Charged particle beam apparatus and method for outputting target image data

Abstract

This charged particle beam device comprises: a sample stage that moves a sample; an imaging unit that acquires observation image data of the sample; an output unit that digitizes an operating state of the charged particle beam device, and outputs time series data of the operating state; a display unit on which the observation image data is displayed, and a graphical user interface for inputting observation setting parameters is displayed; and a computer system that stores time series image data in which the observation image data is arranged in a time series, and executes an arithmetic process that is related to the time series data of the operating state and the observation image data. The charged particle beam device automatically determines, on the basis of the time series data of the operating state, a date and time that match a specific fluctuation pattern that is set in advance, acquires, from the time series image data, observation image data that corresponds to the date and time, and outputs the observation image data as image data of interest.

Description

Technical Field

[0001] The present invention relates to a charged particle beam device and a method for outputting target image data.

Background Art

[0002] With the recent evolution of semiconductor devices, their device structures have become more complex. For semiconductor manufacturers producing advanced devices, how to develop the processes of such devices quickly and efficiently is an important issue. In semiconductor process development, it is essential to optimize the conditions for processing the materials deposited and laminated on a silicon (Si) wafer into the designed shape, and for this purpose, observation of the processed shape of the cross-sectional pattern is required.

[0003] Since the processed patterns of advanced semiconductor devices are fine structures at the nanometer level, charged particle beam devices such as a transmission electron microscope (TEM) and a high-resolution scanning electron microscope (SEM) are used for observing the processed shape of the cross-sectional pattern.

[0004] Currently, the observation of the processed shape of the wafer cross-section using a charged particle beam device is entrusted to manual work by humans, and a lot of labor and time are required for searching the observation field and imaging work. Therefore, for the acceleration and high efficiency of semiconductor process development, there is a demand for a device that can automate this observation work as much as possible and acquire a large amount of observation data at high speed and with less manpower.

[0005] In addition, due to reasons such as the progress of materials informatics technology, the demand for a device that can acquire a large amount of observation data at high speed and with less manpower is increasing also for observation objects other than semiconductors, such as metal materials and biological samples.

[0006] By utilizing machine learning and AI technologies for image processing, it is possible to automatically detect characteristic objects during observation, reducing the workload on the operator when aligning the field of view to the observation target position. For example, in semiconductor process development, it may be necessary to observe multiple wafer samples processed in a series of experimental series. Once a feature discriminator for the object is constructed by machine learning, it is also possible to automate the process from field-of-view search to pattern observation for the same wafer samples. However, in another experimental series, since the pattern layout also changes, it is necessary to newly construct an identification model for the target at regular intervals. Since a large amount of labeled image data is indispensable for constructing an identification model (feature discriminator) by machine learning, there is still a problem that a large workload is generated for the operator to collect the labeled image data himself / herself.

[0007] The labeled image data is an image including the object that the operator focuses on, and it is also possible to newly acquire an image through the observation operation, or to search for and reuse an image that seems to be usable from a past image database. As a means to easily search the image data acquired by the measuring device without imposing a burden on the operator, for example, in Patent Document 1, a technique for searching for an image estimated to be important for the operator from the images stored in the database in an SEM device is disclosed. In that disclosure, an "importance" score is created from the data recording various operation commands performed during observation, such as autofocus, brightness adjustment, and stage movement, and the score is assigned to the acquired image data, so that image data conforming to the importance is narrowed down during image search. However, in the method disclosed in Patent Document 1, when collecting new labeled image data corresponding to a new experimental series, since there are no appropriate images in the stored database, ultimately, it is necessary to manually acquire the labeled image data one by one. Therefore, the method disclosed in Patent Document 1 cannot address the issue of reducing the burden on the operator related to the collection of new labeled image data.

Prior Art Documents

Patent Documents

[0008]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0009] The present disclosure provides a charged particle beam apparatus and a method for outputting target image data, which have a function of semi-automatically generating teacher image data in sample observation using the charged particle beam apparatus, constructing a feature discriminator for automatically recognizing an observation field of view using the teacher image data, and performing automatic imaging based on these.

Means for Solving the Problems

[0010] An example of the charged particle beam apparatus according to the present invention is a sample stage for moving a sample, an imaging unit for acquiring observation image data of the sample, an output unit for quantifying the operating state of the charged particle beam apparatus and outputting time-series data of the operating state, a display unit for displaying the observation image data and displaying a graphical user interface for inputting observation setting parameters, a computer system for storing time-series image data in which the observation image data is arranged in time series, and performing arithmetic processing on the time-series data of the operating state and the observation image data, and includes automatically determining a time corresponding to a preset specific variation pattern based on the time-series data of the operating state, acquiring and outputting, as target image data, the observation image data corresponding to the time from the time-series image data.

[0011] An example of the method according to the present invention is a method for outputting target image data by a charged particle beam apparatus, wherein the charged particle beam apparatus includes a sample stage for moving a sample, An imaging unit that acquires observation image data of the sample; An output unit that quantifies the operating state of the charged particle beam apparatus and outputs time-series data of the operating state; A display unit on which the observation image data is displayed and a graphical user interface for inputting observation setting parameters is displayed; A computer system that stores time-series image data in which the observation image data is arranged in time series, and executes arithmetic processing related to the time-series data of the operating state and the observation image data; Comprising: The method includes: Automatically determining a time that matches a preset specific variation pattern based on the time-series data of the operating state; Obtaining and outputting, as target image data, the observation image data corresponding to the time from the time-series image data; Comprising:

Advantages of the Invention

[0012] According to the charged particle beam apparatus and the method for outputting target image data according to the embodiments of the present disclosure, teacher image data necessary for learning a feature discriminator for field recognition can be generated semi-automatically. As a result, the time and labor for field search during sample observation can be significantly reduced, and automatic imaging of cross-sectional images becomes possible.

Brief Description of the Drawings

[0013]

Figure 1

Figure 2A

Figure 2B

Figure 3A

Figure 3B

Figure 4A

Figure 4B

Figure 4C

Figure 4D

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9A

Figure 9B

Figure 9C

Figure 9D

Figure 10A

Figure 10B

Figure 11

Figure 12

Figure 13A

Figure 13B

Figure 14

Figure 15

Embodiments for Carrying Out the Invention

[0014] For example, in semiconductor device development, when performing cross-sectional observation of a processing pattern in a series of experimental series, it is assumed that a process engineer and an SEM operator conduct the first test observation together, and while observing an actual sample, they match the recognition of the processing pattern shape to be observed and the mark pattern for finding the observation location (pre-observation). If the information at that time is digitized, it includes information regarding the ROI (Region Of Interest) that the process engineer is focusing on.

[0015] The exemplary charged particle beam apparatus of the present disclosure records the screen during observation in pre-observation etc. as "time-series image data" such as a video, and at the same time records "time-series data of the operating state" of the apparatus such as sample stage coordinates and observation magnification, and by analyzing the two in cooperation, it semi-automatically generates teacher image data necessary for learning a feature discriminator for field recognition.

[0016] Specifically, an exemplary charged particle beam apparatus according to the present disclosure includes an imaging unit that irradiates a sample with a charged particle beam to acquire image data of the sample at a predetermined magnification, a computer system that performs an operation process for searching for a field of view when acquiring the image data using the image data, and a display unit on which a graphical user interface (GUI) for inputting setting parameters for the field of view search is displayed. The imaging unit is configured to be able to move the sample by at least two drive axes, and includes a sample stage that can move an imaging field corresponding to position information of the sample required by the computer system. The computer system records the observation image displayed on the display unit as "time-series image data" which is video data or a set of images continuously captured at regular time intervals, and at the same time, records "time-series data of the operating state" of the apparatus. Further, the computer system calculates an event time estimated to be noticed by the operator from a specific variation pattern preset in the time-series data of the operating state, and extracts a plurality of image data matching the event time from the time-series image data as target images. Further, the computer system records information such as the position coordinates of the ROI selected by the operator and the observation magnification from among the target images. Based on the information, the sample stage is moved to the position coordinates of the ROI, and a plurality of images of the ROI are acquired under a plurality of conditions such as different observation magnifications and the tilt angle of the sample stage. The computer system generates teacher image data for learning by cutting out a predetermined range from the image of the ROI, and generates a feature discriminator using the same. The feature discriminator executes a process of outputting position information of the ROI existing one or more on the input new image data for the input new image data.

[0017] Hereinafter, embodiments of the present disclosure will be described in more detail. However, the disclosed content of each embodiment is not limited to the following description only, and a configuration in which the disclosed or suggested element technologies of each embodiment are appropriately combined within the scope of the knowledge of those skilled in the art is also included in the scope of this embodiment.

[0018] [First Embodiment] The first embodiment proposes an automatic observation method of a sample using an automatic recognition function of a field of view realized by implementing a function of semi-automatically generating teacher image data including an ROI of an observation target in a charged particle beam apparatus in which a scanning electron microscope (SEM) is an imaging device.

[0019] FIG. 1 shows a configuration diagram of the scanning electron microscope according to the first embodiment. The scanning electron microscope 10 according to the first embodiment is an example of a charged particle beam apparatus, for example, a field emission scanning electron microscope (FE-SEM). The scanning electron microscope 10 can execute the method of outputting target image data described in the present embodiment.

[0020] As an example, the scanning electron microscope 10 includes an electron gun 11, a focusing lens 13, a deflection lens 14, an objective lens 15, a secondary electron detector 16, a sample stage 17, an image forming unit 31, a control unit 33, a display unit 35, and an input unit 36, and further includes a computer system 32 that executes arithmetic processing necessary for the field of view search function of the present embodiment. Hereinafter, each component will be described.

[0021] The electron gun 11 irradiates the sample with an electron beam (charged particle beam). The electron gun 11 includes a source that emits an electron beam 12 accelerated by a predetermined acceleration voltage. The emitted electron beam 12 is focused by the focusing lens 13 and the objective lens 15 and irradiated onto the sample 20. The deflection lens 14 deflects the electron beam 12 by a magnetic field or an electric field, whereby the surface of the sample 20 is scanned with the electron beam 12.

[0022] The sample stage 17 has a function of translating the sample 20 along a predetermined drive axis or a function of tilting and / or rotating the sample 20 around a predetermined drive axis in order to move the imaging field of view of the scanning electron microscope 10, and includes an actuator such as a motor or a piezo element for this purpose.

