Collision determination device, program, and collision determination method

The collision determination device uses three-dimensional models to accurately assess collisions in charged particle beam devices, improving detection precision and efficiency by employing point cloud and voxel group information.

JP7583011B2Active Publication Date: 2024-11-13JEOL LTD
View PDF 9 Cites 0 Cited by

Patent Information

Application Number
JP2022205954
Authority / Receiving Office
JP · JP
Patent Type
Patents
Current Assignee / Owner
Filing Date
2022-12-22
Publication Date
2024-11-13
Estimated Expiration
2042-12-22

AI Technical Summary

Technical Problem

Conventional collision determination methods in charged particle beam devices only consider two-dimensional planes, leading to inaccurate collision assessments and increased determination time, as they fail to account for depth and actual structural interactions.

Method used

A collision determination device that utilizes point cloud information and voxel group information to represent three-dimensional shapes of dynamic and static objects within the sample chamber, employing a three-dimensional orthogonal coordinate system to accurately determine overlaps and collisions.

Benefits of technology

Enhances collision detection accuracy and reduces determination time by considering three-dimensional structures and their movements, preventing collisions before actual movement occurs.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 0007583011000001
    Figure 0007583011000001
  • Figure 0007583011000002
    Figure 0007583011000002
  • Figure 0007583011000003
    Figure 0007583011000003
Patent Text Reader

Abstract

To determine whether structures collide against each other more accurately than when determining the collision on a two-dimensional plane and to shorten the time required for the determination.SOLUTION: Based upon a three-dimensional model for a dynamic object which shifts in position, point group information is generated which represents a three-dimensional shape of the dynamic object by a point group. Based upon a three-dimensional model for a static object which does not shift in position, point group information is generated which represents a three-dimensional shape of the static object by a point group. Based upon the point group information on the static object, databased voxel group information is generated which represents a three-dimensional shape of the static object by a voxel group. For each coordinate axis, voxel group information representing a static object present on a movement path of the dynamic object is collated with point group information representing the dynamic object so as to determine whether the dynamic object and static object overlap with each other.SELECTED DRAWING: Figure 8
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] The present invention relates to a technique for determining a collision between structures present in a sample chamber of a charged particle beam device. [Background technology]

[0002] Charged particle beam devices such as electron microscopes (e.g., scanning electron microscopes (SEM) and transmission electron microscopes (TEM)), focused ion beam (FIB) devices, and Auger electron spectroscopy (AES) devices are known.

[0003] There is a known technique for determining whether or not structures in a sample chamber of a charged particle beam device collide with each other. Hereinafter, a collision determination method according to the conventional technique will be described with reference to Fig. 36 to Fig. 38. An X-axis, a Y-axis, and a Z-axis that are orthogonal to each other are defined.

[0004] Fig. 36 shows an objective lens 500, a sample holder 502, and a sample 504 mounted on the sample holder 502. The objective lens 500, the sample holder 502, and the sample 504 are examples of structures present in the sample chamber. Fig. 36 shows the structures as viewed in the YZ plane.

[0005] Conventionally, in a two-dimensional YZ plane, the objective lens 500 is represented by a polygon, and the sample holder 502 is represented by a rectangle, and it is determined whether or not the objective lens 500 and the sample holder 502 collide with each other in the YZ plane. Also, one height is set as the height of the sample 504, and it is determined whether or not the objective lens 500 and the sample 504 collide with each other in the YZ plane. When multiple samples are mounted on the sample holder 502, the height of the highest sample among the multiple samples is set as the height of the sample, and it is determined whether or not the objective lens 500 and the sample collide with each other.

[0006] Shown in Figure 37 are a sample holder 502, a sample 504, a stage 506 and a detector 508. Shown in Figure 37 is the structure as viewed in the YZ plane.

[0007] The detector 508 is an example of a structure that moves back and forth. Such a detector 508 is represented by two types of rectangles, one side of which is in contact with the Z axis, in the YZ plane. Also, in the YZ plane, the sample holder 502 is represented by a rectangle, and the stage 506 is represented by two types of rectangles and one type of circle. In the conventional technology, it is determined whether or not these structures collide in the YZ plane.

[0008] Shown in Figure 38 are the sample holder 502, the walls 510 of the sample chamber, and the stage 512. Shown in Figure 38 is the structure as viewed in the XY plane.

[0009] In the XY plane, the wall surface 510 of the sample chamber is expressed by a combination of arcs and line segments, and the sample holder 502 is expressed by a circle. In the conventional technology, the movable ranges of the sample holder 502 and the stage 512 are determined in advance, and it is determined whether or not these structures collide in the XY plane.

[0010] Patent Documents 1 and 2 describe systems that represent objects using voxels and determine whether or not the objects collide with each other. [Prior art documents] [Patent documents]

[0011] [Patent Document 1] Patent No. 2726811 [Patent Document 2] Patent Publication No. 2021-24019 Summary of the Invention [Problem to be solved by the invention]

[0012] In the conventional technology, the presence or absence of a collision between structures is judged within a two-dimensional plane, and the depth is not taken into consideration. Therefore, when it is judged that structures overlap on a two-dimensional plane, it is judged that the structures collide with each other even if they do not actually collide with each other.

[0013] In addition, in the conventional technology, the presence or absence of a collision is determined using a predetermined two-dimensional shape. For example, a shape larger than the actual structure is determined to determine the presence or absence of a collision. Therefore, even if the structures are not actually colliding with each other, it may be determined that the structures are colliding with each other.

[0014] Furthermore, if it were possible to determine whether or not a collision will occur before a structure is actually moved, it would be possible to prevent collisions between structures from occurring. Therefore, it is desirable to be able to determine whether or not a collision will occur before a structure is actually moved.

[0015] The object of the present invention is to determine the presence or absence of a collision between structures more accurately than when the presence or absence of a collision between structures is determined on a two-dimensional plane. Also, even when the presence or absence of a collision is determined using a three-dimensional model, the object is to shorten the time required to determine the presence or absence of a collision compared to when the present invention is not implemented. Also, the object is to prevent collisions between structures in advance. [Means for solving the problem]

[0016] One aspect of the present invention is a collision determination device comprising: a determination means for determining whether or not the dynamic object overlaps with the static object by comparing point cloud information representing the three-dimensional shape of a dynamic object whose position changes within a sample chamber of a charged particle beam device as a point cloud, the point cloud information including position information of the dynamic object, with voxel group information representing the three-dimensional shape of a static object whose position does not change within the sample chamber as a voxel group, the voxel group information including the position information of the static object and stored in a database; and an output means for outputting the determination result of the determination means, wherein a coordinate system including a three-dimensional orthogonal coordinate system composed of three mutually perpendicular coordinate axes is defined, and the determination means determines whether or not the dynamic object overlaps with the static object by comparing, for each coordinate axis, voxel group information representing the static object present on the movement path of the dynamic object with point cloud information representing the dynamic object.

[0017] The coordinate system may further include a tilt axis of the dynamic object, and the determination means may determine whether or not there is overlap between the dynamic object and the static object by comparing, for each coordinate axis, voxel group information representing the static objects present on the movement path of the dynamic object with point cloud information representing the dynamic object while the dynamic object is tilted around the tilt axis.

[0018] The coordinate system may further include a rotation axis of the dynamic object, and point cloud information representing the dynamic object is generated for each predetermined range of rotation angles, and the determination means may determine whether or not there is overlap between the dynamic object and the static object by comparing point cloud information corresponding to the range of rotation angles of the dynamic object with voxel group information representing the static object.

[0019] The determination means may determine whether or not the dynamic object and the static object overlap by comparing, in order, for each coordinate axis, voxel group information representing the static object present on the movement path of the dynamic object with point cloud information representing the dynamic object, and if it is determined that the dynamic object and the static object overlap, may change the order of the coordinate axes to be compared and determine whether or not the dynamic object and the static object overlap by comparing, for each coordinate axis, voxel group information representing the static object present on the movement path of the dynamic object with point cloud information representing the dynamic object.

[0020] Voxel group information is generated for each state of the static object, and the determination means may further determine whether or not there is overlap between the dynamic object and the static object by comparing the voxel group information corresponding to the state of the static object with point cloud information representing the dynamic object.

[0021] The static object may include a structure that moves back and forth between a first position and a second position, and the state of the static object may be a state in which the structure is at the first position, a state in which the structure is at the second position, or a state in which the structure is moving back and forth between the first position and the second position.

[0022] The determination means may further determine whether or not there is overlap between the dynamic object and the static object by comparing point cloud information representing the dynamic object according to the conditions of observation using the charged particle beam device with voxel group information representing the static object.

[0023] The observation conditions may be conditions related to a voltage applied to a sample that is an object of observation.

[0024] The collision determination device may further include a user interface that accepts input of the size of a sample to be mounted on the dynamic object, and the determination means may determine whether or not there is overlap between the dynamic object and the static object by comparing point cloud information representing the point cloud of the dynamic object reflecting the size accepted by the user interface with voxel group information representing the static object.

[0025] The collision determination device may further include a calculation means for calculating a height of the sample based on a tilt operation and a movement operation in a vertical direction of the dynamic object on which the sample is mounted, and the determination means may determine whether or not there is an overlap between the dynamic object and the static object by comparing point cloud information representing a point cloud of the dynamic object reflecting the height calculated by the calculation means with voxel group information representing the static object.

[0026] One aspect of the present invention is a program that causes a computer to function as a determination means for determining whether or not the dynamic object overlaps with the static object by comparing point cloud information representing the three-dimensional shape of a dynamic object whose position changes within a sample chamber of a charged particle beam device as a point cloud, the point cloud information including position information of the dynamic object, with voxel group information representing the three-dimensional shape of a static object whose position does not change within the sample chamber as a voxel group, the voxel group information including the position information of the static object and stored in a database, wherein a coordinate system including a three-dimensional orthogonal coordinate system composed of three mutually perpendicular coordinate axes is defined, and the determination means determines whether or not the dynamic object overlaps with the static object by comparing, for each coordinate axis, voxel group information representing the static object with point cloud information representing the dynamic object.

