, the following may be used. The processor 201 generates an accumulated image obtained by selecting all of the plurality of images and accumulating the images, independently of the user's operation, and generates accumulated images corresponding to each mode of selecting or not selecting each image from the plurality of images. The processor 201 compares these plurality of accumulated images to evaluate the image quality. Based on the evaluation result, the processor 201 stores the accumulated image selected from the plurality of accumulated images and ends the processing flow.

[0092] End button 511 is to close Figure 5 A button for ending the processing flow of the imaging function for preventing image quality degradation including the selection processing implemented by the computer 20 of the charged particle beam device 1 according to the first embodiment is displayed on the screen.

[0093] [database]

[0094] Figure 6 The following shows the databases included in the charged particle beam device 1 according to Embodiment 1. The databases are composed of three: a scanned image DB 610 , a selected state DB 620 , and an accumulated image DB 630 .

[0095] The scanned image DB 610 is a database for associating scanned images, information including environmental information obtained from the imaging device 10 at each scan, and management information, and has the following attributes: The scanned image DB 610 has a scanned image ID 611, a number 612, an image file 613, an imaging timing 614, a probe current 615, and a vacuum degree 616.

[0096] The scanned image ID 611 is information for managing the scanned image. Figure 2 In step S1, the charged particle beam device 1 assigns a new inherent value, which is set as a clue (key) for determining the scanned image, namely, the scanned image ID 611. Through this clue, it is associated with the selection state DB 620. Number 612 is the order in which the images are obtained. Number 612 is management information for determining the image obtained by one scan from the scanned image. Whenever the charged particle beam device 1 obtains an image in step S2, it assigns a count value, which is set as number 612. The selection state DB 620 also has the same attributes. The image file 613 is each image (scanned image) obtained in step S2. In the illustrated example, the image file 613 is not the actual state of the image, but is information as a storage destination to a recording medium, namely, path information of the file. Not limited to this, the image file 613 can also be the actual state of the image or a file obtained by compressing it.

[0097] The shooting timing 614 indicates the time when the image is obtained. In the example shown in the figure, the shooting timing 614 uses the first image obtained as a reference (0) and uses the elapsed time to indicate the timing of each image (scanned image) obtained by scanning the same area multiple times. This is not limited to this, and the shooting timing 614 can also be expressed as a moment. The probe current 615 is Figure 1 The current of the charged particle beam L1 irradiated by the charged particle beam device 1 to the sample 110. The probe current 615 can be calculated based on the control values ​​of the electron gun 101, the focusing lens 102, and the objective lens 104. The vacuum degree 616 is the atmospheric pressure in the sample chamber that is made into a vacuum state by the vacuum pump 107. The vacuum degree 616 can be measured by the environmental sensor 108 installed in the sample chamber.

[0098] Shooting timing 614, probe current 615 and vacuum degree 616 are examples of environmental information when an image is acquired by the shooting device 10. For example, if the probe current 615 decreases, the signal amount of the secondary electron L2 decreases, and the image becomes darker. Alternatively, if the vacuum degree 616 decreases, the focused charged particle beam L1 is scattered by the atmosphere, and the image is blurred. In this way, environmental information related to image quality can be used for selection processing. Therefore, the charged particle beam device 1 associates the environmental information with the image and records it in the database. Through the shooting device 10, in the presence of other environmental information related to image quality, the charged particle beam device 1 adds corresponding items to the attributes of the environmental sensor 108 and the scanned image DB610, and records them, etc.

[0099] The selection status DB 620 is a database for associating the selection status related to the above-mentioned classification with the management information, and has the following attributes: The selection status DB 620 has a selection status ID 621 , a scanned image ID 611 , a number 612 , and a selection status 612 .

[0100] The selection state ID 621 is information for managing the selection state. Figure 2 In step S4, the charged particle beam device 1 assigns a new inherent value when executing the selection process, and sets it as the clue to determine the selection state, namely the selection state ID621. In step S5, the charged particle beam device 1 assigns a new inherent value when determining the correction of the selection state. Therefore, there may be multiple selection state IDs 621 for the scanned image ID611. It is associated with the cumulative image DB630 through the selection state ID621. The selection state 622 has a "selection" indicating an input for cumulative processing or a "non-selection" indicating an input that is not cumulative processing as a state, and each state can also be expressed by a discriminable value. In the example shown in the figure, "selection" is set to 1 and "non-selection" is set to 0.