[0023] The secondary electron detector 16 is an E-T detector, semiconductor detector, etc. equipped with a scintillator, light guide, and photomultiplier tube, and detects secondary electrons 100 emitted from the sample 20 irradiated with the electron beam 12. The detection signal output from the secondary electron detector 16 is transmitted to the image formation unit 31. Note that, together with the secondary electron detector 16, a backscattered electron detector for detecting backscattered electrons or a transmission electron detector for detecting transmitted electrons may be provided.

[0024] The image formation unit 31 is composed of an AD converter that converts the detection signal transmitted from the secondary electron detector 16 into a digital signal, and an arithmetic unit (not shown in the figure) that forms an observation image of the sample 20 based on the digital signal output by the AD converter. As the arithmetic unit, for example, an MPU (Micro Processing Unit), a GPU (Graphic Processing Unit), etc. are used. The observation image formed by the image formation unit 31 is transmitted to the display unit 35 for display, or transmitted to the computer system 32 for various processes.

[0025] As described above, the scanning electron microscope 10 includes an electron gun 11, a secondary electron detector 16, and an image formation unit 31. The electron gun 11, the secondary electron detector 16, and the image formation unit 31 (image constructor) constitute the imaging unit in this embodiment. By using such an imaging unit, an image using an electron beam (charged particle beam) can be obtained.

[0026] The computer system 32 includes an interface unit 900 that inputs and outputs data and commands to and from the outside, a processor 901 (for example, a CPU (Central Processing Unit)) that executes various arithmetic processes on the provided information, and a memory 902 and a storage 903 that constitute storage means.

[0027] The storage 903 is composed of, for example, an HDD (Hard Disk Drive), an SSD (Solid State Drive), etc., and stores software 904 that constitutes the field-of-view search tool of the present embodiment and the teacher data DB (database) 44. The software 904 of the present embodiment is a field-of-view search tool. In one example, when the processor 901 executes the software 904, in cooperation with the memory 902, the feature identifier 45 and the image processing unit 34 are configured as functional blocks.

[0028] The feature identifier 45 extracts a landmark pattern 23 (described later with reference to FIG. 2A, etc.) for field-of-view search from the input image data. For example, the feature identifier 45 determines whether the landmark pattern 23 exists at each position of the image data, and outputs the position where it is determined that the landmark pattern 23 exists (or the position where it is determined that the probability of the landmark pattern 23 existing in the image data is the highest). The image processing unit 34 calculates the position coordinates of the landmark pattern 23 with reference to the position information of the sample stage 17 from the position of the detected landmark pattern on the image.

[0029] The memory 902 shown in FIG. 1 represents a state in which each functional block constituting the software 904 is expanded on the memory space. When the software 904 is executed, the processor 901 executes the functions of each functional block expanded in the memory space.

[0030] The feature identifier 45 is a learned model generated by machine learning, and learning is performed using the image data of the landmark pattern 23 stored in the teacher data DB 44 as teacher data. When new image data is input to the feature identifier 45, the position of the landmark pattern learned on the image data is determined, and the position of the landmark pattern (the center coordinates of the landmark pattern in the following example) in the new image data is output. The output center coordinates are used to specify the ROI (Region Of Interest) during field-of-view search. Also, various position information calculated from the center coordinates is transmitted to the control unit 33 and used for driving control of the sample stage 17.

[0031] When the image processing unit 34 automatically executes edge line detection, focus adjustment, aberration correction, etc. of the wafer surface based on image processing in a cross-sectional image with the sample cross-section facing the field of view, it performs processes such as calculation and / or evaluation of image sharpness.

[0032] The control unit 33 is an arithmetic unit that controls each component in FIG. 1 and processes and transmits data formed by each component, and includes, for example, a CPU, an MPU, etc. The input unit 36 is a device that receives input of observation conditions for observing the sample 20 and receives input of commands such as execution and stop of observation, and can be configured by, for example, a keyboard, a mouse, a touch panel, a liquid crystal display, or a combination thereof. A GUI (Graphical User Interface) that constitutes an operation screen for the operator and captured images (observation image data) are displayed on the display unit 35.

[0033] Next, with reference to FIG. 2A, the relative positional relationship between the sample 20 to be observed and the drive axis of the sample stage 17 will be described. FIG. 2A is a perspective view of a wafer sample, which is an example of an object to be observed by the charged particle beam apparatus of the present embodiment.

[0034] In FIG. 2A, the sample 20 is a coupon sample obtained by dicing a wafer, and has a cut surface 21 and an upper surface 22 on which a processing pattern is formed. The sample 20 is produced through a semiconductor device manufacturing process or a process development process, and a fine structure is formed on the cut surface 21. In many cases, the imaging location intended by the operator of the charged particle beam apparatus exists on the cut surface 21.

[0035] A mark pattern 23 is formed on the upper surface 22. The mark pattern 23 is a shape or structure with a size larger than the above-mentioned fine structure, that is, a pattern that can be used as a mark during field search. As the mark pattern 23, for example, a characteristic shape marker for identifying a chip processing area on the wafer, a processing pattern including label information, etc. can be used.

[0036] The orthogonal axes of XYZ shown in FIG. 2A are coordinate axes indicating the relative positional relationship with respect to the electron beam 12 of the sample 20. The traveling direction of the electron beam 12 is the Z axis, the direction parallel to the first tilt axis 61 of the sample stage 17 is the X axis, and the direction parallel to the second tilt axis 62 is the Y axis. In the present embodiment, the sample 20 is placed on the sample stage 17 such that its longitudinal direction is parallel to the X axis.

[0037] When observing the fine shape of the cross-section 21, the electron beam 12 is irradiated from a direction substantially perpendicular to the cross-section 21, and the region of the cross-section observation field of view 24 is observed. However, the manually cut cross-section 21 often is not completely orthogonal to the upper surface 22, and the attachment angle does not always become the same each time the operator installs the sample 20 on the sample stage 17.

[0038] Therefore, the first tilt axis 61 and the second tilt axis 62 are provided on the sample stage 17 as angle adjustment axes for making the cross-section 21 orthogonal to the electron beam 12. The first tilt axis 61 is a drive axis for rotating the sample 20 within the YZ plane. Since the longitudinal direction of the cross-section 21 is the X axis direction, when adjusting the tilt angle of a so-called tilt image in which the sample 20 is tilted and observed from an oblique direction, the rotation angle of the first tilt axis 61 is adjusted. Similarly, the second tilt axis 62 is a drive axis for rotating the sample 20 within the XZ plane. When the field of view is in the facing position with respect to the cross-section 21, by adjusting the rotation angle of the second tilt axis 62, the image can be rotated about an axis in the vertical direction (Y axis direction) passing through the center of the field of view.

[0039] Using FIG. 2B, the configuration of the sample stage 17 will be described. As shown in the figure, the sample 20 is held and fixed on the sample stage 17. The sample stage 17 is provided with a mechanism for rotating the placement surface of the sample 20 around the first tilt axis 61 or the second tilt axis 62, and the rotation angle is controlled by the control unit 33. Although not shown in the figure, the sample stage 17 shown in FIG. 2B is also provided with an X drive axis, a Y drive axis, and a Z drive axis for independently moving the sample placement surface in the XYZ directions, and a rotation axis for rotating the sample placement surface around the Z drive axis. Thus, the scanning area (i.e., the field of view) of the electron beam 12 can be moved and further rotated in the longitudinal direction, the lateral direction, and the height direction of the sample 20. The moving distances of the X drive axis, the Y drive axis, and the Z drive axis are also controlled by the control unit 33.

[0040] In the present embodiment, a feature discriminator 45 that automatically recognizes the mark pattern 23 is constructed from a tilt image obtained by tilting the sample 20 and observing it from an oblique direction, and high-magnification observation is performed at a position separated from the mark pattern 23 by a predetermined distance with the mark pattern 23 as a reference point. Next, the learning procedure in the feature discriminator 45 of the present embodiment will be described with reference to FIGS. 3A, 3B, 4A, 4B, 4C, 4D, 5, 6, 7, and 8.

[0041] To perform the automatic field-of-view search and automatic imaging using the same in the present embodiment, a feature discriminator 45 that detects the mark pattern 23 is constructed. The flowchart of FIG. 3A shows the workflow performed by the operator when constructing the feature discriminator 45.

[0042] After the process of FIG. 3A is started (step S300), the sample 20 is placed on the sample stage 17 in the charged particle beam apparatus shown in FIG. 1 (step S301). Next, the optical conditions for imaging an image serving as teaching data, such as the acceleration voltage and magnification, are set (step S302).

[0043] Thereafter, the tilt angle of the sample stage 17 is set (step S303), and the first observation is performed (step S304). This first observation refers to the first observation operation for the operator to confirm the mark pattern, the processing pattern to be finally observed, etc. The operator may perform it alone, or in some cases, the SEM operator conveys the area to be noted, etc. while looking at the observation screen under the supervision of the process engineer. Through this step S304, the information necessary for generating the teacher image data is acquired. Subsequently, by executing step S305, the teacher image data is generated, and the generated teacher image data is stored in the storage 903 (step S306).

[0044] The workflow showing step S305 in detail is shown in FIG. 3B. From the data obtained by the first observation in step S304, the extraction of the target image 523 (described later in relation to FIG. 5 etc.) is executed by automatic processing (step S305-1). Next, based on the target image 523, the selection and registration of the ROI by the operator are executed (step S305-2), and based on that information, the additional acquisition of the ROI image is executed by automatic processing (step S305-3). Thereafter, in order to obtain appropriate teacher image data based on the acquired ROI image, post-processing of the image is executed (step S305-4). Finally, the generated teacher image data is confirmed by the operator (step S305-5), and the teacher image data that has passed the confirmation is stored in the storage 903 (step S306). The details of each step will be described later.

[0045] The learning procedure of the feature identifier 45 described above will be described in more detail while also using the drawings of the GUI.

[0046] FIG. 4A shows the main GUI 400 displayed on the display unit 35 of the charged particle beam apparatus of the present embodiment. The main GUI in FIG. 4A and similar GUIs in other figures are graphical user interfaces for inputting observation setting parameters.

[0047] An example of a tilt image is displayed on the main GUI. The main GUI shown in Fig. 4A includes, as an example, a main screen 401, a start / stop button 402 for instructing the start / stop of the charged particle beam apparatus, a magnification adjustment bar 403 for displaying and adjusting the observation magnification, a select panel 404 on which item buttons for selecting imaging condition setting items are displayed, an operation panel 405 for adjusting image quality and the stage, a menu button 406 ("Menu") for calling other operation functions, a sub-screen 407 for displaying an image with a wider field of view than the main screen 401, and an image list area 408 for displaying thumbnail images of the captured images. The GUI described above is merely an example configuration, and a GUI with items added other than those described above or with other items replaced is also applicable.