[0027] One aspect of the present invention is a collision detection method comprising: generating point cloud information representing the three-dimensional shape of a dynamic object as a point cloud based on a three-dimensional model representing the three-dimensional shape of the dynamic object whose position changes within a sample chamber of a charged particle beam device, the point cloud information including position information of the dynamic object; generating point cloud information representing the three-dimensional shape of the static object as a point cloud based on a three-dimensional model representing the three-dimensional shape of a static object whose position does not change within the sample chamber, the point cloud information including position information of the static object; generating databased voxel group information representing the three-dimensional shape of the static object as a voxel group based on the point cloud information of the static object; determining whether or not the dynamic object overlaps with the static object by comparing, for each coordinate axis, voxel group information representing the static object present on the movement path of the dynamic object with point cloud information representing the dynamic object. Effect of the Invention

[0028] According to the present invention, it is possible to more accurately determine whether or not a collision has occurred between structures compared to determining whether or not a collision has occurred between structures on a two-dimensional plane. Even when determining whether or not a collision has occurred using a three-dimensional model, it is possible to reduce the time required to determine whether or not a collision has occurred compared to when the present invention is not implemented. Furthermore, it is possible to prevent collisions between structures from occurring. [Brief description of the drawings]

[0029] [Figure 1] 1 is a block diagram showing a functional configuration of a collision determination device according to an embodiment; [Diagram 2] 1 is a block diagram showing a hardware configuration of a collision determination device according to an embodiment; [Diagram 3] FIG. 2 shows a three-dimensional model of an objective lens and a three-dimensional model of a sample holder. [Figure 4] FIG. 1 shows a point cloud of an objective lens and a point cloud of a sample holder. [Diagram 5] FIG. 2 is a diagram showing a group of voxels of an objective lens. [Figure 6] FIG. 13 is a diagram showing voxel group information stored in a database. [Figure 7] FIG. 2 is a perspective view showing an actual objective lens and a sample holder. [Figure 8] FIG. 1 shows a voxel group of an objective lens and a point group of a sample holder. [Figure 9] FIG. 2 shows a three-dimensional model of a static object and a three-dimensional model of a dynamic object. [Figure 10] FIG. 2 is a diagram illustrating a configuration of a dynamic object. [Figure 11] FIG. 1 is a diagram showing a structure on the YZ plane. [Figure 12] FIG. 2 is a perspective view showing an example of a sample holder. [Figure 13] FIG. 13 is a diagram showing a screen that is an example of a user interface. [Figure 14] FIG. 13 is a diagram showing a screen that is an example of a user interface. [Figure 15] FIG. 2 is a perspective view showing an example of a sample holder. [Figure 16] FIG. 2 shows a sample holder and a sample. [Figure 17] FIG. 2 shows a sample holder and a sample. [Figure 18] FIG. 2 shows a sample holder and a sample. [Figure 19] FIG. [Figure 20] FIG. [Figure 21] FIG. 2 shows a three-dimensional model of a static object and a three-dimensional model of a dynamic object. [Figure 22] FIG. 2 shows a three-dimensional model of a static object and a three-dimensional model of a dynamic object. [Diagram 23] FIG. 2 shows a three-dimensional model of a static object and a three-dimensional model of a dynamic object. [Figure 24] FIG. 2 illustrates a point cloud of a static object. [Diagram 25] FIG. 1 is a diagram showing a structure as viewed in the XZ plane. [Figure 26] FIG. 13 is a diagram showing a movement path in the XZ plane or the YZ plane. [Figure 27] FIG. 2 is a diagram showing the structure as viewed in the YZ plane. [Figure 28] FIG. 13 is a diagram showing a movement path in the YZ plane. [Figure 29] FIG. 2 is a diagram showing the relationship between points and voxels. [Diagram 30] FIG. 13 is a diagram showing a path of tilt movement. [Diagram 31] FIG. 13 is a diagram showing a path of rotational movement. [Diagram 32] FIG. 2 is a diagram showing a three-dimensional model and a point cloud. [Diagram 33] FIG. [Diagram 34] FIG. 2 shows a sample holder. [Diagram 35] FIG. 1 is a diagram showing a three-dimensional model. [Diagram 36] FIG. 1 is a diagram for explaining a collision determination method according to a conventional technique. [Figure 37] FIG. 1 is a diagram for explaining a collision determination method according to a conventional technique. [Figure 38] FIG. 1 is a diagram for explaining a collision determination method according to a conventional technique. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0030] A collision determination device 10 according to an embodiment will be described with reference to Fig. 1. Fig. 1 is a block diagram showing the functional configuration of the collision determination device 10.

[0031] The collision determination device 10 determines whether or not there is a collision between structures present in a sample chamber of a charged particle beam device 12. The charged particle beam device 12 is an electron microscope such as an SEM or a TEM, an FIB device, an AES device, or the like. The collision determination device 10 may be mounted on the charged particle beam device 12, or may be configured as a device separate from the charged particle beam device 12.

[0032] The collision determination device 10 acquires position information indicating the position of each structure from the charged particle beam device 12, and determines whether or not the structures collide with each other based on the position information of each structure.

[0033] In this embodiment, structures present in the sample chamber of the charged particle beam device 12 are classified as dynamic objects or static objects, and the dynamic objects are distinguished from the static objects.

[0034] A dynamic object is a structure whose position changes within the sample chamber, for example, a sample holder, a sample stage on which the sample holder is mounted, and a sample mounted on the sample holder.

[0035] Static objects are structures that do not change position within the sample chamber, such as the walls of the sample chamber, the objective lens, and a detector that moves back and forth along a predetermined path.

[0036] Each structure in the sample chamber is classified in advance as either a dynamic object or a static object. Note that the structures exemplified here are merely examples, and objects other than the structures exemplified here may be classified as dynamic objects or static objects.

[0037] Position information indicating the position of the dynamic object and position information indicating the position of the static object are output from the charged particle beam device 12 to the collision determination device 10, and the collision determination device 10 determines whether or not a collision occurs between the dynamic object and the static object based on the position information indicating the position of the dynamic object and the position information indicating the position of the static object. For example, when a user in the charged particle beam device 12 specifies a movement amount and a movement direction of the dynamic object and gives an instruction to move the dynamic object, information indicating the movement amount and the movement direction is output from the charged particle beam device 12 to the collision determination device 10. The collision determination device 10 determines whether or not a collision occurs between the dynamic object and the static object based on the information.

[0038] The collision determination device 10 includes, for example, a generation unit 14, a database 16, a determination unit 18, an output unit 20, a calculation unit 22, and a display unit 24.

[0039] The generation unit 14 generates point cloud information expressing the three-dimensional shape of the structure as a point cloud based on a three-dimensional model expressing the three-dimensional shape of the structure. The three-dimensional model is, for example, generated in advance and input to the collision determination device 10. The size (e.g., height and width) of the structure may be input by the user. In this case, the generation unit 14 generates a three-dimensional model and point cloud information according to the size input by the user. A user interface for inputting the size may be provided to the user.

[0040] The point cloud information is information that expresses the three-dimensional shape of a structure as a set of points distributed in three-dimensional space (i.e., a point cloud), and includes position information indicating the position of the structure in the sample chamber. The position of the structure is a position in three-dimensional space. More specifically, the point cloud information includes position information indicating the position of each point that expresses the structure. For example, the generation unit 14 obtains position information of the structure from the charged particle beam device 12, and includes the position information in the point cloud information of the structure. The generation unit 14 generates point cloud information for each of the dynamic objects and the static objects.

[0041] Further, the generating unit 14 generates voxel group information that represents the three-dimensional shape of the static object by a voxel group based on the point cloud information of the static object. Specifically, the generating unit 14 defines a unit of the three-dimensional space divided into a grid as a voxel, and links information indicating the presence of a static object (for example, a value of "1") to a voxel that includes a point that represents a static object (i.e., a voxel where a static object exists). The generating unit 14 links information indicating the absence of a static object (for example, a value of "0") to a voxel that does not include a point that represents a static object (i.e., a voxel where a static object does not exist). In this way, the generating unit 14 expresses the three-dimensional shape of the static object by information (information indicating the presence or absence of a static object) that is linked to each voxel distributed in the three-dimensional space. The generating unit 14 creates a database of information indicating the presence or absence of a static object, and generates voxel group information that represents the three-dimensional shape of the static object by a voxel group. The voxel group information includes position information of the static object in the sample chamber. More specifically, the voxel group information includes position information indicating the position of each voxel that represents a structure.

[0042] The database 16 stores the voxel group information of static objects in a database.

[0043] The determination unit 18 determines whether or not the dynamic object and the static object overlap by comparing the point cloud information representing the dynamic object with the voxel group information representing the static object. If the dynamic object and the static object overlap, the determination unit 18 determines that the dynamic object and the static object collide. If the dynamic object and the static object do not overlap, the determination unit 18 determines that the dynamic object and the static object do not collide.

[0044] The determination unit 18 may also determine whether or not the dynamic object overlaps with the static object by comparing, for each coordinate axis, voxel group information representing a static object present on the movement path of the dynamic object with point group information representing the dynamic object. For example, a coordinate system including a three-dimensional orthogonal coordinate system consisting of three mutually orthogonal coordinate axes is defined, and the determination unit 18 determines, for each coordinate axis, whether or not the dynamic object overlaps with the static object. The coordinate axes may include the tilt axis and the rotation axis of the dynamic object.

[0045] In addition, point cloud information representing a dynamic object for each range of a predetermined rotation angle is generated in advance. In this case, the determination unit 18 determines whether or not the dynamic object overlaps with the static object by comparing point cloud information according to the range of the rotation angle of the dynamic object with voxel group information representing the static object. This can reduce the time required to determine a collision, compared to a case in which point cloud information representing a rotated dynamic object is generated and a collision is determined every time the dynamic object rotates.

[0046] Also, voxel group information may be generated in advance for each state of a static object. In this case, the determination unit 18 determines whether or not a dynamic object overlaps with a static object by comparing voxel group information corresponding to the state of the static object with point group information representing the dynamic object. For example, when a change occurs in the state of a static object, voxel group information for each state is generated in advance in response to the change. In this way, the time required to determine a collision can be shortened compared to the case where voxel group information representing a static object corresponding to a state is generated each time a change occurs in the state to determine a collision.

[0047] The output unit 20 outputs information indicating the determination result by the determination unit 18. The output unit 20 may display the information indicating the determination result on the display unit 24, may express the determination result by sound and generate the sound from a speaker, may transmit the information to an external device (e.g., a user's terminal device, etc.), or may print the information on a medium such as paper. In addition, when the determination unit 18 is making a determination, the output unit 20 may display information indicating that the determination unit 18 is making a determination on the display unit 24.

[0048] The calculation unit 22 calculates the height of the sample placed in the sample chamber. Specifically, the calculation unit 22 calculates the height of the sample by a tilt operation of the dynamic object and a movement operation in the height direction. Note that the calculation unit 22 does not need to be included in the collision determination device 10.

[0049] The display unit 24 is a display such as a liquid crystal display or an EL display. Various information such as information indicating the determination result by the determination unit 18 is displayed on the display unit 24. The display unit 24 may function as an example of a user interface.