[0101] The accumulated image DB 630 is a database for associating accumulated images, image quality evaluation index values ​​for evaluating accumulated images, and management information, and has the following attributes: The accumulated image DB 630 has an accumulated image ID 631 , a selection state ID 621 , an image file 632 , a line profile 633 , a resolution 634 , and a granularity 635 .

[0102] The accumulated image ID 631 is information for managing accumulated images. Figure 2 In step S6, the charged particle beam device 1 assigns a new inherent value, which is set as a clue to determine the cumulative image, that is, the cumulative image ID 631. The image file 632 is the cumulative image obtained in step S6. In the example shown in the figure, the image file 632 is not the actual state of the image, but is information as a storage destination to the recording medium, that is, the path information of the file. Not limited to this, the image file 632 can also be the actual state of the image or a file after compressing the image.

[0103] The line profile 633 is a set of numerical columns based on the shading values. In the illustrated example, the line profile 633 is not the actual state of the set, but is information of the storage destination to the recording medium, that is, the path information of the file. Not limited to this, the line profile 633 may also be the actual state of the set or the state after the set is compressed. The sharpness 634 is the calculated index value of sharpness. The granularity 635 is the calculated index value of granularity.

[0104] Line profile 633, sharpness 634, and granularity 635 are image quality evaluation index values ​​used to evaluate the cumulative image. For example, when it is difficult to evaluate the edge of a structure by visual inspection of the image, it is sometimes possible to evaluate it by using line profiles. In the cumulative image DB 630, if there are image quality evaluation index values ​​suitable for judging image quality, the attributes are added and recorded. If necessary, Figure 5 In the screen, a control that can be displayed in a graph is added corresponding to the additional image quality evaluation index value.

[0105] [Data flow during shooting function]

[0106] Figure 7 The data flow showing the relationship between the database, the selection process, and the accumulation process in the imaging function for preventing image quality degradation of the charged particle beam device 1 according to the first embodiment is shown.

[0107] Environmental information 701 is the collective data of shooting timing 614, probe current 615, and vacuum degree 616 recorded in scanned image DB 610. Scanned image 702 is the collective data of image files 613 recorded in scanned image DB 610. Non-selected image 703 is the collective data of image files whose selection status 622 is "non-selected" (0) recorded in selection status DB 620. Selected image 704 is the collective data of image files whose selection status 622 is "selected" (1) recorded in selection status DB 620. Cumulative image 705 is one of image files 632 recorded in cumulative image DB 630.

[0108] The selection process 706 is equivalent to Figure 2 Step S4 in is executed by the processor 201. The selection process 706 is a process of inputting the environmental information 701 and the scanned image 702, and classifying them into two images, namely, the selected image 704, which is an image required to generate the target image, and the non-selected image 703, which is an image not required to generate the target image. Here, the image required to generate the target image refers to the scanned image 702 that does not contain a degraded image. In embodiment 1, in order to automatically classify the provided data set in a manner without a reference such as a threshold, unsupervised machine learning is applied in the selection process 706. For example, the k-means method in unsupervised machine learning is applied. The k-means method can use the average of the clusters and classify them into the assigned k number of clusters. The computer 20 first randomly sets two clusters, so that the scanned images 702 belong to the clusters with high similarity respectively, and sets the center of gravity for the clusters according to the clusters. The computer 20 again makes each scanned image 702 belong to the cluster with high similarity, and sets the center of gravity for the clusters according to each cluster. The computer 20 repeatedly performs this operation to converge the scanned images into two clusters with similar properties, and uses the two clusters as the selected image 704 and the non-selected image 703. The similarity and the centroid can be calculated from the feature quantities obtained from the scanned images.

[0109] In addition, when multiple causes of image quality degradation occur, an extended algorithm of the k-means method is used. This extended algorithm is a method for automatically estimating the number of clusters for the distribution of all defined feature spaces. In this method, for each scanned image 702 obtained, it is classified only according to the features corresponding to the causes of image quality degradation that have occurred, and the cluster containing the reference image among the three or more clusters after classification is used as the scanned image 702 that does not contain the degraded image. The computer 20 stores the time from the start of step S2 to the stabilization of the shooting device 10 as a parameter, and uses the image obtained at the stable timing as the reference image. Alternatively, the user can also specify the reference image through the GUI.

[0110] For example, if a certain number of scans is reached, the entire image area may turn white due to charging. Alternatively, the entire image area may turn black due to contamination. Due to these phenomena, the contrast of the image obtained by one scan is reduced. Therefore, the k-means method using contrast as a feature quantity can be used to classify images into images with maintained contrast and images with reduced contrast.