[0048] At the start of step S304, the operator presses a record button 451 ("Record") in the operation panel 405 to start recording time-series image data 50 (described later in relation to Fig. 5 etc.) being observed and time-series data 51 of the operating state of the apparatus (described later in relation to Fig. 5 etc.). The time-series image data 50 is data in which observation image data is arranged in time series and is stored in the computer system 32. Note that, as will be described later, the computer system 32 can execute arithmetic processing relating to the time-series data 51 of the operating state and the observation image data.

[0049] As an example, Fig. 4A shows a state in which a tilt image is displayed on the main screen 401 during observation, and the tilt image includes a cross-section 21, an upper surface 22, and a mark pattern 23. In step S304, the operator checks the mark pattern 23 that is the starting point of the observation location while looking at the tilt image, moves the sample stage 17 based on that as a reference point, or tilts the sample stage 17 so that the cross-section 21 of the sample 20 faces the observation surface, and observes the processing pattern to be acquired. When the first observation in step S304 is completed, the operator presses the record button 451 again to complete the data recording. During data recording, a recording mark 450 is displayed in the upper left corner of the main screen 401, and the display of the recording mark 450 disappears when the data recording is completed.

[0050] After the completion of step S304 (first observation), step S305 (generation of teacher image data) is performed. FIG. 4B shows a configuration example of the GUI screen used by the operator in step S305. When the GUI of FIG. 4B is to be displayed from the state where the GUI of FIG. 4B is not displayed, if one of the options, "Auto FOV search", is selected from the select buttons displayed by pressing the menu button 406 in FIG. 4A, the teacher data generation tool screen shown in FIG. 4B will be pop-up displayed. In the operation of step S305, the tab 510 of "Training Data Generate" is selected.

[0051] On the teacher data generation tool screen shown in FIG. 4B, the operator presses the input button 511 ("Input") in the data selection area ("Data Select"), and selects the time-series image data 50 stored in the storage 903 in step S304 and the time-series data 51 of the operating state. The selected files are displayed in the time-series image data name display column 516 and the time-series data name display column 517 of the operating state. If you want to select another file, press the clear button 512 ("Clear") to clear the registered file, and then select the appropriate file again.

[0052] Subsequently, in the process of step S305-1 (extraction of the target image), when the operator presses the image extraction button 513 ("Image Extract"), the time-series image data 50 and the time-series data 51 of the operating state are automatically acquired. As a result, the target image 523 estimated to be the one the operator is interested in is displayed, and the time-series data 51 of the operating state is output. In this way, in step S305-1, the processor 901 functions as an output unit that quantifies the operating state of the scanning electron microscope 10 and outputs the time-series data 51 of the operating state.

[0053] When the process of extracting the target image in step S305-1 is completed, a message "Done" is displayed in the status display column 518 indicating the operating state.

[0054] FIG. 5 schematically shows the data analysis processing operation executed by the computer system 32 in step S305-1 of extracting a target image. At the top of FIG. 5, time-series image data 50 is shown, indicating that a number of observed images are stored along the time axis. When the time-series image data 50 is in a video format, it becomes a data set of aggregated image data. For example, if it is 30 fps (frames per second), 30 images are included per second.

[0055] When the time-series image data 50 is video data, the time-series image data 50 can be efficiently processed using a known video processing program or the like. Also, when the time-series image data 50 is a set data set of still image data, video processing becomes unnecessary and the processing is simplified.

[0056] At the bottom of FIG. 5, time-series data 51 of the operating state is shown. Here, as an example, time-series data 51-1 of the stage X coordinate, time-series data 51-2 of the stage Y coordinate, and time-series data 51-3 of the observation magnification are shown. The time-series data 51 of the operating state is not limited to what is shown and can be configured as data including at least one of the following. - Position information of the sample stage 17 - Tilt angle information of the sample stage 17 - Rotation angle information of the sample stage 17 - Observation magnification information - Current information of the objective lens 15 - Stigma current information - Accelerating voltage information of the charged particle beam - Evaluation value of the image being observed By using such time-series data 51 of the operating state, various patterns for extracting a target image can be defined.

[0057] In particular, when the time-series data 51 of the operating state includes the evaluation value of the image being observed, the evaluation value can be configured to include at least one of the following. - Sharpness calculated by high-frequency component analysis of the image - Luminance feature amount calculated based on the luminance distribution of the image By using such time-series data 51 of the operating state, an image with high sharpness or an image with an appropriate luminance distribution can be extracted, and subsequent image processing can be performed with higher accuracy.

[0058] In the computer system 32, a specific variation pattern (target event) is set in advance for the time-series data 51 of the operating state. The computer system 32 automatically determines the time that matches the target event based on the time-series data 51 of the operating state.

[0059] In the present embodiment, two conditions for the target event are set: (1) gazing for a certain period of time or more, and (2) increasing the magnification at the same position. (1) The gazing event for a certain period of time or more is defined on the condition that the operator holds the position of the stage and the observation magnification without changing them at all for a certain period of time. This event corresponds to a pattern in which the sample stage stops for a predetermined time (first time) and the magnification during the predetermined time (first time) is fixed. On the other hand, in (2) the magnification increase event at the same position, it is defined on the condition that the operator increases only the observation magnification without changing the X and Y coordinates of the stage at all. This event corresponds to a pattern in which the sample stage 17 stops for a predetermined time (second time) and the magnification changes during the predetermined time (second time). Note that the first time and the second time may be the same time or different times.

[0060] An example in which a variation pattern that matches the event conditions (1) and (2) is searched from the time-series data 51 of the operating state and the time at that time is determined are the times T11, T12, T13, and T14 shown in FIG. 5. The time T11 and the time T13 correspond to (1) the gazing event for a certain period of time or more, and the time T12 and the time T14 correspond to (2) the magnification increase event at the same position. Note that FIG. 5 shows an overview, and the times indicated by the times T11, T12, T13, and T14 are not exact.

[0061] The computer system 32 acquires and outputs, as the target image data, the observation image data corresponding to the times T11, T12, T13, and T14 from the time-series image data 50. The target image data is stored, for example, as the target image 523 in the storage 903. Further, as the device state data attached to each target image 523, the target image attached information 541 in which the time of the event, the coordinates of the stage, the tilt angle information, the observation magnification, etc. are recorded is stored in the storage 903 in a CSV format file.

[0062] By defining such an event, an appropriate target image 523 can be acquired. In the present embodiment, both (1) the time matching the fixation event for a certain period of time or more and (2) the time matching the magnification increase event at the same position are extracted, but only one of them may be extracted.

[0063] Returning to FIG. 4B, the process of step S305-2 (selection and registration of ROI) will be described. In step S305-2, the selection of the ROI is executed using the target image 523 extracted in step S305-1. When the operator presses the ROI selection button 514 ("ROI Select"), the ROI selection GUI shown in FIG. 4C is displayed. In the ROI selection GUI, an image name list 522 of the target image 523 is displayed on the left side of the screen.

[0064] When any target image name in the image name list 522 is selected with the mouse, the image name is reversely displayed, and the selected target image is displayed on the main screen 532. The operator selects, by mouse operation, the area recognized by the operator as the ROI with respect to the target image 523 displayed on the main screen 532 (for example, the area including the mark pattern 23 (i.e., the ROI) is selected using the pointer 531 and the selection tool 530).

[0065] After that, press the registration button 526 ("Register") to register the information of the ROI. In this way, the scanning electron microscope 10 receives the designation of the ROI in the image of interest via the ROI selection GUI. The registered ROI is displayed with a thick frame like the registered ROI 529. When registration is executed here, the image data of the registered ROI 529 is stored in the storage 903, and the ROI - attached information is stored in the storage 903 as information attached to the ROI.

[0066] The ROI - attached information includes the stage coordinates and the tilt angle of the stage in the real space of the ROI, the size of the ROI, the observation magnification, etc., and is stored as a CSV - format file. The ROI - attached information can be the same as, for example, the image - attached information 541 of the image of interest including the ROI, but may be the corrected image - attached information 541 according to the position of the ROI in the image of interest.

[0067] The ROI - attached information may include at least one of the following. - Tilt angle information of the sample stage 17 - Rotation angle information of the sample stage 17 - Observation magnification information In this way, for each ROI, the conditions for imaging the ROI can be memorized, and re - imaging of the ROI or its periphery becomes easy.

[0068] In this way, the scanning electron microscope 10 stores the image data of the ROI 529 and the ROI - attached information. According to such a configuration, the operator can automatically memorize the attached information by designating an arbitrary ROI, and the work efficiency is improved. In particular, when the ROI - attached information includes the position information of the sample stage 17 corresponding to the ROI, the position information of the sample stage 17 is automatically memorized according to the designation of the ROI, and the work efficiency is further improved.

[0069] If an incorrect location is registered, the ROI can be selected with the pointer 531 and the registration can be deleted by pressing the clear button 527 ("Clear"). For the target image on which the ROI selection and registration process has been performed, a processed mark 524 is displayed in the image name list 522. Also, if there are a large number of target images 523 and they cannot be fully displayed in the image name list 522 on the GUI at once, the list can be scrolled and displayed using the scroll button 525.

[0070] After the operator has selected and registered all the desired ROIs and presses the end button 528 ("Exit"), the GUI in FIG. 4C is closed. In conjunction with this, "Done" is displayed in the status display bar 519 in FIG. 4B, and the process of ROI selection and registration in step S305-2 ends.

[0071] Returning to FIG. 4B, the process of step S305-3 (additional acquisition of ROI images) will be described. In step S305-3, additional ROI images are automatically acquired using the ROI information registered in step S305-2 and the automatic imaging function. When the setting button 521 ("Setting") in the GUI of FIG. 4B is pressed, the setting GUI of FIG. 4D is displayed.

[0072] To construct a highly accurate feature discriminator 45, it is desirable to prepare teacher image data acquired under various conditions. The setting GUI in FIG. 4D is provided with a setting panel 533 for the imaging magnification, a setting panel 534 for the first tilt angle of the sample stage (rotation angle around the first tilt axis 61 in FIG. 2B), and a setting panel 535 for the second tilt angle of the sample stage (rotation angle around the second tilt axis 62 in FIG. 2B).