[0050] FIG. 2 shows a hardware configuration of the collision determination device 10. As shown in FIG.

[0051] The collision determination device 10 includes a communication device 26 , a UI 28 , a storage device 30 , and a processor 32 .

[0052] The communication device 26 includes one or more communication interfaces having a communication chip, a communication circuit, etc., and has a function of transmitting information to other devices and a function of receiving information from other devices. The communication device 26 may have a wireless communication function such as short-range wireless communication or Wi-Fi (registered trademark), or may have a wired communication function.

[0053] The UI 28 is a user interface and includes a display and an input device. The display is a liquid crystal display, an EL display, or the like. The input device is a keyboard, a mouse, an input key, an operation panel, or the like. The UI 28 may be a UI such as a touch panel that combines a display and an input device.

[0054] The storage device 30 is a device constituting one or more storage areas for storing data. The storage device 30 is, for example, a hard disk drive (HDD), a solid state drive (SSD), various types of memory (for example, RAM, DRAM, ROM, etc.), other storage devices (for example, optical disks, etc.), or a combination thereof. For example, the above-mentioned point cloud information and voxel group information are stored in the storage device 30. Data representing a three-dimensional model of each structure may be stored in the storage device 30. In addition, the database 16 is constructed by the storage device 30. Of course, the above-mentioned point cloud information and voxel group information may be stored in an external device other than the collision determination device 10 (for example, an external storage device, a server, or other device), or the database 16 may be constructed in the external device.

[0055] The processor 32 controls the operation of each part of the collision determination device 10. The processor 32 includes, for example, a central processing unit (CPU), a graphics processing unit (GPU), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), a digital signal processor (DSP), other programmable logic devices, or electronic circuits.

[0056] 1 is realized by a processor 32. The UI 28 and the storage device 30 may be used to execute each function.

[0057] For example, each function of the collision determination device 10 is realized by cooperation between hardware resources and software resources. For example, a CPU constituting the processor 32 reads and executes a program stored in the storage device 30 to realize each function. The program is stored in the storage device 30 via a recording medium such as a CD or DVD, or via a communication path such as a network. As another example, each function of the collision determination device 10 may be realized by hardware resources such as electronic circuits.

[0058] The following describes the embodiments in detail.

[0059] An example of a structure will be described with reference to Fig. 3. Fig. 3 shows a three-dimensional model of the objective lens 34 and a three-dimensional model of the sample holder 36. The objective lens 34 and the sample holder 36 are examples of structures present in the sample chamber of the charged particle beam device 12. The three-dimensional model of the objective lens 34 is a model that three-dimensionally represents the objective lens 34. The three-dimensional model of the sample holder 36 is a model that three-dimensionally represents the sample holder 36. These three-dimensional models are generated in advance and input to the collision determination device 10. Data of these three-dimensional models may be stored in the storage device 30.

[0060] The objective lens 34 is an example of a static object, which is a structure that is assumed to not change its position in the sample chamber, and the sample holder 36 is an example of a dynamic object, which is a structure that is assumed to change its position in the sample chamber.

[0061] The generating unit 14 generates point cloud information expressing the three-dimensional shape of the objective lens 34 as a point cloud based on the three-dimensional model of the objective lens 34. Similarly, the generating unit 14 generates point cloud information expressing the three-dimensional shape of the sample holder 36 as a point cloud based on the three-dimensional model of the sample holder 36.

[0062] 4 shows an example of a point cloud representing a structure. Point cloud 38 is a point cloud that represents the three-dimensional shape of objective lens 34 by a set of points distributed in three-dimensional space. Point cloud 40 is a point cloud that represents the three-dimensional shape of sample holder 36 by a set of points distributed in three-dimensional space. Point cloud information indicating point cloud 38 and point cloud information indicating point cloud 40 are stored in storage device 30.

[0063] The point cloud information indicating the point cloud 38 includes position information indicating the position of each point included in the point cloud 38 (the position of each point in three-dimensional space). In other words, the point cloud information indicating the point cloud 38 includes position information indicating the position of each point expressing the objective lens 34 in the sample chamber. For example, position information indicating the position of the objective lens 34 in the sample chamber is output from the charged particle beam device 12 to the collision determination device 10, and the generation unit 14 includes the position information in the point cloud information of the point cloud 38.

[0064] The point cloud information indicating the point cloud 40 includes position information indicating the position of each point included in the point cloud 40 (the position of each point in three-dimensional space). That is, the point cloud information indicating the point cloud 40 includes position information indicating the position of each point expressing the sample holder 36 in the sample chamber. For example, position information indicating the position of the sample holder 36 in the sample chamber is output from the charged particle beam device 12 to the collision determination device 10, and the generation unit 14 includes the position information in the point cloud information of the point cloud 40.

[0065] Furthermore, the generating unit 14 generates voxel group information that represents the three-dimensional shape of the objective lens 34 by a voxel group based on the point group information that represents the point group 38 of the objective lens 34, which is a static object. A voxel group 42 that represents the objective lens 34 is shown in FIG.

[0066] Here, the voxel group information of the objective lens 34 will be described in detail with reference to Fig. 6. The voxel group information stored in a database is shown in Fig. 6. An X-axis, a Y-axis, and a Z-axis that are orthogonal to each other are defined, and a plurality of lattice-like voxels are distributed in a three-dimensional space.

[0067] The generation unit 14 defines a unit of three-dimensional space divided into a grid as a voxel, and associates information indicating the presence or absence of the objective lens 34 with each voxel based on the point cloud information indicated by the point cloud 38. The value "1" in Fig. 6 is information indicating that the objective lens 34 is present, and the value "0" is information indicating that the objective lens 34 is not present.

[0068] A voxel associated with the value "1" is a voxel where the objective lens 34 exists (i.e., a voxel where a point included in the point cloud 38 exists). A voxel associated with the value "0" is a voxel where the objective lens 34 does not exist (i.e., a voxel where a point included in the point cloud 38 does not exist).

[0069] As shown in Fig. 6, the generation unit 14 generates voxel group information that expresses the three-dimensional shape of the objective lens 34 with information (value "1" or "0") associated with each voxel distributed in three-dimensional space, by linking a value "1" or a value "0" to each voxel based on the point cloud information indicating the point cloud 38. In this way, information indicating the presence or absence of the objective lens 34 is organized into a database. The databased voxel group information is stored in the database 16. The voxel group information that expresses the three-dimensional shape of the objective lens 34 includes position information that indicates the position of each voxel that expresses the objective lens 34.

[0070] 5, in order to visually explain the three-dimensional shape of the objective lens 34 with voxels, for convenience, only voxels associated with the value "1" are shown, and voxels associated with the value "0" are not shown. In other words, the voxel group 42 shown in FIG. 5 is expressed only by voxels associated with the value "1", but in the database representing the objective lens 34, voxels in which the objective lens 34 exists are associated with the value "1", and voxels in which the objective lens 34 does not exist are associated with the value "0".

[0071] An outline of the determination process by the determination unit 18 will be described with reference to Fig. 7 and Fig. 8. Fig. 7 shows an actual objective lens 44 and an actual sample holder 46. Fig. 8 shows a voxel group 42 representing the objective lens 44 and a point group 40 representing the sample holder 46. For ease of explanation, Fig. 8 shows only voxels associated with a value "1".

[0072] The determination unit 18 determines whether or not the objective lens 44 and the sample holder 46 overlap by comparing databased voxel group information indicating the voxel group 42 of the objective lens 44 with point cloud information indicating the point cloud 40 of the sample holder 46. For example, position information indicating the position of the sample holder 46 is output from the charged particle beam device 12 to the collision determination device 10, and point cloud information including the position information is generated by the generation unit 14. The determination unit 18 determines whether or not the objective lens 44 and the sample holder 46 overlap by comparing the point cloud information generated in this manner with voxel group information indicating the objective lens 44.

[0073] If even one point included in the point group 40 exists among the voxels included in the voxel group 42, the determination unit 18 determines that the objective lens 44 and the sample holder 46 overlap. In other words, the determination unit 18 determines that the objective lens 44 and the sample holder 46 collide with each other.

[0074] When there is no point included in the point group 40 among the voxels included in the voxel group 42, the determination unit 18 determines that the objective lens 44 and the sample holder 46 do not overlap. In other words, the determination unit 18 determines that the objective lens 44 and the sample holder 46 do not collide with each other.

[0075] 8, since a plurality of points included in the point group 40 exist among the voxels included in the voxel group 42, the determination unit 18 determines that the objective lens 44 and the sample holder 46 overlap. In this case, the output unit 20 may cause the display unit 24 to display information (e.g., warning information) indicating that the objective lens 44 and the sample holder 46 are colliding. When it is determined that the objective lens 44 and the sample holder 46 are not colliding, the output unit 20 may cause the display unit 24 to display information indicating that the objective lens 44 and the sample holder 46 are not colliding.

[0076] According to this embodiment, since the point cloud information and the voxel group information are each generated based on a three-dimensional model, it is possible to determine whether or not a collision occurs between a dynamic object and a static object, taking into consideration not only a two-dimensional plane but also the depth. As a result, it is possible to more accurately determine whether or not a collision occurs, compared to determining whether or not structures collide with each other on a two-dimensional plane.

[0077] In addition, by using a three-dimensional model, it is possible to accurately determine whether or not there will be a collision between structures with complex shapes, whether or not there will be a collision between structures moving along complex paths, and whether or not there will be a collision taking into account actual movement.

[0078] Furthermore, by determining whether or not there is a collision between a dynamic object and a static object before actually moving the dynamic object, the collision can be avoided in advance.

[0079] Each embodiment will be described below.

[0080] Example 1 A first embodiment will be described with reference to FIGS.

[0081] 9 shows a three-dimensional model of a static object 48 and a three-dimensional model of a dynamic object 50. Here, as an example, the static object 48 is a structure such as an objective lens, and the dynamic object 50 is a structure such as a sample holder and a stage.

[0082] 10 shows a schematic configuration of a dynamic object 50. The dynamic object 50 includes a sample holder 52 and a stage 54. An X-axis, a Y-axis, and a Z-axis that are orthogonal to each other are defined, and the stage 54 is a structure that can move along each of the X-axis, the Y-axis, and the Z-axis. The stage 54 can rotate about the R-axis as a rotation axis, and can tilt about the T-axis as a tilt axis. Here, as an example, the X-axis and the T-axis are coincident.

[0083] For example, the stage 54 includes an X-axis stage movable along the X-axis, a Y-axis stage movable along the Y-axis, a Z-axis stage movable along the Z-axis, a T-axis stage tiltable about the T-axis, and an R-axis stage rotatable about the R-axis. "X", "Y", "Z", "T" and "R" shown in Fig. 10 correspond to the respective axis stages.