[0111] exist Figure 8 On the upper side of (a), the scanned images are arranged at the acquired times t1 to t5, and the time when the scanned images are acquired is time t5. On the lower side, a graph with the contrast calculated based on the arranged images as the vertical axis is shown. When observing the transition of the scanned images, the structure that is black at times t1 to t3 (in this example, a structure in the shape of the letter A) gradually turns white from time t4. In this graph, the transition of the contrast decreasing from time t4 is shown, and the two clusters represented by cluster C1 and cluster C2 can be classified by the k-means method using contrast as a feature quantity. In addition, the cross represents the center of gravity of each cluster, and the portion belonging to the cluster is surrounded by a dotted line. In the example shown in the figure, the image acquired at time t1 is used as a reference image, and cluster C1 containing the reference image can be judged as a scanned image that does not contain a degraded image. In this way, according to embodiment 1, even for the contrast that changes each time a scanned image is obtained, classification can be performed without setting a threshold value or the like.

[0112] In addition, when the image is deformed due to vibration before the scan is completed, by using the image of the structure captured without deformation in the scanned image and the correlation value of each image as the feature value, it is possible to classify it into images similar to the structure without deformation and deformed images through the k-means method.

[0113] Figure 8 (b) is to Figure 8 The vertical axis of the graph in (a) is a graph of correlation values. The correlation value at time t1 is the maximum value because it is the correlation of the same image. The correlation values ​​at time t2, time t4, and time t5 are the same value because the structure is not deformed. Since only the structure at time t3 is deformed, the correlation value is lowered, and the two clusters, cluster C3 and cluster C4, can be classified by the k-means method using the correlation value as the feature value. In the example shown in the figure, the image obtained at time t1 is used as the reference image, and cluster C3 containing the reference image can be judged as a scanned image that does not contain a degraded image.

[0114] In the above example, the feature quantity is obtained based on the information included in the image obtained by one scan, but the imaging timing 614, the probe current 615, and the vacuum degree 616 included in the environmental information 701 may also be defined as the feature quantity. Similarly, the relationship between the image and environmental information obtained by one scan and the image and environmental information obtained by the next scan, such as the differential value of the probe current, may also be defined as the feature quantity.

[0115] Thus, processor 201 performs Figure 7 In the selection process 706 , machine learning that defines features suitable for each cause of image quality degradation is used, so that only the scanned images 702 that do not include degraded images can be selected as the selected images 704 .

[0116] Figure 7 The cumulative processing of 707 is equivalent to Figure 2 Step S6 in the above step is executed by the processor 201. The accumulation processing 707 inputs the selected image 704 and outputs the accumulated image 705. For the set data of the input image, the accumulation processing 707 accumulates the shading value of each pixel as the shading value of the corresponding pixel of the accumulated image 705. Thus, the accumulated image 705, which is the target image with suppressed random noise, is obtained. The accumulation processing 707 is not limited to this, and the normalized accumulation described in Patent Document 1 can also be applied.

[0117] In addition, the accumulation processing 707 includes a process called drift correction processing described in Patent Document 2, which is a process of accumulating the shading value of each pixel while performing image alignment. In realizing the shooting function to prevent image quality degradation, if the charged particle beam device 1 is aligned to an erroneous position by the drift correction processing, Figure 2 In step S5, the selection state 622 of the image that is the cause is changed and removed from the input of the accumulation process 707, thereby improving the accuracy.

[0118] (Implementation Method 2)

[0119] use Fig. 9 as well as Fig.10 , a charged particle beam device according to Embodiment 2 of the present invention is described. The basic structure of Embodiment 2 is the same as that of Embodiment 1. The following describes the structural parts of Embodiment 2 that are different from those of Embodiment 1. In addition to the conventional observation function, the charged particle beam device of Embodiment 2 also has a function of observing the sample 110 ( Figure 1 ) is applied to the observation image of the monitor (display device 21) to select and accumulate images. Thus, the image quality of the observation image displayed on the monitor (display device 21) is prevented from being deteriorated.

[0120] [Observe the data flow in the function]

[0121] Fig. 9 In the embodiment 2, as the structure of the charged particle beam device of the second embodiment, the difference from the charged particle beam device 1 of the first embodiment, that is, the data flow is shown. In the second embodiment, Fig. 9 The environmental information 701, scanned image 702, and accumulated image 705 represent the observation function of real-time update, so they are not aggregate data, but current (latest) values ​​or images. For example, the environmental information 701 uses a black frame to represent the current acquired value, and a gray frame to represent the past acquired value. In addition, in the second embodiment, the above Figure 7 The accumulated image 705 is updated in real time as an observation image, so there is no need to store the image before the update in the database for a long time. In addition, in the data flow of the observation function in the second embodiment, in order to display the observation image requested by the user and set the conditions of the imaging device 10, the association between the computer 20, the display device 21, and the operating device 22 is added.