[0073] The operator inputs the start value, end value, and step value between them in each setting panel. For example, if the start value of the imaging magnification is set to "×0.1k", the end value is set to "×1.0k", and the step value is set to "0.1k", a total of 10 imaging magnification conditions from ×0.1k to ×1.0k in 0.1 increments can be set. Note that the unit "k" represents kilo, that is, 1000.

[0074] Similarly, if the set value of the first tilt angle (the rotation angle around the first tilt axis 61 in FIG. 2B) is set in 1° increments from 1° to 5°, five angular conditions can be set. Also, if the second tilt angle (the rotation angle around the second tilt axis 62 in FIG. 2B) is to be fixed at 0°, both the start value and the end value are set to 0°.

[0075] When the above settings are made, conditions combining the imaging magnification and the tilt angle are generated. In this example, a total of 10×5 = 50 imaging conditions are set. These imaging conditions are saved in the storage 903 as additional imaging conditions. That is, the additional imaging conditions of the present embodiment include one or more combinations of the imaging magnification, the first tilt angle, and the second tilt angle. Also, when the automatic radio buttons ("Auto") of the respective setting panels 533, 534, 535 are checked, the default conditions preset for the corresponding setting panel are applied, reducing the operator's workload when there is no need to change the settings each time.

[0076] After completing the setting of the above imaging conditions, returning to the GUI of FIG. 4B again, the operator presses the automatic collection button 515 ("Auto Collect"), and the additional imaging of the ROI in step S305-3 is automatically executed.

[0077] FIG. 6 shows a flowchart of additional acquisition of the ROI image. In step S305-3A, the additional imaging conditions are saved using the setting GUI as described in FIG. 4D. In step S305-3B, the sample stage 17 is moved to the center coordinates of the next ROI (when this step is executed for the first time, it is the first ROI, for example, the ROI described at the head among the plurality of ROIs described in the ROI attached information). In this way, the scanning electron microscope 10 moves the sample stage 17 to a position where the ROI can be imaged using the ROI attached information.

[0078] In step S305-3C, the stage is tilted to the next set angle described in the additional imaging conditions (when this step is executed for the first time, it is the first set angle, for example, the set angle described at the beginning among the set angles described in the additional imaging conditions). In step S305-3D, the observation magnification is set to the lowest magnification of the scanning electron microscope 10.

[0079] In step S305-3E, based on the observation image at the lowest magnification, field center correction is performed. As shown in FIG. 4A, in the tilt image during field observation, the upper surface 22 (wafer surface) and the cross-sectional surface 21 of the sample are observed, and the boundary line between the two can be visually recognized as the edge line. As a field center correction method, the edge line is detected by image processing, and the actual position coordinates of the edge line are calculated from the position information of the edge line on the image and the position information of the sample stage 17, and the sample stage 17 is moved so as to be at the center of the field of view. As an image processing algorithm for detecting the edge line, line detection by Hough transform or the like can be used. Further, in order to improve the detection accuracy, pre-processing such as processing with a Sobel filter may be performed to emphasize the edge line.

[0080] In step S305-3F, the magnification is changed to the next magnification condition described in the additional imaging conditions. Then, in step S305-3G, an ROI-containing image 440 with the ROI captured at the center of the field of view is acquired. In step S305-3H, the imaging magnification is determined. It is determined whether the final magnification (for example, the highest magnification among the magnifications described in the additional imaging conditions) has been reached. If the final magnification has not been reached, the process returns to step S305-3F, the next magnification is set, and the process of acquiring the ROI-containing image 440 is repeated.

[0081] If the final magnification is reached in step S305-3H, the process proceeds to step S305-3I. In step S305-3I, it is determined whether the set angle is the final angle (for example, the larger angle among the tilt angles described in the additional imaging conditions). If it is not the final angle, the process returns to step S305-3C, and after performing field center correction at the lowest magnification at the next set angle, the process of acquiring the ROI-containing image 440 while changing the magnification is repeated.

[0082] Note that, for simplicity of explanation, in FIG. 6, the angle loop (the loop from step S305-3I to step S305-3C) is shown as a single one, but actually loops are executed for each of the first tilt angle and the second tilt angle, resulting in a double loop.

[0083] When the last angle condition is reached in step S305-3I, proceed to step S305-3J. In step S305-3J, determine whether the ROI is the last ROI among the ROIs described in the additional imaging conditions. If it is not the last ROI, return to step S305-3B, and repeat the process of obtaining the ROI-containing image 440 under multiple conditions while changing the tilt and magnification conditions of the stage with the center coordinates of the next ROI.

[0084] In this way, the scanning electron microscope 10 acquires additional image data of the ROI under a plurality of imaging conditions. Note that the plurality of imaging conditions can include conditions where at least one of the magnification, the tilt angle of the sample stage, and the rotation angle of the sample stage is different. Using such a plurality of imaging conditions increases the likelihood of obtaining an image in which the landmark pattern 23 appears appropriately.

[0085] When the last ROI is reached in step S305-3J, the process of acquiring additional ROI images ends (step S305-3K). When the above is completed, in the teacher data generation tool screen of FIG. 4B, the display in the status display bar 520 becomes "Completed" ("Done"), and the process of acquiring additional ROI images in step S305-3 (FIG. 3B) is completed.

[0086] When step S305-3 is completed, next proceed to the post-processing step of the image in step S305-4 (FIG. 3B). The process of step S305-4 will be described with reference to FIGS. 7 and 4B.

[0087] FIG. 7 is a schematic diagram showing the relationship between the ROI-containing image 440 obtained in step S305-3, the correct image 429 of the training data used for constructing the feature discriminator 45, and the incorrect image 430. In the present embodiment, the feature discriminator 45 is constructed by a cascade classifier. In the cascade classifier, two types of image data sets, i.e., the correct image and the incorrect image including the target ROI, are used as the training image data. In step S305-4, in the ROI-containing image 440, the region including the mark pattern 23 or a part thereof is saved as the correct image 429, and the region not including the mark pattern 23 or a part thereof is saved as the incorrect image 430.

[0088] Return to the GUI of FIG. 4B and explain the operation in step S305-4. In the image folder display column 545, when step S305-3 ends, the folder in which the image data composed of a large number of acquired ROI-containing images 440 is saved is displayed.

[0089] If you want to use an image set other than the image set additionally acquired in the immediately preceding step S305-3, click the image folder display column 545 and select an appropriate image folder.

[0090] Thereafter, when the post-processing button 544 ("Post Process") is pressed, as described in FIG. 7, the regions of the correct image 429 and the incorrect image 430 are automatically cut out and saved in a new training image data folder. The automatic cutting process at this time can be appropriately designed based on known techniques. For example, each region may be specified using an algorithm not relying on machine learning, or if a general feature discriminator 45 based on machine learning is available, it may be used. Since it can be corrected in a later step S305-5 when an incorrect specification is made, it is not essential to increase the specification accuracy in this step S305-4.

[0091] The correct answer image 429 and the incorrect answer image 430 are stored in different subfolders. For all the ROI-containing images 440 obtained in step S305-4, the same processing is executed. When the processing is completed, the post-processing step of the image in step S305-4 is completed, and the status display column 546 of the GUI in FIG. 4B becomes "Completed" ("Done").

[0092] Subsequently, the process proceeds to the step of checking the teacher image data in step S305-5 (FIG. 3B). FIG. 8 shows a configuration example of the GUI screen used in the step of S305-5. In order to display the GUI in FIG. 8 from the state where the screen is not displayed, when "Automatic Field of View Search" is selected from the select button displayed by pressing the menu button 406 in FIG. 4A, the learning tool screen shown in FIG. 8 is pop-up displayed.

[0093] In the operation of step S305-5, the learning tab 411 ("Train") is selected. The operator presses the input button 412 ("Input") for data selection ("Data Select") and selects the folder in which the teacher image data set generated in step S305-4 is stored. The selected folder name is displayed in the image folder display column 413. Press the clear button 414 ("Clear") to change the designation of the image folder and start the selection operation again.

[0094] When the folder of the teacher image data is selected, the teacher image data tab 417 ("Folder: Training_data") is displayed at the lower part of the GUI, the correct answer image 429 is displayed on the image display screen 418, and the incorrect answer image 430 is displayed on the image display screen 428. Images that cannot be fully displayed on the image display screens 418 and 428 can be scrolled and displayed by the scroll button 419.

[0095] Each teacher image data is provided with a check mark input column 420, and in the initial setting, check marks 421 are given to all images. The operator checks the displayed teacher image data and deletes the check mark 421 by mouse operation if there is an inappropriate image.

[0096] Once the confirmation is complete, the operator presses the update button 422 ("Update") to update the teacher image data. At this time, the update display button 427 ("Data Update") at the upper part of the GUI blinks to indicate to the operator that the data update has been performed. Also, if the operator wants to reselect the teacher image data multiple times, pressing the reset button 423 ("Reset") initializes the check mark input field 420 and returns it to the state where check marks 421 are assigned to all images.

[0097] When the update button 422 is pressed, all the images with check marks 421 at that time are copied as teacher image data and temporarily stored in the teacher data DB 44. At this time, the content of the original teacher image data set remains unchanged from the state immediately after the input button 412 is pressed, so there is no data corruption when it is reused later. On the other hand, when you want to save the updated teacher image data set as a new data set, you can press the save button 431 ("Save") to save it as a data set with a new name.

[0098] Through the above steps, the confirmation process of the teacher image data in step S305-5 is completed, and accordingly, the storage process of the teacher data in step S306 is also completed (Fig. 3B). If there is no particular need to update the teacher image data displayed by pressing the data selection input button 412, the processes of steps S305-5 and S306 are completed as they are.

[0099] Next, the process of step S307 (learning of the feature identifier) will be described (Fig. 3A). When the setting of the teacher image data is completed, the operator presses the learning button 415 ("Train") at the upper part of the GUI in Fig. 8 or the learning button 424 ("Train") at the lower part of the GUI to start the learning. To the right of the learning button 415 and the learning button 424, a status display bar cell 416 and a status display bar cell 425 indicating the status are respectively displayed. When the learning in step S307 is completed, "Done" is displayed in the status display bar cells 416 and 425. In this way, the process of Fig. 3A is completed (step S308).

[0100] When a plurality of feature identifiers 45 are available, by clicking on the model name input field 426 ("Model Name") and selecting a learned model in the storage, the feature identifier 45 to be used for the field of view search can be selected.