[0084] When the X-axis stage moves along the X-axis, the sample holder 52 on the stage 54 also moves along the X-axis. The same applies to the Y-axis and Z-axis. When the T-axis stage tilts about the T-axis, the sample holder 52 also tilts about the T-axis. When the R-axis stage rotates about the R-axis, the sample holder 52 also rotates about the R-axis. In the example shown in FIG. 10, since the R-axis stage is provided on the Y-axis stage, when the Y-axis stage moves, the sample holder 52 on the stage 54 and the R-axis stage move along the Y-axis.

[0085] The generating unit 14 converts the three-dimensional model of the static object 48 into a point cloud to generate point cloud information of the static object 48, and converts the three-dimensional model of the dynamic object 50 into a point cloud to generate point cloud information of the dynamic object 50. The generating unit 14 also converts the point cloud of the static object 48 into a voxel group to generate voxel group information of the static object 48. As the three-dimensional model of the dynamic object 50, a three-dimensional model in which the coordinates of all axes of the stage 54 (i.e., the coordinates on the X-axis, Y-axis, Z-axis, T-axis, and R-axis) are "0" is used. In addition, the X-axis, Y-axis, and Z-axis of the three-dimensional space that define the coordinates of each point cloud coincide with the X-axis, Y-axis, and Z-axis when the stage 54 is not tilted (i.e., when T=0).

[0086] The determination unit 18 determines whether or not the static object 48 and the dynamic object 50 overlap by comparing the voxel group information of the static object 48 with the point cloud information of the dynamic object 50. If it is determined that the static object 48 and the dynamic object 50 overlap, it is determined that the position of the stage 54 at that time is a position where a collision between the static object 48 and the dynamic object 50 occurs. If it is determined that the static object 48 and the dynamic object 50 do not overlap, it is determined that the position of the stage 54 at that time is a position where a collision between the static object 48 and the dynamic object 50 does not occur. In this manner, it is determined whether or not a collision between the static object 48 and the dynamic object 50 occurs.

[0087] When determining whether or not a collision has occurred using a three-dimensional point cloud and a voxel group, the processing speed for the determination is slower and the time required for the determination is longer than when determining whether or not a collision has occurred using a two-dimensional model.

[0088] Therefore, in the first embodiment, the determination unit 18 first uses a two-dimensional model to determine whether or not there is a collision between the static object 48 and the dynamic object 50. When it is determined that there is a collision between the static object 48 and the dynamic object 50 in the determination using the two-dimensional model, the determination unit 18 uses a three-dimensional model to determine whether or not there is a collision between the static object 48 and the dynamic object 50 in accordance with the above-described embodiment.

[0089] With reference to Fig. 11, collision determination using a two-dimensional model will be described. Fig. 11 shows structures as viewed from within the YZ plane. A static object 56 is an object as viewed from within the YZ plane when the static object 48 is viewed. A dynamic object 58 is an object as viewed from within the YZ plane when the dynamic object 50 is viewed. For example, the generation unit 14 expresses the two-dimensional static object 56 and the two-dimensional dynamic object 58 as polygons, and the determination unit 18 determines whether or not the static object 56 and the dynamic object 58 overlap within the YZ plane.

[0090] When it is determined that a two-dimensional static object 56 and a two-dimensional dynamic object 58 overlap in the YZ plane, the determination unit 18 determines whether or not there is overlap between the static object 48 and the dynamic object 50 by comparing the voxel group information of the three-dimensional static object 48 with the point cloud information of the three-dimensional dynamic object 50.

[0091] If it is determined that the two-dimensional static object 56 and the two-dimensional dynamic object 58 do not overlap in the YZ plane, the determination unit 18 does not determine whether or not there is overlap between the point group generated from the three-dimensional model and the voxel group.

[0092] In this way, by first determining whether or not a collision has occurred using a two-dimensional model, the amount of calculation required for the determination can be reduced, and the time required for the determination process can be shortened. Furthermore, if it is determined that a collision has occurred using the two-dimensional model, the presence or absence of a collision can be determined next using a three-dimensional model, thereby making it possible to accurately determine whether or not a collision has occurred.

[0093] Depending on the type of the charged particle beam device 12, the arrangement of each axis may differ. For example, in an SEM, a Z-axis stage is placed below a T-axis stage. Although the direction of movement of the Z-axis stage does not change due to the tilt of the dynamic object about the T-axis, the position of the T-axis stage changes depending on the arrangement position of the Z-axis stage. The generation unit 14 generates point cloud information reflecting this change, and the determination unit 18 uses the point cloud information to determine the presence or absence of a collision.

[0094] Example 2 In the second embodiment, a user interface for inputting the size of a sample to be mounted on a dynamic object is provided to a user, and the user inputs the size of the sample. The generation unit 14 generates point cloud information indicating a point cloud of the dynamic object reflecting the input size. For example, the dynamic object is a sample holder, and the generation unit 14 generates point cloud information indicating a point cloud of the dynamic object including the sample holder and the sample mounted on the sample holder. The determination unit 18 determines whether or not the dynamic object overlaps with the static object by comparing the point cloud information indicating the point cloud of the dynamic object reflecting the size of the sample with voxel group information indicating the voxel group of the static object.

[0095] The second embodiment will be described in detail below with reference to FIGS.

[0096] A sample holder 60 is shown in Fig. 12. The sample holder 60 is capable of mounting a plurality of samples. For example, a plurality of mounting parts 62 (five mounting parts 62 in the example shown in Fig. 12) are provided on the sample holder 60, and a sample is mounted on each mounting part 62. For example, a sample is mounted at the center of the surface of each mounting part 62. The samples mounted on each mounting part 62 may be samples of different sizes and shapes, or may be samples of the same size and shape.

[0097] A screen 64 is shown in Fig. 13. The screen 64 is an example of a user interface for inputting the size of a sample, and is displayed on a display constituting the display unit 24. A list 66 of sample holder candidates is displayed on the screen 64. In the example shown in Fig. 13, four sample holders are included in the list 66 as candidates, and an image representing each candidate is displayed on the screen 64. For example, when a user operates the UI 28 to instruct the display of the screen 64, the processor 32 causes the screen 64 to be displayed on the display.

[0098] Once a sample holder is selected by the user from the list 66, the processor 32 displays an image 68 on the screen 64 representing the selected sample holder.

[0099] Also displayed on the screen 64 are an image 70 that diagrammatically represents the sample, and an input field 72 for inputting the size of the sample. Here, the size of the sample is the height of the sample, and the input field 72 is a field for inputting the height of the sample. Specifically, the input field 72 is a bar for adjusting the height of the sample, and the user adjusts the bar to input the height of the sample.

[0100] The generating unit 14 generates point cloud information representing a point cloud of a dynamic object reflecting the height input by the user. Here, as an example, the dynamic object includes a sample holder 60, a mounting unit 62, and a sample, and the generating unit 14 generates point cloud information representing a point cloud of the dynamic object reflecting the height of the sample. For example, the generating unit 14 increases the height of the mounting unit 62 by the height of the sample, and generates point cloud information representing a point cloud of the dynamic object reflecting the increased height.

[0101] Furthermore, the processor 32 may display on the display an image 74 representing a dynamic object reflecting the height of the sample input by the user, as shown in Fig. 14. For example, the sample holder 60 and the mounting unit 62 are shown in the image 74, and an image 76 representing an object having the input height is included in the image 74. The object corresponds to a portion of the mounting unit 62 where the height has increased by the input height. The shape of the object may be a cylinder, a rectangular parallelepiped, or the like.

[0102] As shown in Fig. 15, multiple specimens 78 may be mounted on the specimen holder 60. The sizes and shapes of the specimens 78 may be different from each other or may be the same. In the example shown in Fig. 15, the sizes and shapes of the specimens 78 are different from each other.

[0103] For example, the user inputs on the screen 64 the height of the sample 78 that is the tallest among the multiple samples 78. The generation unit 14 defines the input height as the height of each sample 78, and generates point cloud information that indicates a point cloud of a dynamic object in which the height is reflected.

[0104] As another example, the user may input the height of each sample 78 on the screen 64. In this case, the generation unit 14 generates point cloud information indicating a point cloud of a dynamic object in which the height of each sample 78 is reflected.

[0105] The determination unit 18 determines whether or not the dynamic object overlaps with the static object by comparing point cloud information indicating a point cloud of the dynamic object reflecting the height of the sample input by the user with voxel group information indicating a voxel group of the static object. Also, in the second embodiment, the determination unit 18 may perform a determination using a two-dimensional model as in the first embodiment, and then perform a determination using a three-dimensional model.

[0106] According to the second embodiment, the presence or absence of a collision is determined using point cloud information reflecting the height input by the user, so that the presence or absence of a collision can be determined for any height.

[0107] In the above example, a case where a sample is mounted on the sample holder has been described, but the process according to the second embodiment may be executed when a part other than the sample is attached to the sample holder. As a result, point cloud information indicating a point cloud of a dynamic object reflecting the size of the attached part is generated, and the point cloud information is collated with the voxel group information of the static object.

[0108] Example 3 In the third embodiment, the calculation unit 22 is used, and the height of the sample is calculated by the calculation unit 22.

[0109] Example 3 will be described below with reference to Figs. 16 to 18. Figs. 16 to 18 show a sample holder 80, a sample 82 placed on the sample holder 80, and an object 84 present on a surface 82a of the sample 82. Figs. 16 to 18 show each object as viewed in the YZ plane. In Example 3, the height of the sample 82 is calculated by tracking the object on the electron image while tilting the sample holder 80 (i.e., while tilting the stage). A specific procedure will be described below.

[0110] (1) First, in the charged particle beam device 12, the stage on which the sample holder 80 is mounted is moved horizontally without being tilted.

[0111] (2) The user observes the electron image generated by imaging with the charged particle beam device 12, and determines the object 84 on the electron image. Then, as shown in Fig. 16, the user moves the stage so that the object 84 is centered on the electron image. This moves the sample holder 80 and the sample 82. At this time, the stage is not tilted.

[0112] 17, the charged particle beam device 12 tilts the sample holder 80 by tilting the stage around the T-axis, which is the tilt axis. The tilt angle is, for example, a predetermined angle. Since the Y-axis and the Z-axis are determined based on the stage, when the stage is tilted, the Y-axis and the Z-axis also tilt by the same tilt angle.

[0113] (4) Next, the user observes the electron image, specifies the position of the object 84 on the electron image, and inputs information indicating the position to the charged particle beam device 12. For example, the information indicating the position is input to the charged particle beam device 12 via a UI.