[0122] Fig. 9 The selection processing 901 of embodiment 1 has the same function as the selection processing 701 of embodiment 1, and is classified into a scanned image 702 that does not include a degraded image and images other than this. In the observation image that can dynamically obtain various fields of view, in order to classify a scanned image as the current scanned image 702, a past scanned image 702 and a correct selection state 622 are required. Therefore, the charged particle beam device of embodiment 2 pre-learns the learning data DB903 described later, and applies supervised machine learning based on the classification of the past scanned image 702 to the selection processing 901. For example, in supervised machine learning, a method using a convolutional neural network (CNN) is applied. This method uses CNN to find features from data such as images and classify them according to supervision. In this case, the selection processing 901 inputs the shading value of a scanned image as the current scanned image 702, and obtains output from each unit corresponding to the selected image 704 and the non-selected image 703. In the learning of CNN, the shade values ​​of one scanned image as the past scanned image 702 are sequentially input, and the weight is updated based on the error between the output from the obtained unit and the output from the unit corresponding to the correct selection state 622. By repeatedly learning CNN, it is possible to find features from the image and classify the selected image 704 and the non-selected image 703. In addition, there may be a unit corresponding to the non-selected image 703 for each cause of image degradation.

[0123] Fig. 9The cumulative processing 902 has the same effect as the selection processing 701 of the first embodiment, and the cumulative image 705 with suppressed random noise is obtained from the selected image 704. In the case of the observation image that can dynamically obtain various fields of view, if the same processing as the first embodiment is performed, the responsiveness will be impaired when it becomes the collection data of images with different fields of view. Therefore, a filter called a recursive filter that simulates the afterglow characteristics of a fluorescent body is usually applied. The method is as follows: when accumulating the light and dark values ​​of the pixels, the proportion is changed to perform the accumulation so that the newer the image, the greater the proportion of the image. In this way, the responsiveness is not impaired, and the cumulative image 705 with suppressed random noise is obtained.

[0124] The learning data DB 903 is a DB having data in the same format as the selection state DB 620. The learning data DB 903 associates the correct selection state 622 with the past scanned image 702 through the error correction 904 described later. In order to improve the accuracy, the learning data DB 903 may be divided according to the optical conditions determined by the sample 110 observed by the imaging device 10, the deflection coil 103, etc., and the type of the detector 105. In this case, in the selection process 901, the machine learning learned under the same conditions in the divided learning data DB 903 is used.

[0125] After the use of the observation function for preventing image quality degradation in the second embodiment, error correction 904 is performed in a stopped state. The error correction 904 is performed based on the error correction operation performed by the user. The charged particle beam device uses the non-selected image 703 and the selected image 704 of the selection state DB 620 added during the use of the observation function as input, corrects the erroneous selection state 622 based on the user input from the operating device 22, etc., and adds it to the learning data DB 903. After the error correction 904 is performed, the learning data DB 903 is accumulated every time the observation function of the second embodiment is started again, and it can be expected that the accuracy and versatility will be improved.

[0126] In addition, by using the non-selected image 703 and the selected image 704 of the selected state DB 620 added during the use of the observation function as input, the work of generating the learning data DB 903 is only to correct the erroneous selected state 622. That is, according to the second embodiment, although it depends on the accuracy of the accumulation processing 902, it is expected that the work of generating the learning data DB 903 can be reduced.

[0127] [Observe the GUI in the function]

[0128] Fig.10 The charged particle beam device according to the second embodiment Fig. 9 An example of error correction 904 is displayed on a screen of the display device 21 as a user interface. Fig.10The “error correction” screen has a selection image column 1001 , a non-selection image column 1002 , an image display unit 1003 , a correction unit 1004 , a number display unit 1005 , an update button 1006 , and a cancel button 1007 .

[0129] The selection image column 1001 lists and displays the selection images 704 ( Fig. 9 The non-selected image column 1002 lists and displays the non-selected images 703 ( Fig. 9 The image display unit 1003 is a region for displaying each image as the non-selected image 703 or the selected image 704.