[0101] As a machine learning method, in this embodiment, a cascade classifier is used, but an object detection algorithm using a deep neural network (DNN) can also be used.

[0102] Next, after the learning of the feature identifier 45 is completed, the field of view search and the automatic imaging sequence utilizing it will be described with reference to Figs. 9A to 12. Fig. 9A is a flowchart showing the entire automatic imaging sequence. After the process of Fig. 9A is started (step S500), first, a new sample 20 is placed on the sample stage 17 (step S501), and then, the condition setting during the field of view search is performed (step S502).

[0103] The condition setting step during the field of view search in step S502 is composed of, as shown in Fig. 9B, an optical condition setting step during the field of view search (step S502-1) and a stage condition setting step during the field of view search (step S502-2). In step S502, operations are performed using the GUI600 shown in Fig. 10A and the main GUI400 shown in Fig. 10B, the details of which will be described later.

[0104] When the condition setting for the visual field search is completed, a test run of the visual field search (step S503) is executed (Fig. 9A). The test run is a step of acquiring a tilted image of the sample 20 at a preset magnification and outputting the center coordinates of the landmark pattern from the feature identifier 45. Depending on the number of imaging locations of interest and the set magnification, the tilted image of the sample cross-section may fit into one image or multiple images may need to be captured.

[0105] When capturing multiple images, after the images are acquired, the computer system 32 automatically moves the sample stage 17 by a certain distance in the X-axis direction, then acquires the next image, further moves the sample stage 17 by a certain distance, and further acquires the next image, repeating the image acquisition in this way. For the multiple tilted images acquired in this manner, the feature identifier 45 is operated to detect the landmark pattern 23. The detection result is displayed on the main GUI 400 in a form where a marker (for example, a rectangular frame) indicating the ROI is superimposed on the acquired image. The operator checks whether the ROI of the landmark pattern included in the image can be correctly extracted from the obtained output result.

[0106] If the operator determines that a problem has occurred as a result of the test run, after performing the troubleshooting process in step S504-2, step S503 is re-executed or resumed. Possible problems include, for example, when the landmark pattern 23 in the visual field cannot be found even when the feature identifier 45 is operated and the center coordinates of the landmark pattern 23 are not output, or when an area other than the landmark pattern 23 is misrecognized as the landmark pattern 23 and incorrect center coordinates are output. Also, when a problem related to the imaging device or the entire device, such as an abnormality in the optical system, occurs, the execution process of the test run may be temporarily interrupted.

[0107] If no problem occurs and the test run is successful, the conditions for image auto-capture (i.e., image acquisition at high magnification) are set (step S505). Note that steps S503 of the test run and step S504 for checking for malfunction can also be omitted. After the condition setting during the field of view search (step S502), the process can proceed to the condition setting step for image auto-capture (step S505), and the actual operation can start immediately.

[0108] As shown in FIG. 9C, step S505 includes an optical condition setting step (step S505-1) during high magnification imaging, a stage condition setting step for the alignment condition (step S505-2), and a setting step for the final observation position (step S505-3).

[0109] Here, the GUI used when executing the flowcharts of FIGS. 9B and 9C will be described. FIG. 10A shows the GUI600 used by the operator when setting the conditions for the field of view search (step S502), and FIG. 10B shows an example of the main GUI400 which is the main screen.

[0110] The main GUI400 is the same as the GUI described in FIG. 4A. As described above, when the operator selects the field of view search button from the select buttons displayed by pressing the menu button 406, the screen shown in FIG. 10A is popped up and displayed. If the GUI shown in FIG. 10A is not displayed, selecting the auto recipe tab 601 ("Auto Recipe") will switch the screen to the GUI shown in FIG. 10A.

[0111] The upper part of the GUI600 shown in FIG. 10A can set both the imaging conditions during the field of view search (steps S502 and S506) and the automatic imaging of high magnification images (step S508). By pressing either the radio button in the field of view search column 602 ("FOV search") or the high magnification imaging column 603 ("High magnification capture"), the setting screens for both can be switched.

[0112] Below the radio button, a setting panel for various setting items of imaging conditions is displayed. For example, in the case of FIG. 10A, on the upper part of the GUI600, a stage state setting panel 604, a magnification setting panel 605, a final observation position setting panel 607, etc. are displayed.

[0113] The position number selection column 621 is a column for selecting the registration number of the position where the field of view search is performed. In the present embodiment, two registration numbers, P1 and P2, can be set. When there is only one position where the field of view search is performed, only P1 needs to be set. On the other hand, when the field of view search is performed while scanning the stage within a certain range, by setting both P1 and P2, it is possible to scan between the two points to search for the field of view.

[0114] The imaging number column 630 is a column for selecting a condition set for automatically capturing a high-magnification image. When a condition set is registered, by operating the imaging number column 630, any of the registered condition sets can be referred to.

[0115] The stage state setting panel 604 is a setting column for registering the XYZ coordinate information of the sample stage 17, the first tilt angle (the rotation angle around the first tilt axis 61 in FIG. 2B), and the second tilt angle (the rotation angle around the second tilt axis 62 in FIG. 2B) in the computer system 32. Although a tilted image of the sample cross-section is displayed on the main screen 401 of the main GUI400, in each display column of the X coordinate information, Y coordinate information, Z coordinate information, first tilt angle (the first tilt axis 61 in FIG. 2B), and second tilt angle (the second tilt axis 62 in FIG. 2B) of the stage state setting panel 604, the stage information of the state of the image displayed on the main screen 401 is displayed.

[0116] When the registration button 612 ("Register") is pressed in a state where P1 or P2 is selected in the position number selection column 621, the current stage state (the state of the drive axis) is registered in the computer system 32 as the stage information of the selected position number (P1 or P2).

[0117] For convenience of explanation, FIG. 10A shows a configuration example of a setting panel displayed on both the screens during field-of-view search (steps S502 and S506) and during automatic imaging of a high-magnification image (step S508). In reality, however, only the necessary setting panels corresponding to the selection by radio buttons are displayed on each screen. For example, when the radio button in the field-of-view search column 602 is selected, the position number selection column 621 and the stage state setting panel 604 are displayed. When the radio button in the high-magnification imaging column 603 is selected, the imaging number column 630, the magnification setting panel 605, and the final observation position setting panel 607 are displayed.

[0118] Registration can be canceled by pressing the clear button 613 ("Clear"). The operations of the registration button 612 and the clear button 613 are common in the following description.

[0119] The execution button 614 ("Run") is a button for instructing the computer system 32 to start the field-of-view search. By pressing this button, step S503 (test run) in FIG. 9A can be started.

[0120] The resume button 615 ("Resume") is a button for resuming the process when the process automatically stops due to malfunction or the like in step S504 of FIG. 9A. After the processing in step S504-2 and after the cause of the malfunction has been eliminated, pressing this button can resume the test run process from the step where the process automatically stopped. Pressing the stop button 616 ("Stop") can stop the ongoing field-of-view search halfway.

[0121] When you want to finely adjust the field of view of the tilted image, press the adjustment button 609, and each of the XYZ coordinates or the tilt angle of the sample stage 17 will change in the positive or negative direction. The image after the change is displayed in real time on the main screen 401, and the operator registers the state of the sample stage 17 where the most appropriate field of view is obtained while viewing the image. In addition, if the field of view is adjusted so that the front-facing image of the cross-section 21 is reflected on the main screen 401 with the radio button in the high-magnification imaging field 603 selected, the stage condition in that state represents the stage condition for the front-facing condition. By registering this stage condition in the computer system 32, step S505-2 in FIG. 9C can be executed.

[0122] In addition to the manual adjustment described above, the setting and / or registration of the stage front-facing condition may be automatically adjusted based on a predetermined algorithm. As an algorithm for adjusting the tilt angle of the sample 20, an algorithm that obtains tilt images at various tilt angles and calculates the tilt angle by numerical calculation based on the edge line of the wafer extracted from the image can be adopted.

[0123] The magnification setting panel 605 (Fig. 10A) is a setting field for setting intermediate magnifications when increasing the magnification from the imaging magnification during field-of-view search (i.e., the starting magnification during high-magnification imaging) to the final magnification (the final magnification during high-magnification imaging). In the right column of the location displayed as "Current", the imaging magnification of the tilt image currently displayed on the main screen 401 is shown. To the right of "Final" in the middle section is a setting field for setting the final magnification, and the final magnification is selected using adjustment buttons similar to those on the stage state setting panel 604. The lower section "Step*" is a setting field for setting which step the intermediate magnification is from the imaging magnification of the tilt image. When the adjustment button on the right side of the setting field is operated, a number is displayed in the "*" column. For example, it is displayed as "Step1", "Step2", etc. Further to the right of the adjustment button on the right side of the setting field, a magnification setting field for setting the imaging magnification at each step is shown, and the intermediate magnification is set by operating the adjustment button in the same way. According to such a GUI, when expanding the magnification step by step during high-magnification imaging, it is possible to individually set which step is what magnification. After the setting is completed, when the registration button 612 is pressed, the set final magnification and intermediate magnifications are registered in the computer system 32.

[0124] The final observation position setting panel 607 is a setting field for setting the center position of the field of view when imaging at the final magnification according to the relative position from the mark pattern 23. On the main screen 401, a tilt image of the sample cross-section is shown together with the ROI 25 for mark pattern setting. However, the operator can set the relative position information of the final observation position with respect to the mark pattern 23 by operating the pointer 409 to drag and drop the selection tool 410 to the desired final observation position 436 (see Fig. 10B). In the final observation position setting panel 607, the distance in the X direction from the center coordinates of the ROI 25 is displayed in either the left display column ("Left") or the right display column ("Right"), and the distance in the Z direction is displayed in either the upper display column ("Above") or the lower display column ("Below").

[0125] When setting a plurality of final observation positions, it is set by repeating the drag-and-drop of the selection tool 410. Also, as will be described later, you may directly input a numerical value to the necessary one among the left display column, the right display column, the upper display column, and the lower display column using a keyboard, a numeric keypad, etc. provided in the input unit 36. This method is highly convenient for the operator, for example, when imaging a plurality of images at a fixed interval (for example, an equal interval pitch) from a reference position based on a position separated by a predetermined distance from the mark pattern 23.

[0126] The setting of the optical conditions during the field search and the high-magnification image imaging is performed using the GUI400 which is the main GUI. With the GUI600 being displayed, when a button related to the optical conditions on the selection panel 404 or the operation panel 405 of the GUI400 is pressed, a setting screen for the optical conditions is displayed.