[0114] 18, the charged particle beam device 12 moves the stage along the Z axis based on the position input by the user so that the target object 84 is centered on the electron image. By moving the stage, the sample holder 80 and the sample 82 are moved along the Z axis.

[0115] (6) The above steps (3) to (5) are repeated until the object 84 does not move on the electron image when the specimen holder 80 is tilted. For example, the state shown in Fig. 18 is the state in which the object 84 does not move on the electron image when the specimen holder 80 is tilted.

[0116] (7) Next, the calculation unit 22 identifies the Z coordinate of the surface 82a of the sample 82, and calculates the height of the sample 82 based on the Z coordinate.

[0117] In the following, a method for calculating the height of the sample 82 when the Z axis exists on the T axis will be described. The state in which the object 84 does not move on the electron image even when the sample holder 80 is tilted is a state in which the position of the object 84 and the position of the T axis coincide with each other and the surface 82a of the sample 82 is in contact with the T axis, as shown in FIG. 18. In this case, even if the sample holder 80 is tilted around the T axis, the object 84 does not move. The calculation unit 22 calculates the Z coordinate of the surface 82a when the surface 82a is in contact with the T axis when the height of the sample 82 is "0", and calculates the difference between the calculated Z coordinate and the Z coordinate calculated in (7) as the height of the sample 82. This makes it possible to calculate the height more accurately.

[0118] When the height of the sample 82 is calculated as described above, the generation unit 14 generates point cloud information indicating a point cloud of a dynamic object in which the height is reflected. The determination unit 18 compares the point cloud information indicating the point cloud of the dynamic object in which the height is reflected with voxel group information indicating a voxel group of a static object, thereby determining whether or not the dynamic object overlaps with the static object.

[0119] Alternatively, the object 84 may be specified by the user, and the calculation unit 22 may track the position of the object 84 after the tilt operation by applying pattern matching to the object 84 depicted in the electronic image.

[0120] Example 4 In the fourth embodiment, the generation unit 14 generates point cloud information indicating a point cloud of a dynamic object according to the conditions of observation by the charged particle beam device 12. The determination unit 18 compares the point cloud information according to the observation conditions with voxel group information indicating a voxel group of a static object, thereby determining whether or not the dynamic object overlaps with the static object.

[0121] For example, the observation condition is a condition related to the voltage applied to the sample to be observed. When the sample and the objective lens approach each other while a voltage is being applied to the sample, a discharge may occur. In order to avoid this, the generation unit 14 sets a discharge danger area according to the magnitude of the voltage in the dynamic object, the static object, or both the dynamic object and the static object. That is, the generation unit 14 may set the discharge danger area so as to cover a three-dimensional model of the dynamic object including the sample holder and the sample, may set the discharge danger area so as to cover a three-dimensional model of the static object, or may set the discharge danger area so as to cover each of the three-dimensional models of the dynamic object and the static object. The object in which the discharge danger area is set may be set by the user or may be set in advance.

[0122] Here, as an example, the generating unit 14 sets a discharge danger area according to the magnitude of the voltage so as to cover a three-dimensional model of a dynamic object including a sample holder and a sample. Of course, the generating unit 14 may set a discharge danger area so as to cover a three-dimensional model of a static object including an objective lens, or may set a discharge danger area in both the dynamic object and the static object. An area in which a discharge may occur is set as the discharge danger area. The generating unit 14 generates a three-dimensional model of a dynamic object including a sample holder, a sample, and a discharge danger area. For example, the generating unit 14 generates a three-dimensional model including a sample holder, a sample, and a discharge danger area by setting a discharge danger area for an original three-dimensional model representing the sample holder and the sample.

[0123] The generation unit 14 generates a three-dimensional model of the dynamic object by increasing the size of the discharge danger area as the voltage increases. For example, the generation unit 14 generates a three-dimensional model of the dynamic object by increasing the size of the discharge danger area when the voltage is 5 kV compared to the size of the discharge danger area when the voltage is 2 kV. The generation unit 14 may change the shape of the discharge danger area depending on the magnitude of the voltage.

[0124] The generating unit 14 generates point cloud information representing a dynamic object including the specimen holder, the specimen, and the discharge danger area as a point cloud based on a three-dimensional model including the specimen holder, the specimen, and the discharge danger area. In other words, the discharge danger area is treated as a dynamic object.

[0125] The judgment unit 18 judges whether or not there is overlap between the dynamic object and the static object by comparing point cloud information, which represents the dynamic object including the sample holder, the sample, and the discharge danger area as a point cloud, with voxel group information, which indicates the voxel group of the static object.

[0126] As described above, by setting a discharge danger zone, it is possible to prevent the occurrence of discharge even when observation is performed under observation conditions in which discharge may occur.

[0127] Example 5 In the fifth embodiment, the state inside the sample chamber of the charged particle beam device 12 is displayed in a three-dimensional model. Hereinafter, the fifth embodiment will be described with reference to Fig. 19 and Fig. 20. Fig. 19 and Fig. 20 show a screen 86.

[0128] Screen 86 is displayed on a display included in display unit 24. Processor 32 generates a three-dimensional model representing structures present in the sample chamber, and displays the three-dimensional model on screen 86. For example, an objective lens, a sample holder, and a sample are structures present in the sample chamber, and three-dimensional models representing these are generated and displayed on screen 86.

[0129] Fig. 19 shows a three-dimensional model in a zoomed-out state, and Fig. 20 shows a three-dimensional model in a zoomed-in state. For example, a bar for instructing a zoomed-in or zoomed-out display is displayed on the screen 86, and the user can instruct the zoom-in or zoom-out by operating the bar.

[0130] By displaying a three-dimensional model representing the structures present in the sample chamber, the user can confirm the positional relationship of the structures, etc. For example, the user can confirm the positional relationship between the objective lens and the surface of the sample.

[0131] Also, a three-dimensional model representing a structure such as a detector may be generated and displayed, allowing a user to confirm the positional relationship between the detector and the surface of the sample.

[0132] Example 6 In the sixth embodiment, a voxel group is generated in advance for each state of a static object, and voxel group information is generated in advance for each state of the static object and stored in database 16. Determination unit 18 reads out the voxel group information corresponding to the state of the static object from database 16, and compares the voxel group information corresponding to the state of the static object with point cloud information that represents the dynamic object as a point cloud, thereby determining whether or not the dynamic object overlaps with the static object.

[0133] Specifically, the static object includes a structure capable of reciprocating along a predetermined trajectory. The structure is an object that reciprocates, but since it is an object that moves along a predetermined trajectory, it can be considered as a static object existing on the trajectory.

[0134] For example, a structure is an object that moves back and forth between a first position and a second position. The state of a static object is a state in which the structure is at the first position (hereinafter referred to as the "first state"), a state in which the structure is at the second position (hereinafter referred to as the "second state"), or a state in which the structure is moving back and forth between the first position and the second position (hereinafter referred to as the "third state"). For example, either the first position or the second position corresponds to a start position of the reciprocating movement, and the other corresponds to an end position of the reciprocating movement.

[0135] Based on a three-dimensional model (hereinafter referred to as the "first three-dimensional model") representing the static object when the structure is at the first position, the generation unit 14 generates in advance point cloud information representing the static object when the structure is at the first position as a point cloud, and generates in advance voxel group information representing the static object when the structure is at the first position as a voxel group based on the point cloud information. Hereinafter, the voxel group representing the static object when the structure is at the first position is referred to as the "first voxel group," and the voxel group information representing the first voxel group is referred to as the "first voxel group information."

[0136] Furthermore, based on a three-dimensional model representing the static object when the structure is at the second position (hereinafter referred to as the "second three-dimensional model"), the generation unit 14 generates in advance point cloud information representing the static object when the structure is at the second position as a point cloud, and based on the point cloud information, generates in advance voxel group information representing the static object when the structure is at the second position as a voxel group. Hereinafter, the voxel group representing the static object when the structure is at the second position is referred to as the "second voxel group," and the voxel group information representing the second voxel group is referred to as the "second voxel group information."

[0137] Furthermore, the generation unit 14 generates in advance point cloud information representing, as a point cloud, the static objects when the structure is moving back and forth between the first position and the second position, based on a three-dimensional model (hereinafter referred to as a "third three-dimensional model") representing the static objects when the structure is moving back and forth between the first position and the second position, and generates in advance voxel group information representing, as a voxel group, the static objects when the structure is moving back and forth between the first position and the second position, based on the point cloud information. Hereinafter, the voxel group representing the static objects when the structure is moving back and forth between the first position and the second position is referred to as a "third voxel group," and the voxel group information representing the third voxel group is referred to as "third voxel group information."

[0138] The third three-dimensional model is a three-dimensional model that represents the trajectory of the movement of the structure. Therefore, the third voxel group is a voxel group that represents the trajectory of the movement of the structure. In other words, the third three-dimensional model is a three-dimensional model that represents the trajectory of the movement of the structure when the structure moves between the first position and the second position, and the third voxel group is a voxel group that represents the trajectory of the movement of the structure when the structure moves between the first position and the second position.

[0139] The first voxel group information, the second voxel group information, and the third voxel group information are generated in advance and stored in the database 16.

[0140] Hereinafter, the sixth embodiment will be described with reference to Figures 21 to 23. Figures 21 to 23 show a three-dimensional model of a static object and a three-dimensional model of a dynamic object.

[0141] In Figures 21 to 23, a three-dimensional model of the objective lens and a three-dimensional model of the detector are shown as three-dimensional models of static objects, and a three-dimensional model of the sample holder is shown as a three-dimensional model of a dynamic object.

[0142] Three-dimensional model 88 is a three-dimensional model of the objective lens, three-dimensional models 90A, 90B, and 90C are three-dimensional models of the detectors, and three-dimensional model 92 is a three-dimensional model of the sample holder.

[0143] The detector is an example of a structure that can move back and forth along a predetermined trajectory. Specifically, the detector can move back and forth along the direction indicated by the arrow A in Fig. 21. In other words, a trajectory is set along the direction indicated by the arrow A, and the detector can move back and forth on the trajectory.

[0144] The three-dimensional model 90A shown in FIG. 21 is a three-dimensional model representing a detector present at a first position. The three-dimensional models 88 and 90A are three-dimensional models representing a static object (i.e., an objective lens and a detector) when the detector is at a first position. The state of the static object when the detector is at the first position is a state (in state) in which the detector is inserted between the objective lens and the sample holder. The three-dimensional models 88 and 90A shown in FIG. 21 correspond to an example of a first three-dimensional model. The generation unit 14 generates in advance point cloud information representing a static object when the detector is at the first position with a point cloud based on the three-dimensional models 88 and 90A shown in FIG. 21, and generates in advance first voxel group information representing a static object when the detector is at the first position with a voxel group (first voxel group) based on the point cloud information. The first voxel group information is stored in the database 16 in association with information indicating the in state (i.e., information indicating the first position).