[0130] The correction unit 1004 is a part that allows the user to input correction of errors in the selection state of the image. For example, a cross mark (×) in the correction unit 1004 indicates a mark that is corrected due to an error in the selection state, and no mark indicates a mark that is not corrected in the selection state. When the user presses the correction unit 1004, the cross mark and no mark are reversed. In the example shown in the figure, the correction unit 1004 is in the state of the cross mark and correction is performed. The number display unit 1005 indicates the order in which the images are acquired.

[0131] Update button 1006 is used to update the Fig. 9 When the user presses the update button 1006, the charged particle beam device copies the amount of data that has not been added to the learning data DB 903 from the selection state DB 620, and corrects the selection state 622 according to the state input to the correction unit 1004. The charged particle beam device adds the generated data to the learning data DB 903, closes Fig.10 When the charged particle beam device updates the learning data DB 903, in order to divide the data according to the sample 110 observed by the imaging device 10, the optical conditions determined by the deflection coil 103, and the type of the detector 105, a screen for setting the category may be displayed to prompt the user to input. The cancel button 1007 is a button that can cancel the update. When the user presses the cancel button 1007, the charged particle beam device does not update the learning data DB 903 and closes. Fig.10 screen.

[0132] [Observation function to prevent image quality degradation]

[0133] Regarding the charged particle beam device of the second embodiment, for example, it is assumed that floor vibration caused by carrying a heavy object is transmitted to the imaging device 10. In this case, the position deviation of the probe scanning the sample is reflected in the current scan image 702 ( Fig. 9) is deformed in one image. Here, in the case of the conventional observation function, the cumulative processing 902 ( Fig. 9 ), the weight of the current scanned image 702, i.e., one image, is increased for accumulation, so that the contour of the fine structure is displayed as a double observation image on the monitor (display device 21).

[0134] On the other hand, in the case of applying the above-mentioned image selection and image accumulation observation function in the second embodiment, it is possible to perform learning in advance on the learning data DB 903, based on the past scanned image 702 ( Fig. 9 ), the current scanned image 702 is classified as a degraded image. Thus, the current scanned image 702 is excluded from the target of the accumulation process 902, and the contour of the fine structure displayed on the monitor (display device 21) is not a double observation image.

[0135] When the user as the operator wants to display an observation image in which the contour of the fine structure is not doubled, the user performs a switching operation 905 from the conventional observation function to the observation function to which the image selection and image accumulation according to the second embodiment are applied. Fig. 9 ).

[0136] By the selection process 901, when the selection state 622 is only non-selected, the observation image of the monitor (display device 21) is no longer updated, so the user performs a switching operation 905 from the observation function to which image selection and image accumulation are applied to the conventional observation function. Alternatively, when the selection state 622 becomes only non-selected over a fixed period of time, the processor 201 may automatically switch from the observation function to which image selection and image accumulation are applied to the conventional observation function.

[0137] Description of Reference Numerals

[0138] 1 ... charged particle beam device, 10 ... imaging device, 20 ... computer, 201 ... processor, 202 ... memory, 21 ... display device, 22 ... operating device.

Claims

1. A charged particle beam device, characterized in that: have: an imaging device that irradiates a sample with a charged particle beam and images information of the sample; and a computer having a memory for storing each image obtained by scanning the same area multiple times using the imaging device and storage time information as a parameter, and a processor for performing image processing, The computer performs the following processing: Classifying the images obtained by scanning the same area a plurality of times into image groups by unsupervised machine learning, and defining the image group including the designated reference image as the image group not including degradation; Performing image accumulation on a group of images that do not contain the degradation; as well as An image obtained by accumulating the images is output.

2. The charged particle beam device according to claim 1, characterized in that The computer, in the unsupervised machine learning, selects a cluster including a reference image among clusters classified based on the characteristics of the cause of image quality degradation as a group of images not including degradation.

3. The charged particle beam device according to claim 2, characterized in that The characteristics of the cause of image quality degradation include any one of contrast, a correlation value of image deformation, imaging timing, probe current, vacuum degree, and a differential value of the probe current.

4. The charged particle beam device according to claim 1, characterized in that The charged particle beam device includes an operation device for receiving user input and a display device for displaying a graphical user interface. The computer displays the images obtained by scanning and time information during scanning on the screen of the display device, A reference image is specified according to an operation of the user on the screen.

5. The charged particle beam device according to claim 1, characterized in that The charged particle beam device includes an operation device for receiving user input and a display device for displaying a graphical user interface. The computer displays the images and the selected states of the images on the screen in association with a graph of environmental information.

Images