[0127] For example, when the scan button 437 ("Scan") is pressed in the GUI of FIG. 10B, a scan speed setting panel 608 is displayed, and the operator can operate the setting knob 611 while looking at the indicator 610 to set an appropriate value for the scan speed during imaging. After the setting, when the registration button 612 is pressed, the set scan speed is registered in the computer system 32.

[0128] In the above manner, while switching the radio buttons in the field search column 602 and the high-magnification imaging column 603, the optical conditions such as the acceleration voltage and the beam current value are set and registered in the computer system 32, whereby the conditions used in step S502-1 of FIG. 9B and step S505-1 of FIG. 9C can be determined. Note that the scan speed during the imaging of the tilt image can be set to be higher than the scan speed of the image at the final magnification. Also, the scanning electron microscope 10 can switch the scan speed according to the set speed.

[0129] In the above description of the upper part of FIG. 10A, numerical input to the display fields provided in each setting panel (stage state setting panel 604, magnification setting panel 605, and final observation position setting panel 607) can be performed using the adjustment button 609. Instead of or in addition to this, it is also possible to directly input numerical values using a keyboard, numeric keypad, etc. provided in the input unit 36.

[0130] When a plurality of feature discriminators 45 can be used, in the GUI 600 of FIG. 10A, by clicking on the model name input field 624 ("Model Name") and selecting a learned model in the storage, the feature discriminator 45 used for field of view search can be selected.

[0131] When the above-described settings in the upper part of the GUI 600 of FIG. 10A are made, the information is also reflected in the lower area. In the lower part of the GUI 600, the setting conditions for field of view search ("FOV search") and the setting conditions for high-magnification image shooting are listed and displayed.

[0132] Conversely, it is also possible to directly edit the lower setting list in the same way as inputting in the upper panel. For example, if "Fix" is selected in the field of view search mode setting panel 622 and stage coordinates and tilt angles are set in the row of P1 in the search position setting table 625, field of view search at a fixed position can be set (at this time, the P2 row in the search position setting table 625 becomes an input-disabled state).

[0133] Alternatively, if "Scan" is selected in the field of view search mode setting panel 622 and different stage coordinates are set for both P1 and P2 in the search position setting table 625, a process of searching for a mark pattern while scanning from P1 to P2 can be set.

[0134] In the high-magnification imaging recipe setting table 626 as well, multiple imaging locations and imaging conditions can be directly input. For each imaging number ("Capture No"), the relative position from the mark pattern 23 can be set respectively, and observation magnification, tilt angle, etc. can be set, which is convenient when continuously setting high-magnification imaging under a number of conditions. When there are many imaging numbers and the high-magnification imaging recipe setting table 626 cannot be fully displayed within the screen, it can be scrolled and displayed using the scroll button 627.

[0135] Regarding the detection signal as well, it is possible to select whether to use either the secondary electron image (SE image) or the backscattered electron image (BSE image), and it is also possible to capture a shape image and a Z-contrast image in the same field of view. This high-magnification imaging recipe can be loaded by entering the recipe name in the automatic recipe name input field 629 ("Auto Recipe Data") and then pressing the import button 628 ("Import") if it has been previously described in a CSV file or the like, and it will be reflected in the GUI600.

[0136] Returning to FIG. 9A, the description of the flowchart is resumed. When the condition setting for image auto-capture is completed in step S505, the execution of the actual field-of-view search is started (step S506). FIG. 11 shows a configuration example of the GUI used by the operator when executing the actual field-of-view search in the procedure after step S506 of FIG. 9A. The screen switches to the GUI of FIG. 11 when the operator selects from the menu button 406 shown in the main GUI400 or selects the auto-capture tab 619 ("Auto Capture") instead of the auto-recipe tab 601 of the GUI in FIG. 10A. When the operator presses the start button 617, the procedure after step S506 of FIG. 9A is started.

[0137] In step S506, the tilted images of the sample cross-section within the range specified as the imaging condition are captured. The image data obtained from the captured images is sequentially input into the feature discriminator 45, and the center coordinate data of the mark pattern is output. Serial numbers such as ROI1, ROI2, etc. are assigned to the output center coordinate data, and it is stored in the storage 903 together with the aforementioned supplementary information.

[0138] When the field-of-view search is completed, the control unit 33 calculates the movement amount of the sample stage 17 from the current stage position information and the center coordinate data of each ROI, and performs a field-of-view movement to the position of the mark pattern 23 (step S507). After the field-of-view movement, a high-magnification image at the final observation position is acquired according to the high-magnification image auto-capture conditions set in step S506 (step S508). Hereinafter, the details of step S508 will be described with reference to FIG. 9D.

[0139] After performing the field-of-view movement to the position of the mark pattern 23 in step S507 of FIG. 9A, the control unit 33 performs a field-of-view movement to the final observation position according to the relative position information set in the final observation position setting panel 607 of FIG. 10A (step S508-1). Next, in step S508-2, the stage condition is adjusted to the facing condition. In this step, the control unit 33 calculates the stage movement amount from the difference between the stage condition set with the radio button in the high-magnification imaging field 603 of the GUI in FIG. 10A pressed and the stage condition at the end of step S508-1 (or the stage condition set with the radio button in the field-of-view search field 602 pressed), and operates the sample stage 17.

[0140] By executing step S508-1 and step S508-2, the observation field-of-view moves to the final observation position and becomes the facing condition with respect to the sample cross-section, so the magnification is increased in that field-of-view (step S508-3). The magnification is increased step by step according to the intermediate magnification set in the magnification setting panel 605 of FIG. 10A.

[0141] In step S508-4, the computer system 32 performs focus adjustment and aberration correction processing. As an algorithm for the correction processing, an image is acquired while sweeping the current values of the objective lens and the aberration correction coil within a predetermined range, a fast Fourier transform (FFT) or a Wavelet transform is performed on the acquired image to evaluate the image sharpness, and a method of deriving setting conditions with a high score can be used. Other aberration correction processing may be included as necessary. In step S508-5, the computer system 32 performs imaging at the magnified magnification and acquires image data in the current field of view.

[0142] In step S508-6, the computer system 32 performs first field shift correction. The first field shift correction in the present embodiment includes correction processing for the horizontal line of the image and correction processing for the positional shift of the field center, but other necessary field shift correction processing may be performed according to the magnification.

[0143] First, the correction processing for the horizontal line will be described. As shown in FIG. 2A, the observation sample in the present embodiment is a coupon sample, and there are an upper surface 22 (wafer surface) of the coupon sample on which a mark pattern 23 is formed and a cross section 21. In the cross-sectional image of the cross section 21 under the stage facing condition, the upper surface 22 of the coupon sample is visually recognized as an edge line. Therefore, in this step, the edge line is automatically detected from the image data acquired in step S508-5, and the field shift in the XZ plane of the acquired image is corrected so that the edge line coincides with the horizontal line (a virtual horizontal reference line passing through the center of the field of view) in the image. Specifically, the actual position coordinates of the edge line are derived by the processor 901 from the position information of the edge line on the image and the position information of the sample stage 17, and the rotation angle of the first tilt axis is adjusted by the control unit 33 to move the field of view so that the edge line is positioned at the center of the field of view. As an image processing algorithm for detecting the edge line, line detection by Hough transform or the like can be used. Further, in order to improve the detection accuracy, preprocessing such as processing with a Sobel filter may be performed to emphasize the edge line.

[0144] Next, the position deviation correction process for the center of the field of view will be described. Immediately after the field of view movement in step S508-1, the position set on the final observation position panel 607 in FIG. 10A is located at the center of the field of view. However, when the observation magnification is increased in step S508-3, the center of the field of view may shift. Therefore, the computer system 32 extracts image data for an appropriate number of pixels around the center of the field of view from the image before magnification, and uses this image data as a template to perform pattern matching on the image data obtained in step S508-5. The center coordinates of the region detected by the matching are the original center of the field of view. The computer system 32 calculates the difference between the center coordinates of the detected region and the coordinates of the center of the field of view of the image data obtained in step S508-5, and transmits it to the control unit 33 as the control amount for the sample stage 17. The control unit 33 drives the X drive axis or the Y drive axis according to the received control amount, and further drives the second tilt axis depending on the magnification, to correct the shift of the center of the field of view.

[0145] Note that if the computer system 32 is equipped with another feature identifier that has learned the images obtained during the magnification process as teacher data, the coordinate data of the center of the field of view can be obtained by directly inputting the image data obtained in step S508-5 into the other feature identifier without using template matching.

[0146] Also, the correction of the field of view shift in this step may be performed by image shift instead of adjusting the sample stage 17. In that case, the adjustment amount of the field of view shift is converted by the computer system 32 into control information regarding the scanning range of the electron beam in the XY directions, and sent to the control unit 33. The control unit 33 controls the deflection lens 14 based on the received control information, and performs the adjustment of the field of view shift by image shift.

[0147] In step S508-7, it is determined whether the adjustment amount of the first field shift correction executed in step S508-6 is appropriate. In FIG. 2B, since the height of the sample 20 (the distance between the cross-sectional plane 21 and its opposing surface in FIG. 2A) is known, the distance R between the rotation center of the second tilt axis 62 and the cross-sectional plane 21 is also known. When performing field shift correction using the second tilt axis 62 in step S508-6, in principle, θ is adjusted so that the product Rθ of the rotation angle θ of the second tilt axis 62 and the distance R is equal to the field movement amount on the image. However, due to various reasons such as the horizontal accuracy of the wafer mounting surface of the sample stage 17 and the inclination of the cross-sectional plane 21 (resulting from the sample shape), it is difficult to accurately measure R correctly. Therefore, the rotation angle θ calculated in the first field shift correction step may be insufficient or excessive due to the accuracy of R. Also, even in the case of field shift correction by adjusting the X drive axis or the Y drive axis, due to problems such as mechanical accuracy, it may occur that the original field center is not located at the field center in the image after field shift correction. If it is not appropriate, the process proceeds to step S508-8, and if it is appropriate, the process proceeds to step S508-9.

[0148] In step S508-8, a second field shift correction is executed. In the second field shift correction, by image processing, the adjustment amount of the rotation angle θ of the shortage or excess, or the adjustment amounts of the X drive axis and the Y drive axis are obtained, and the sample stage 17 is readjusted. When the original field center is not located at the field center, in step S508-8, the image before the execution of the specified distance movement and the image after the execution of the movement are compared, the actually moved distance is measured, and the shortage amount is added for correction. When there is no object for the above processing in the field of view, the magnification is changed to the low magnification side, and after an object that can be identified in the image is included in the field of view, the above processing is performed.