[0145] The three-dimensional model 90B shown in FIG. 22 is a three-dimensional model representing the detector at the second position. The three-dimensional models 88 and 90B are three-dimensional models representing the static object (i.e., the objective lens and the detector) when the detector is at the second position. The state of the static object when the detector is at the second position is a state (out state) in which the detector is retreated from the region between the objective lens and the sample holder to an external region. The three-dimensional models 88 and 90B shown in FIG. 22 correspond to an example of a second three-dimensional model. The generation unit 14 generates in advance point cloud information representing the static object when the detector is at the second position with a point cloud based on the three-dimensional models 88 and 90B shown in FIG. 22, and generates in advance second voxel group information representing the static object when the detector is at the second position with a voxel group (second voxel group) based on the point cloud information. The second voxel group information is stored in the database 16 in association with information indicating the out state (i.e., information indicating the second position).

[0146] The three-dimensional model 90C shown in FIG. 23 is a three-dimensional model that represents the state when the detector moves back and forth between the first position and the second position. That is, the three-dimensional model 90C is a three-dimensional model that represents the trajectory when the detector moves back and forth (that is, a three-dimensional model that represents the path of the detector's back and forth movement). The three-dimensional model 90C also represents the trajectory of the back and forth movement of a protrusion (e.g., a screw, etc.) included in the detector. The state of the static object (i.e., the objective lens and the detector) at this time is an in-out state, that is, a state in which the detector state is switching from the in state to the out state, or a state in which the detector state is switching from the out state to the in state. The three-dimensional models 88 and 90C shown in FIG. 23 correspond to an example of a third three-dimensional model. 23, the generation unit 14 generates in advance point cloud information expressing, by a point cloud, static objects when the detector moves back and forth between the first position and the second position, and generates in advance third voxel group information expressing, by a voxel group (third voxel group), static objects when the detector starts to move back and forth between the first position and the second position, based on the point cloud information. The third voxel group information is stored in the database 16 in association with information indicating an in-out state.

[0147] As described above, voxel group information for each state of a static object is generated in advance and stored in database 16. That is, voxel group information when the static object is in the IN state, voxel group information when the static object is in the OUT state, and voxel group information when the static object is in an IN / OUT state are stored in database 16 in advance.

[0148] The determination unit 18 determines whether or not there is an overlap between the dynamic object and the static object by comparing voxel group information corresponding to the state of the static object with point group information representing the dynamic object.

[0149] For example, before analysis, when the stage (the stage on which the sample holder is mounted) is moved while the detector is in the OUT state, the determination unit 18 checks the voxel group information when the static object is in the OUT state with the point cloud information of the dynamic object to determine whether or not the dynamic object overlaps with the static object. At this time, the voxel group information when the static object is in the IN state and the voxel group information when the static object is in / out are not used.

[0150] When the state of the detector is switched from the out state to the in state depending on the timing of the analysis, etc., the determination unit 18 determines whether or not the dynamic object overlaps with the static object by comparing the voxel group information when the static object is in-out with the point cloud information of the dynamic object. At this time, the voxel group information when the static object is in the in state and the voxel group information when the static object is in the out state are not used. When it is determined that the dynamic object overlaps with the static object, the detector is not inserted between the objective lens and the sample holder. When it is not determined that the dynamic object overlaps with the static object, the detector is inserted between the objective lens and the sample holder.

[0151] When the stage is moved while the detector is in the IN state, the determination unit 18 determines whether or not the dynamic object overlaps with the static object by comparing the voxel group information when the static object is in the IN state with the point cloud information of the dynamic object. At this time, the voxel group information when the static object is in the OUT state and the voxel group information when the static object is in the IN / OUT state are not used.

[0152] When the detector state is switched from the in state to the out state at a stage such as after analysis, the determination unit 18 determines whether or not there is overlap between the dynamic object and the static object by comparing the voxel group information when the static object is in an in-out state with the point group information of the dynamic object.

[0153] In this manner, in the sixth embodiment, voxel group information for each state of a static object is generated in advance, and the voxel group information corresponding to the state of the static object is compared with the point cloud information of the dynamic object. This makes it possible to reduce the time required to determine overlap, compared with the case where the voxel group information corresponding to the state of the static object is generated each time comparison is performed.

[0154] For example, as a comparative example, when the detector moves and the state of the detector is switched from the in state to the out state or from the out state to the in state, it is possible to generate voxel group information for each position of the detector during the movement, and compare the voxel group information corresponding to the position of the detector with the point cloud information of the dynamic object. In this case, it is necessary to generate voxel group information for each position of the detector when the detector is moving. However, if the voxel group information for each position of the detector is generated each time, it takes time to generate it, and the time required for the overlap determination increases accordingly. In contrast, according to the sixth embodiment, since the voxel group information when the state of the static object is in the in-out state is generated in advance, the time required for the determination can be shortened compared to the case where the voxel group information is generated each time while the detector is moving. The voxel group when the state of the static object is in the in-out state includes a voxel group representing the trajectory of the movement of the detector, that is, the path when the detector moves. Therefore, by using the voxel group information when the state of the static object is in the in-out state, it is possible to determine whether or not a collision occurs when the detector is moving back and forth.

[0155] Another example of the sixth embodiment will be described with reference to Fig. 24. Fig. 24 shows a point cloud 94 representing the objective lens in the normal state and a point cloud 96 representing the approach prohibition area.

[0156] Depending on the observation mode, it may be desired to prevent a dynamic object from approaching the objective lens more closely than in a normal state. In this case, the generation unit 14 generates a composite point cloud 98 representing the objective lens and the approach prohibition region by combining a point cloud 94 representing the objective lens in a normal state with a point cloud 96 representing the approach prohibition region. The approach prohibition region may be set in advance or may be set by the user. The generation unit 14 generates voxel group information based on the composite point cloud 98.

[0157] When an observation mode requiring a no-approach area is executed, the judgment unit 18 judges whether or not there is overlap between the dynamic object and the static object by comparing the voxel group information generated based on the composite point group 98 with the point group information of the dynamic object.

[0158] Example 7 In the seventh embodiment, the determination unit 18 determines whether or not there is overlap between the dynamic object and the static object by comparing voxel group information representing a static object present on the movement path of the dynamic object with point group information representing the dynamic object.

[0159] For example, a coordinate system including a three-dimensional orthogonal coordinate system composed of three mutually orthogonal coordinate axes is defined. The determination unit 18 determines whether or not the dynamic object overlaps with the static object by comparing, for each coordinate axis, voxel group information representing a static object present on the movement path of the dynamic object with point group information representing the dynamic object.

[0160] The coordinate system further includes a tilt axis of the dynamic object. When the dynamic object tilts around the tilt axis, the determination unit 18 checks, for each coordinate axis, voxel group information representing static objects present on the moving path of the tilted dynamic object with point group information representing the dynamic object to determine whether the dynamic object overlaps with the static object.

[0161] Example 7 will be described below with reference to Fig. 25 to Fig. 30. Fig. 25 shows a static object 100 and a dynamic object 102. The static object 100 includes, for example, an objective lens. The dynamic object 102 includes, for example, a stage, a sample holder mounted on the stage, and a sample mounted on the sample holder.

[0162] As described with reference to FIG. 10, an X-axis, a Y-axis, and a Z-axis that are mutually orthogonal are defined, and the stage is a structure that can move along each of the X-axis, the Y-axis, and the Z-axis. The stage can also rotate around the R-axis as a rotation axis, and can tilt around the T-axis as a tilt axis. Here, as an example, the X-axis and the T-axis are coincident. For example, the stage includes an X-axis stage, a Y-axis stage, a Z-axis stage, a T-axis stage, and an R-axis stage. FIG. 25 shows an object as viewed in the XZ plane.

[0163] When determining whether a dynamic object overlaps with a static object, it is possible to view the voxel group on a plane. This plane is, for example, an XZ plane or a YZ plane. When the voxel group is viewed on a plane, the voxel group can be treated as a collection of squares. The determination unit 18 determines whether or not a voxel associated with a value "1" exists on the moving path of the dynamic object on each of the X-axis, Y-axis, Z-axis, and T-axis. If a voxel associated with a value "1" exists on the moving path of the dynamic object, the determination unit 18 determines that the dynamic object and the static object overlap at the position of the voxel associated with the value "1". If a voxel associated with a value "0" exists on the moving path of the dynamic object, the determination unit 18 determines that the dynamic object and the static object do not overlap at the position of the voxel associated with the value "0". The determination unit 18 determines whether or not there is an overlap for each coordinate axis.

[0164] With reference to Fig. 26, a case where the dynamic object is not rotating around the T axis (i.e., the tilt angle is = 0°) will be described. In Fig. 26, the XZ plane or the YZ plane is shown. Here, movement within the XZ plane will be described. In Fig. 26, the voxels are expressed as two-dimensional square grids.

[0165] The movement directions of the X-axis stage, the Y-axis stage, and the Z-axis stage correspond to the X-axis, the Y-axis, and the Z-axis that define a three-dimensional coordinate system. Therefore, the determination unit 18 can calculate the coordinates of the target position by calculating the coordinates indicating the current position of the point cloud representing the dynamic object 102 and adding the movement amount of the dynamic object 102 to the coordinates. For example, in the charged particle beam device 12, when the movement amount and the movement direction of the dynamic object 102 are specified and an instruction to move is given, information indicating the movement amount and the movement direction is output from the charged particle beam device 12 to the collision determination device 10. The determination unit 18 calculates the coordinates of the target position based on the information.

[0166] For example, when the movement path from the current position to the target position is viewed on the XZ plane, the movement path along the X axis and the movement path along the Z axis are each expressed as a line segment connecting two points. For example, the movement path along the X axis is shown as a movement path from the current position 104 to the target position 106, and when viewed on the XZ plane, the movement path is expressed as a line segment 108 connecting two points. If a voxel (a square represented as a two-dimensional square in FIG. 26) associated with the value "1" exists on the line segment 108, the determination unit 18 determines that the dynamic object 102 and the static object 100 overlap when the dynamic object 102 moves along the X axis in the XZ plane. If a voxel associated with the value "1" does not exist on the line segment 108, the determination unit 18 determines that the dynamic object 102 and the static object 100 do not overlap when the dynamic object 102 moves along the X axis in the XZ plane. The presence or absence of overlap is determined similarly for movement along the Z axis, and also for movement in the YZ plane.

[0167] A case where the dynamic object 102 is tilted around the T axis (i.e., tilt angle>0) will be described with reference to Fig. 27 and Fig. 28. Fig. 27 shows the dynamic object 102 in a tilted state. Fig. 28 shows the YZ plane. In Fig. 28, voxels are represented as two-dimensional square grids.

[0168] Since the movement direction of the X-axis stage and the T-axis are parallel, tilting around the T-axis does not affect the movement direction of the X-axis stage. Therefore, the presence or absence of overlap on the X-axis is determined by the same method as when the dynamic object 102 is not tilted around the T-axis. On the other hand, the movement directions on the Y-axis, Z-axis, and T-axis are tilted according to the tilt of the dynamic object 102 around the T-axis.

[0169] The movement directions of the Y-axis stage and the Z-axis stage change according to the tilt of the dynamic object 102 about the T-axis, but neither axis stage moves in the movement direction of the X-axis stage. Therefore, the presence or absence of overlap on the Y-axis and the Z-axis can be determined by examining voxels on a diagonal line on the YZ plane, just like when the dynamic object 102 is not tilted about the T-axis.

[0170] The determination unit 18 regards the voxels as pixels on the YZ plane, draws an approximate line, and checks whether the value associated with the voxels on the diagonal line is "1" or "0." In Fig. 28, the movement path of the dynamic object in the YZ plane is shown as a movement path from a current position 110 to a target position 112. When viewed in the YZ plane, the movement path is expressed as a line segment 114 connecting two points.

[0171] If a voxel (a square cell in FIG. 28) associated with the value "1" exists on the line segment 114, the determination unit 18 determines that the dynamic object 102 and the static object 100 overlap when the dynamic object 102 moves within the YZ plane. If a voxel associated with the value "1" does not exist on the line segment 114, the determination unit 18 determines that the dynamic object 102 and the static object 100 do not overlap when the dynamic object 102 moves within the YZ plane.

[0172] Strictly speaking, an error may occur when an approximation line is used. However, when converting a point group into a voxel group, as shown in FIG. 29, by treating each point 110 included in the point group as a sphere 113, the region 115 in which the object exists can be expanded. For example, the inside of the sphere 113 and the part overlapping with the surface of the sphere 113 are included in the region 115 in which the object exists. This expansion makes it possible to ignore errors when checking voxels on the line. This makes it possible to determine whether there is an overlap without checking all voxels on the line. As a result, the time required for the determination can be shortened.

[0173] The tilt movement around the T axis will be described below. Since the X-axis stage and the T axis are parallel, the tilt movement around the T axis does not affect the movement direction of the X-axis stage. Therefore, the determination unit 18 calculates the coordinates of the current position of the dynamic object 102, and then rotates the dynamic object 102 around the T axis in the YZ plane by the amount of movement, thereby calculating the coordinates of the target position.

[0174] The path of the tilt movement at this time is shown in Figure 30. The path of the tilt movement at this time is represented by an arc-shaped line segment 120 connecting a current position 116 and a target position 118. This line segment 120 is a line segment that indicates an arc with the T-axis as the center of rotation within the YZ plane.

[0175] For example, the determination unit 18 regards the voxels as pixels in the YZ plane, draws an approximate arc curve, and checks whether the value associated with the voxels on the line segment 120 is "1" or "0."

[0176] If a voxel (a square cell in FIG. 30) associated with the value "1" exists on the line segment 120, the determination unit 18 determines that the dynamic object 102 and the static object 100 overlap when the dynamic object 102 tilts around the T axis in the YZ plane. If a voxel associated with the value "1" does not exist on the line segment 120, the determination unit 18 determines that the dynamic object 102 and the static object 100 do not overlap when the dynamic object 102 tilts around the T axis in the YZ plane.

[0177] In addition, in a device where the Z axis is placed below the T axis (e.g., SEM), the direction of movement of the Z axis stage does not change due to the tilt of the dynamic object around the T axis, so there is no need to check voxels on a diagonal line, and it is sufficient to only check voxels in the vertical direction. Even if the T axis is not parallel to the X axis, as long as the T axis is parallel to the Y axis, it is sufficient to determine the presence or absence of a collision on the XZ plane.

[0178] Example 8 Rotational movement around the R axis will be described with reference to Fig. 31. A path 122 along which a dynamic object rotates is shown in Fig. 31. As an example, it is assumed here that the R axis is placed at the top of the coordinate axes in a stage included in the dynamic object.

[0179] When the R axis is placed at the top, only the sample holder and the R axis stage move during rotational movement, and the axis stages placed below the R axis stage do not move. Therefore, when determining whether there is overlap during rotational movement, it is sufficient to determine whether there is overlap between the sample holder and the R axis stage and the static object. On the other hand, when tilting movement around the T axis occurs, it is difficult to determine whether there is overlap during rotational movement on a two-dimensional plane.

[0180] Therefore, it is conceivable to calculate a point cloud (i.e., a point cloud representing the trajectory of the rotational movement of a dynamic object) that fills the path of the rotational movement, convert the coordinates of the point cloud into coordinates that indicate the position of an arbitrary axis stage, and then compare the point cloud information with the voxel group information. In this case, in order to compare the point cloud with the voxel group, it is necessary to calculate the position of each point included in the point cloud at an interval equal to or less than the size of the voxel (in other words, the density). However, the farther the position of the point included in the rotating point cloud is from the R axis, the longer the path of the rotational movement becomes, and the number of points to be calculated increases accordingly. In addition, the amount of calculation increases depending on factors such as the amount of rotation around the R axis and the total number of points representing the sample holder and the R axis stage. If the amount of calculation increases, the time required to determine whether there is an overlap increases accordingly, and the speed of the determination decreases.

[0181] In order to deal with this, in the eighth embodiment, the generating unit 14 generates point cloud information representing the dynamic object for each range of a predetermined rotation angle centered on the axis R based on a three-dimensional model of the dynamic object. Specifically, the generating unit 14 rotates the three-dimensional model of the dynamic object by a predetermined rotation angle centered on the R axis, and generates a three-dimensional model representing the trajectory of the rotation movement. The generating unit 14 generates a point cloud representing the trajectory of the rotation movement based on the three-dimensional model representing the trajectory of the rotation movement, and generates point cloud information representing the point cloud. Since the trajectory of the rotation movement represents the path of the rotation movement, the three-dimensional model representing the trajectory of the rotation movement is a model that fills the path of the rotation movement. Therefore, the point cloud representing the trajectory of the rotation movement is a point cloud that fills the path of the rotation movement. Hereinafter, the three-dimensional model representing the trajectory of the rotation movement is referred to as a "rotation model", and the point cloud generated based on the rotation model is referred to as a "rotation point cloud". In addition, the three-dimensional model of the dynamic object when not rotating is referred to as a "standard model".

[0182] A specific example of the eighth embodiment will be described with reference to Fig. 32. In Fig. 32, a standard model 124, a rotated model 126, and a group of rotated points 128 of a dynamic object are shown.

[0183] The standard model 124 is a three-dimensional model that represents the sample holder when not rotated around the R axis. In other words, the standard model 124 is a three-dimensional model that represents the sample holder when the rotation angle is 0°.

[0184] The rotation model 126 is a three-dimensional model that represents the trajectory of the rotational movement when the standard model 124 is rotated 90° around the R axis (i.e., the path of the rotational movement of the standard model 124). The rotation model 126 is a three-dimensional model that represents the trajectory of the rotational movement in a rotation angle range of 0° to 90°, that is, a three-dimensional model that covers the range of 0° to 90°. The rotation model 126 is generated by the generation unit 14.

[0185] The generating unit 14 generates a rotation point group 128 representing a trajectory of a rotational movement based on the rotation model 126, and generates in advance point cloud information representing the rotation point group 128. The rotation point group 128 is a point group representing a trajectory of a rotational movement in a rotation angle range of 0° to 90°, and is a point group covering the range of 0° to 90°. The point cloud information representing the rotation point group 128 is stored in the storage device 30 in advance.

[0186] When determining whether or not there is overlap between a dynamic object and a static object when the rotation angle range of the dynamic object is 0° to 90° (determining whether or not there is overlap within the range of 0° to 90°), the determination unit 18 determines whether or not there is overlap between the dynamic object and the static object within that range of rotation angle by comparing point group information representing the rotation point group 128 with voxel group information of the static object. This makes it possible to determine whether or not there is overlap within the range in which the dynamic object rotates.

[0187] The following describes a determination when the rotation amount of a dynamic object is greater than the previously assumed rotation amount. For example, a rotation point group 128 corresponding to a rotation angle range of 0° to 90° is generated in advance, and the presence or absence of overlap is determined within the range of 0° to 360°.

[0188] The generator 14 converts the coordinates of the rotation point group 128 into coordinates within a target angle range, thereby generating a rotation point group that covers the target angle range.

[0189] Specifically, the generation unit 14 generates a rotation point group covering a range of 90° to 180° by rotating the rotation point group 128 by 90° around the R axis, generates a rotation point group covering a range of 180° to 270° by rotating the rotation point group 128 by 180°, and generates a rotation point group covering a range of 270° to 360° by rotating the rotation point group 128 by 270°. In this way, a rotation point group covering a range of 0° to 360° is generated, and the determination unit 18 determines whether or not there is an overlap between the dynamic object and the static object by collating point group information representing the rotation point group with voxel group information of the static object.

[0190] An example of determining whether or not there is an overlap in the range of 0° to 140° will be described below with reference to Fig. 33. Fig. 33 shows a point group of a dynamic object.

[0191] As described above, the rotation point group 128 is generated in advance. The generation unit 14 generates the rotation point group 130 covering the range of 50° to 140° by rotating the rotation point group 128 by 50° around the R axis. The rotation point group 132 is a point group consisting of the rotation point groups 128 and 130. The determination unit 18 determines whether or not there is an overlap between the dynamic object and the static object by collating point group information representing the rotation point group 132 with voxel group information of the static object. In this way, when the rotation point group 128 is generated in advance, the rotation point group 132 covering the range of 0° to 140° can be generated by one calculation. Including the number of times to calculate the rotation point group 128, only two calculations are required in total.

[0192] In contrast, when using a point cloud generated based on a standard model instead of a rotation point cloud generated based on a rotation model, a point cloud is generated each time by rotating the standard model around the R axis for each unit rotation angle (e.g., 0.1° or 1°, etc.), and the generated point cloud is used to determine whether or not there is overlap. For example, when determining whether or not there is overlap in the range of 0° to 140°, if the unit rotation angle is 1°, 141 calculations are required. The smaller the unit rotation angle, the greater the amount of calculations, and as a result, the longer the time required to determine whether or not there is overlap. In the eighth embodiment, such calculations are not required, and therefore the time required for determination can be shortened.