[0149] Note that the second field shift correction in this step may be executed using image shift instead of adjusting the sample stage 17. In some cases, the first field shift correction process and the second field shift correction process described above are collectively referred to as "fine adjustment".

[0150] In step S508-9, it is determined whether the current imaging magnification matches the final observation magnification set in the magnification setting panel 605 of FIG. 10A. If they match, the process proceeds to the next step S508-10. If they do not match, the process returns to step S508-3, and the processing from step S508-3 to step S508-8 is repeated.

[0151] In step S508-10, the optical conditions at the time of high-magnification image imaging set in the GUI 400 of FIG. 10A are changed, and imaging is performed according to the optical conditions in step S508-11. Thus, step S508 ends, and the process proceeds to step S509 in FIG. 9A.

[0152] In step S509, it is determined whether imaging has been completed for all ROIs extracted in the field-of-view search based on the serial number of the ROI imaged in step S508 at the final observation position. If not, the process returns to step S507 to perform a field-of-view movement to the next ROI. If it has been completed, the automatic imaging process of this embodiment is terminated (step S510).

[0153] During the execution of the automatic imaging process, a status indicating the progress of the automatic imaging process is displayed on the GUI shown in FIG. 11. The status bar 618 displays the ratio of the imaged ROIs to the total number of ROIs. In the details column 620 of the imaged images, the serial number of the image that has not been imaged or is being imaged, the coordinates (stage conditions), and the serial number of the ROI corresponding to the mark pattern at each imaging location are displayed.

[0154] As described above, since the scanning electron microscope 10 executes the automatic imaging process using the feature discriminator 45, the manual field-of-view search operation is unnecessary, and the work efficiency is improved. In addition, since the generation of the feature discriminator 45 is performed based on the attached information 541 of the target image, the scanning electron microscope 10 can execute the automatic imaging process based on the attached information 541 of the target image. According to such a configuration, since the operation of automatic imaging (stage conditions, magnification, etc.) is determined based on the actual sample, an imaging operation suitable for the sample can be realized.

[0155] FIG. 12 shows the state of the main GUI 400 after the sequence of the automatic imaging process is completed. In the main screen 401, the captured high-magnification image is displayed, and in the sub-screen 407, a tilt image of the cross-section 21 with a wider field of view than the main screen 401 is displayed. In the image list area 408, the high-magnification images 439 of each imaging location are displayed as thumbnails. The high-magnification image displayed on the main screen 401 is a cross-sectional image with a high magnification to the extent that the shape of the processing pattern 26 formed on the wafer can be confirmed. In this example, the magnification is ×200k as displayed in the magnification adjustment column 403. Further, in order to highlight the imaging location of the high-magnification image, a marker 438 indicating the mark pattern and the final imaging position is displayed on the sub-screen 407.

[0156] Regarding the charged particle beam apparatus described above, a feature discriminator 45 for the mark pattern 23 was constructed using 200 sets of teacher data and a cascade classifier, and as a result of implementing the flow of FIGS. 9A to 9D as an automatic imaging sequence in the apparatus, a good automatic cross-section observation operation was confirmed. Also, the manual working time required for preparing the teacher image data was 640 minutes in the conventional fully manual method, whereas it was only 5 minutes in the semi-automated method of the present embodiment, and the present embodiment obtained the effect of reducing the working burden to 1 / 128.

[0157] As described above, according to the scanning electron microscope 10 and the method for outputting the target image data according to the present embodiment, the teacher image data necessary for the learning of the feature discriminator 45 can be generated semi-automatically. As a result, the time and labor for field search during sample observation can be significantly reduced, and automatic imaging of the cross-sectional image becomes possible.

[0158] In this embodiment, a configuration for semi-automating the generation of teacher image data in step S305 has been shown. However, in some cases, the operator may manually generate additional teacher image data. Even in that case, the main GUI in FIG. 4A can be used. The operator selects, with the pointer 409 and the selection tool 410, an area including the mark pattern 23 that the feature identifier 45 is to automatically detect on the tilt image displayed on the main screen 401. When selecting or editing an image, the operator presses the edit button ("Edit") in the operation panel 405 in the GUI shown in FIG. 4A. When this button is pressed, editing tools for image data such as "Cut", "Copy", or "Save" are displayed on the screen, and further, the pointer 409 and the selection tool 410 are displayed in the main screen 401.

[0159] FIG. 4A shows a state where one ROI is selected, and a marker indicating the ROI 25 is displayed on the main screen 401. The operator uses these editing tools to cut out the selected area from the image data of the tilt image and save it as image data in the storage 903 (step S306 in FIG. 3A). The saved image data becomes the teacher image data used for machine learning. Although only one ROI is selected in FIG. 4A, multiple ROIs may be selected in one image. Note that when saving, not only the image data but also additional information such as optical conditions at the time of imaging, such as magnification and scanning conditions, and stage conditions (conditions related to the setting of the sample stage 17) can be saved in the storage 903.

[0160] Also, in this embodiment, in step S305-3 of FIG. 3B, additional acquisition of ROI images was performed to increase the quantity of teacher image data. However, if a sufficient amount of teacher image data for constructing the feature discriminator 45 was obtained by the selection and registration of the ROI in step S305-2, step S305-3 is not necessarily required. Further, when additional acquisition of ROI images is performed, it is also possible to use only the image data related to the additional acquisition as teacher image data (that is, it is not necessary to use the image data acquired in step S305-2). Thus, the feature discriminator 45 can be generated by machine learning using an image set including at least one of the ROI image data or the additional ROI image data as teacher data.

[0161] Also, in this embodiment, although the time-series image data 50 has been described assuming it is video data, the time-series image data 50 may be a data set of a plurality of images obtained by continuously capturing observation images at regular time intervals.

[0162] Also, in this embodiment, as the types of data in the time-series data 51 of the operating state, the stage X coordinate, the stage Y coordinate, and the observation magnification have been collected. However, further, as other state data, the first tilt angle of the stage, the second tilt angle of the stage, the current information of the objective lens, the sigma current information, the evaluation value of the image being observed, etc. may be collected. As the evaluation value of the image being observed, for example, sharpness calculated by high-frequency component analysis of the image, luminance feature amounts such as luminance average or luminance dispersion calculated from the luminance distribution of the image, etc. can be used.

[0163] [Second Embodiment] Next, a scanning electron microscope according to the second embodiment will be described. The second embodiment proposes a scanning electron microscope provided with a sample stage 17 having a structure different from that of the first embodiment. The target sample, the flow of automatic imaging, and the method of constructing the field recognition function are the same as those of the first embodiment, but the configuration of the sample stage 17 is different.

[0164] FIG. 13A shows a schematic diagram of the sample stage 17. In the present embodiment, the second tilt axis 62 is provided along the Z-axis direction in the drawing. Further, the first tilt axis 61 is installed on the base 17X below the sample stage 17 provided with the second tilt axis 62. By means of a fixing jig, the upper surface 22 of the sample 20 is fixed so as to be orthogonal to the upper surface of the sample stage 17.

[0165] FIG. 13B shows the state after rotating the second tilt axis 62 by 90° from the state of FIG. 13A (the X-Z plane is parallel to the paper surface) (the Y-Z plane is parallel to the paper surface). In this arrangement, the upper surface 22 of the sample 20 and the first tilt axis 61 are orthogonal.

[0166] By rotating the first tilt axis 61 from this state, the inclination of the cross-sectional plane 21 with respect to the electron beam 12 can be adjusted. Also, when acquiring a tilt image for searching the mark pattern 23, after returning to the state of FIG. 13A, the tilt image can be observed by rotating the first tilt axis 61. Although not shown, the sample stage of the present embodiment is provided with an X drive axis and a Y drive axis for independently moving the sample placement surface in the XY directions, and the observation field of view can be translated in the longitudinal direction of the sample.

[0167] [Third Embodiment] Next, a scanning electron microscope according to the third embodiment will be described. In the present embodiment, two feature discriminators 45 are provided to perform field search in a low magnification image and field search in a high magnification image. Since the overall configuration of the charged particle beam apparatus in which the automatic imaging sequence of the present embodiment is executed and the GUI used by the operator are the same as those in the first embodiment, duplicate explanations may be omitted in the following description. While appropriately referring to FIGS. 3A - B, FIGS. 9A - D, and FIG. 10A as necessary, the description will focus on the differences.

[0168] In this embodiment, the construction flow of the feature identifier 45 is the same as that of the first embodiment shown in FIGS. 3A and 3B, except that there are two objects to be automatically recognized. In this embodiment, a first feature identifier for automatically detecting the mark pattern 23 at a low magnification is constructed in the same manner as in the first embodiment, and further, a second feature identifier for automatically detecting the processing pattern 26 (FIG. 12) at the final observation magnification is also constructed in the same manner as in the first embodiment.

[0169] In this embodiment, up to the middle of the automatic imaging sequence is the same as the flow shown in FIG. 9A. In step S506 of FIG. 9A, using the constructed first feature identifier, a field search is performed with the mark pattern 23 as the target object. In this embodiment, the processing executed during high magnification imaging at the final observation position in step S508 is partially different from that of the first embodiment.

[0170] FIG. 14 shows a flowchart of the main part of the automatic imaging sequence in this embodiment. The processing from step S508-1 to step S508-10 is the same as the flowchart of the first embodiment (FIG. 9D).

[0171] Thereafter, in S508-10A, using the second feature identifier, a second field search is performed with the processing pattern 26 as the target object (ROI). If the processing pattern 26 is not found in the observed field of view for some reason such as dust adhering to the processing pattern or the position of the processing pattern deviating from the assumption (step S508-10B: NO), the sample stage 17 is moved in the X or Y direction by one field of view or a preset distance (step S508-10C). Then, the second field search is performed again (step S508-10A). This process is repeated until the processing pattern 26 of the target object is confirmed. If the ROI is detected in the determination of step S508-10A (step S508-10B: YES), high magnification cross-sectional image acquisition is performed in step S508-11.

[0172] In this embodiment, the setting of the second field-of-view search can be executed by the GUI in FIG. 10A. In the high-magnification imaging recipe setting table 626 at the lower part of FIG. 10A, there is provided a feature discriminator setting column 623 ("Model") for setting a second feature discriminator used for high-magnification observation. When an operator sets a previously constructed second feature discriminator in the feature discriminator setting column 623, when the automatic imaging sequence is executed, according to the processing of the flowchart in FIG. 14, the second field-of-view search (step S508-10A) is executed.

[0173] Note that when using a sequence that does not execute the second field-of-view search as in the first embodiment, the scanner does not set anything in the feature discriminator setting column 623 for high-magnification observation, and only sets the feature discriminator in the model name input column 624 of the upper field-of-view search setting panel ("FOV search").

[0174] This embodiment is effective, for example, in preventing a situation where unexpected dust adheres to the cross-section of the processing pattern and the shape of the processing pattern cannot be seen in the automatically captured image. Alternatively, it is also effective when automatically observing not only the cross-sectional image facing the cross-section 21 but also the tilted image with the sample slightly tilted at a high magnification. In the observation of the tilted image at a high magnification, due to the tilt of the sample from the facing state, the observation location often goes out of the field of view. Therefore, the function of automatically searching for the observation field of view after tilting is effective in reducing the work burden on the operator.

[0175] [Fourth Embodiment] Next, a scanning electron microscope according to the fourth embodiment will be described. In this embodiment, an example of an observation method in which the present invention is applied to the observation of a metal material structure instead of a semiconductor sample will be described.

[0176] In the cross-section of the metal material structure, features such as the types and distributions of multiple different phases, the shapes and compositions of each phase appear, and the operator focuses on them to select a field of view and obtain an observation image. In the case of a metal structure with a complex structure, there are often multiple regions that should be focused on. Especially in the experiments of material development, in order to observe the metal material structure under new manufacturing conditions, in many cases, it is only after observation that it becomes clear what kind of structural features it has.

[0177] In such a case, the operator observes various locations within the observation sample (first observation), first grasps the overall image of what kind of structure is formed, then determines the tissue features to be focused on, and obtains an observation image in a field of view that includes the region where it appears. It is often the case that the operator returns to the position where the tissue features to be focused on appeared and conducts a detailed observation, and it takes time to search for that field of view.

[0178] Also, during the first observation, when the operator is observing while being confused as to whether a certain tissue feature is worthy of detailed analysis and it is later determined that it was a region that should be analyzed in detail, there may be cases where the record of the position where the tissue appeared is forgotten and the operator cannot return. In order to efficiently proceed with the observation of such a material structure within limited machine time, a function that can extract the field of view focused on by the operator later is effective, and the technology of the present disclosure is applicable. Hereinafter, the present embodiment will be described, but the assumed device configuration is the same as that in FIG. 1 and is the same as each of the above-described embodiments.

[0179] As an example of observing a metal material, FIG. 15 shows a schematic cross-sectional view of a polycrystalline structure 71 of a metal material having a polycrystalline structure. As shown in the figure, in this polycrystalline structure, in addition to the main phase, a plurality of different phases coexist. In such a structure, while the operator confirms features such as the region of the first main phase 81, the region of the first different phase 91, the region of the third different phase 93 surrounded by the second different phase 92, the region of the fourth different phase 94 surrounded by the second different phase 92, the region of the second different phase 92 existing alone, the region of the second main phase 82 with a constricted shape, the region of the third main phase 83 with a large aspect ratio, etc., the operator proceeds with the observation of the entire sample (first observation) while moving to other fields of view while considering which region should be analyzed in particular in detail.

[0180] When performing the initial observation, in this embodiment, the operator presses the recording button 451 on the main GUI in FIG. 4A to start recording the time-series image data 50 during observation and the time-series data 51 of the operating state of the apparatus.

[0181] When the initial observation is completed, the operator presses the image extraction button 513 on the GUI in FIG. 4B. As a result, a target image 523 estimated to be the one the operator focused on is output according to the algorithm described in FIG. 5.

[0182] After that, similar to the first embodiment, the operator presses the ROI selection button 514 on the GUI in FIG. 4B to activate the ROI selection GUI in FIG. 4C. Then, after selecting and registering the ROI, the operator presses the setting button 521 on the GUI in FIG. 4B to activate the GUI for setting additional imaging in FIG. 4D. After setting the observation magnification there, by pressing the automatic collection button 515 in FIG. 4B, an observation image of the region including the target tissue can be automatically acquired.

[0183] Although this embodiment is not aimed at semi-automatically generating teacher image data like the first embodiment, by outputting target image data with a scanning electron microscope, additional observation of the target area can also be automated in the observation of metal material structures, and the workload of the operator can be reduced. That is, even when the feature discriminator 45 is not generated, the workload can be reduced.

[0184] Note that the present invention is not limited to the above-described embodiments and includes various modifications. For example, the above-described embodiments have been described in detail for easy understanding of the present invention and are not necessarily limited to those having all the configurations described. Also, a part of the configuration of one embodiment can be replaced with the configuration of another embodiment, or the configuration of another embodiment can be added to the configuration of one embodiment. Also, for a part of the configuration of each embodiment, addition, deletion, or replacement with other configurations is possible. Further, the above-described configurations, functions, processing units, processing means, etc. may be realized in hardware, for example, by designing a part or all of them with an integrated circuit.

Explanation of Symbols

[0185] 10…Scanning electron microscope (charged particle beam device) 11…Electron gun (imaging unit) 12…Electron beam 13…Focusing lens 14…Deflection lens 15…Objective lens 16…Secondary electron detector (detector, imaging unit) 17…Sample stage 20…Sample 21…Cross-section 22…Upper surface 23…Mark pattern 24…Cross-sectional observation field of view 25…ROI (Region of Interest) 26…Processing pattern 31…Image forming unit (image composer, imaging unit) 32…Computer system 33…Control unit 34…Image processing unit 35…Display unit 36…Input unit 44…Teacher image database 45…Feature identifier 50…Time-series image data 51…Time-series data of operating state 61…First tilt axis 62…Second tilt axis 71…Polycrystalline structure 81…First main phase 82…Second main phase 83…Third main phase 91…First secondary phase 92…Second secondary phase 93…Third secondary phase 94…Fourth secondary phase 400…Main GUI (Graphical User Interface) 523…Image of interest (image data of interest) 541…Additional information of the image of interest 600…GUI (Graphical User Interface) 901…Processor (Output Unit)

Claims

1. A charged particle beam device, comprising a sample stage for moving a sample, an imaging unit for acquiring observation image data of the sample, an output unit for quantifying the operating state of the charged particle beam device and outputting time-series data of the operating state, a display unit for displaying the observation image data and displaying a graphical user interface for inputting observation setting parameters, a computer system for storing time-series image data obtained by arranging the observation image data in time series and performing arithmetic processing on the time-series data of the operating state and the observation image data, provided with, based on the time-series data of the operating state, automatically determining a time that matches a preset specific variation pattern, and acquiring and outputting, as attention image data, the observation image data corresponding to the time from the time-series image data, A charged particle beam device, characterized in that.

2. In the charged particle beam device according to Claim 1, the imaging unit, - an electron gun for irradiating the sample with a charged particle beam, - a detector, - an image composer, A charged particle beam device, characterized by comprising.

3. In the charged particle beam device according to Claim 1, the specific variation pattern is, - a pattern in which the sample stage stops for a predetermined first time and the magnification is fixed during the first time, or - a pattern in which the sample stage stops for a predetermined second time and the magnification changes during the second time, A charged particle beam device, characterized by including at least one of.

4. In the charged particle beam device according to claim 1, the time-series data of the operating state is - the position information of the sample stage, - the tilt angle information of the sample stage, - the rotation angle information of the sample stage, - the magnification information, - the current information of the objective lens, - the stigmator current information, - the acceleration voltage information of the charged particle beam, - the evaluation value of the image during observation, and characterized in that it includes at least one of the above. A charged particle beam device.

5. In the charged particle beam device according to claim 4, the time-series data of the operating state includes the evaluation value of the image during observation, the evaluation value of the image during observation is - the sharpness calculated by high-frequency component analysis of the image, or - the luminance feature amount calculated based on the luminance distribution of the image, and characterized in that it includes at least one of the above. A charged particle beam device.

6. In the charged particle beam device according to claim 1, the time-series image data is video data. A charged particle beam device characterized by this.

7. In the charged particle beam device according to claim 1, the time-series image data is a set data set of still image data. A charged particle beam device characterized by this.

8. In the charged particle beam device according to claim 1, the graphical user interface displays the target image data, the charged particle beam device receives, via the graphical user interface, the designation of a region of interest in the image related to the target image data, and the charged particle beam device stores the image data of the region of interest and the incidental information of the region of interest. A charged particle beam device characterized by this.

9. In the charged particle beam device according to claim 8, the additional information of the region of interest includes position information of the sample stage corresponding to the region of interest. A charged particle beam device characterized by this.

10. In the charged particle beam device according to claim 8, the additional information of the region of interest is - tilt angle information of the sample stage, - rotation angle information of the sample stage, or - magnification information, A charged particle beam device characterized by including at least one of these.

11. In the charged particle beam device according to claim 8, the charged particle beam device moves the sample stage to a position where the region of interest can be imaged using the additional information of the region of interest, the charged particle beam device acquires additional image data of the region of interest under a plurality of imaging conditions, the plurality of imaging conditions include a plurality of imaging conditions in which at least one of magnification, tilt angle of the sample stage, or rotation angle of the sample stage is different, A charged particle beam device characterized by this.

12. In the charged particle beam device according to claim 11, the charged particle beam device includes a feature discriminator, and the feature discriminator is generated by machine learning using, as teacher data, an image set including at least one of the image data of the region of interest or the additional image data of the region of interest, the charged particle beam device executes automatic imaging processing using the feature discriminator, A charged particle beam device characterized by this.

13. In the charged particle beam device according to claim 1, a charged particle beam device characterized by executing automatic imaging processing based on the additional information of the target image data.

14. A method for outputting target image data by a charged particle beam device, wherein the charged particle beam device includes a sample stage for moving a sample, an imaging unit for acquiring observation image data of the sample, an output unit for quantifying the operating state of the charged particle beam device and outputting time-series data of the operating state, a display unit for displaying the observation image data and displaying a graphical user interface for inputting observation setting parameters, and a computer system for storing time-series image data obtained by arranging the observation image data in time series, and performing arithmetic processing on the time-series data of the operating state and the observation image data, and comprising the method comprising automatically determining a time that matches a preset specific variation pattern based on the time-series data of the operating state, and acquiring and outputting, as target image data, the observation image data corresponding to the time from the time-series image data, characterized by comprising the above steps.

Images