[0193] The narrower the range of rotation angles covered by the rotation model, the greater the number of coordinate transformations, which increases the amount of calculation and the time required for determination. For example, when a rotation model covering the range of 0° to 10° is generated in advance, the number of coordinate transformations is 36 when determining whether or not there is an overlap in the range of 0° to 360°. In this way, the number of coordinate transformations increases, but the calculation time is shorter than when a point cloud for filling the path of rotational movement is calculated without using a rotation model. Therefore, the time required for determination can be shortened compared to when a rotation model is not used.

[0194] Alternatively, multiple rotation models with different rotation angle ranges may be generated in advance. For example, the generation unit 14 generates a rotation model covering the range of 0° to 90° and a rotation model covering the range of 0° to 10°, and generates a rotation point group covering the range of 0° to 90° and a rotation point group covering the range of 0° to 10°.

[0195] When the range of rotation angles of the dynamic object is less than 90°, the generation unit 14 generates a rotation point group covering the range of rotation angles based on the rotation point group covering the range of 0° to 10°. The determination unit 18 compares point group information representing the rotation point group generated by the generation unit 14 with voxel group information of the static object. In this way, by performing coordinate transformation a maximum of nine times, it is possible to determine the presence or absence of overlap within a range of less than 90°.

[0196] For example, when determining whether or not there is overlap in the range of 0° to 80°, the generation unit 14 generates a rotation point group covering the range of 0° to 80° by rotating the rotation point group covering the range of 0° to 10° by 10° increments for a total of eight times. The determination unit 18 determines whether or not there is overlap by collating point group information representing the rotation point group covering the range of 0° to 80° with voxel group information of the static object.

[0197] On the other hand, when determining whether or not there is overlap in a range of 90° or more, the determination unit 18 determines whether or not there is overlap by comparing point group information representing a rotation point group covering the range of 0° to 90° with voxel group information of the static object.

[0198] Furthermore, the generating unit 14 may generate a rotation point cloud according to the purpose and characteristics of a dynamic object (eg, a sample holder).

[0199] For example, when it is assumed that the sample holder is used with a rotation angle in the range of 0° to 90°, the generating unit 14 generates a rotation point group covering the range of 0° to 90° as a rotation point group for the sample holder in advance. When the sample holder is used, the determining unit 18 determines whether or not there is an overlap using the rotation point group covering the range of 0° to 90°.

[0200] When it is assumed that the sample holder will be used with a rotation angle in the range of 0° to 50°, the generating unit 14 generates a rotation point group covering the range of 0° to 50° as a rotation point group for the sample holder in advance. When the sample holder is used, the determining unit 18 determines whether there is an overlap using the rotation point group covering the range of 0° to 50°.

[0201] The following describes how to determine when the rotation amount of the dynamic object is smaller than the previously assumed rotation amount. When high accuracy is required for the rotation position and calculation speed is not important, the generator 14 generates a three-dimensional model that fills the path of the rotation movement by rotating the standard model, and generates point cloud information based on the three-dimensional model.

[0202] When high accuracy is not required for the rotation position and calculation speed is important, the determination unit 18 determines whether or not there is overlap by using the rotation point group that covers the smallest range of angles among multiple rotation point groups generated in advance. For example, in a case where a rotation point group covering a range of 0° to 90° and a rotation point group covering a range of 0° to 10° are generated in advance, the determination of whether or not there is overlap when the rotation angle is 3° will be described. In this case, the determination unit 18 determines whether or not there is overlap by using the rotation point group covering a range of 0° to 10°. As a result, although the rotation angle exceeds 7°, the calculation cost is low and the time required for the determination can be shortened.

[0203] The generating unit 14 may generate in advance a group of rotation points that covers a finer range. For example, the generating unit 14 generates in advance a group of rotation points that covers a range of 0° to 1°. By using this group of rotation points, it is possible to improve the accuracy of the rotation position and shorten the time required for determination.

[0204] The above-mentioned sample holder is a structure having an asymmetric shape with respect to the R axis. When such a structure is included in a dynamic object, it is necessary to precisely determine whether or not the dynamic object and a static object collide with each other due to the rotation of the dynamic object. In the above-mentioned example, in order to prevent a decrease in the speed of the determination and to increase the accuracy of the determination, a plurality of rotation point groups with different ranges of rotation angles are generated in advance.

[0205] On the other hand, when a sample holder having a rotationally symmetric shape is used, the shape may have little effect on collision determination. In this case, the generation unit 14 may generate point cloud information covering a range of 0° to 360°. For example, as shown in FIG. 34, when a sample holder 134 having a rotationally symmetric shape is included in a dynamic object, the generation unit 14 generates a rotation model covering a range of 0° to 360° by rotating a three-dimensional model of the sample holder 134 by 360° around the R axis. FIG. 35 shows a rotation model 136 covering a range of 0° to 360°. By using the rotation model 136, it is not necessary to set the range of the rotation angle in detail, and the amount of calculation is reduced, so that the presence or absence of overlap can be determined more quickly.

[0206] Example 9 In the ninth embodiment, the determination unit 18 determines whether or not the dynamic object overlaps with the static object by comparing, for each coordinate axis, the voxel group information representing a static object present on the moving path of the dynamic object with the point group information representing the dynamic object. When it is determined that the dynamic object overlaps with the static object in the determination for a certain coordinate axis, the determination unit 18 changes the coordinate axis to be compared and determines whether or not the dynamic object overlaps with the static object for each coordinate axis. In this way, when the dynamic object and the static object overlap, the determination unit 18 searches for a path where the dynamic object and the static object do not overlap by changing the coordinate axis to be compared. The output unit 20 may display information indicating a path where the dynamic object and the static object do not overlap on a display.

[0207] For example, the dynamic object is moved or rotated in the order of the X-axis, the Y-axis, and the R-axis. That is, the dynamic object is moved along the X-axis to a target position on the X-axis, then moved along the Y-axis to a target position on the Y-axis, and then rotated to a target angle around the R-axis. In this case, the determination unit 18 determines whether or not the dynamic object overlaps with the static object for each coordinate axis according to the order. For example, when the amount of movement along the X-axis and its movement direction, the amount of movement along the Y-axis and its movement direction, and the amount of movement (i.e., rotation amount) and its movement direction (i.e., rotation direction) around the R-axis are input to the charged particle beam device 12, information indicating these amounts of movement and movement directions is output from the charged particle beam device 12 to the collision determination device 10. Based on the information, the determination unit 18 calculates a target position of movement along the X-axis, a target position of movement along the Y-axis, and a target angle of rotation around the R-axis.

[0208] The determination unit 18 determines whether or not the dynamic object and the static object overlap when the dynamic object moves to a destination position along the X-axis. If the dynamic object and the static object do not overlap when the dynamic object moves to a destination position along the Y-axis, the determination unit 18 then determines whether or not the dynamic object and the static object overlap when the dynamic object moves to a destination position along the Y-axis. If the dynamic object and the static object do not overlap when the dynamic object moves to a destination position along the Y-axis, the determination unit 18 then determines whether or not the dynamic object and the static object overlap when the dynamic object rotates to a destination angle around the R-axis. The output unit 20 displays information indicating the determination result on the display during each determination.

[0209] For example, when it is determined that the dynamic object and the static object overlap when the dynamic object moves along the Y axis, the output unit 20 causes the display to display information indicating that the dynamic object and the static object overlap due to the movement along the Y axis. The output unit 20 may cause the display to display information prompting the user to change the movement path of the dynamic object from the movement path along the Y axis to the movement path along the R axis.

[0210] Next, the determination unit 18 determines whether or not the dynamic object and the static object overlap when the dynamic object rotates to a target angle around the R axis. If it is determined that the dynamic object and the static object do not overlap as a result of the rotational movement, the determination unit 18 determines whether or not the dynamic object and the static object overlap when the dynamic object moves along the Y axis. Furthermore, the output unit 20 may display, on the display, information that prompts the user to move the dynamic object along the Y axis after rotating the dynamic object around the R axis.

[0211] For example, if it is determined that the dynamic object and the static object do not overlap when the dynamic object moves in the order of the X-axis, Y-axis, and R-axis, the output unit 20 displays information indicating that on the display. In this case, when the dynamic object is moved in the order of the X-axis, Y-axis, and R-axis, the dynamic object can be moved to a target position and a target angle without colliding with the static object.

[0212] On the other hand, if it is determined that the dynamic object and the static object overlap when the dynamic object moves in the order of the X-axis, Y-axis, and R-axis, but it is determined that the dynamic object and the static object do not overlap when the dynamic object moves in the order of the X-axis, R-axis, and Y-axis, the output unit 20 displays information indicating that on the display. When the dynamic object is moved in the order of the X-axis, R-axis, and Y-axis in accordance with that information, the dynamic object can be moved to a target position and a target angle without colliding with the static object.

[0213] Each of the above-described embodiments may be implemented independently, or a plurality of the embodiments may be combined and implemented. [Explanation of symbols]

[0214] 10 collision determination device, 12 charged particle beam device, 14 generation unit, 16 database, 18 determination unit, 20 output unit, 22 calculation unit, 24 display unit.

Claims

1. a determination means for determining whether or not the dynamic object overlaps with the static object by comparing point cloud information representing a three-dimensional shape of a dynamic object whose position changes within a sample chamber of a charged particle beam device as a point cloud, the point cloud information including position information of the dynamic object, with voxel group information representing a three-dimensional shape of a static object whose position does not change within the sample chamber as a voxel group, the voxel group information including position information of the static object and compiled into a database; an output means for outputting a determination result of the determination means; Including, A coordinate system is defined that includes a three-dimensional orthogonal coordinate system that is composed of three mutually orthogonal coordinate axes, the determining means determines whether or not the dynamic object overlaps with the static object by comparing, for each coordinate axis, voxel group information representing the static object present on the movement path of the dynamic object with point group information representing the dynamic object; A collision determination device comprising:

2. The collision determination device according to claim 1 , the coordinate system further includes a tilt axis of the dynamic object; the determination means, in a state in which the dynamic object is tilted around a tilt axis, compares, for each coordinate axis, voxel group information representing the static object present on the movement path of the dynamic object with point group information representing the dynamic object, thereby determining whether or not the dynamic object overlaps with the static object; A collision determination device comprising:

3. The collision determination device according to claim 1 , the coordinate system further includes an axis of rotation of the dynamic object; point cloud information representing the dynamic object is generated for each range of a predetermined rotation angle; the determining means determines whether or not the dynamic object and the static object overlap by comparing point group information corresponding to a range of a rotation angle of the dynamic object with voxel group information representing the static object. A collision determination device comprising:

4. The collision determination device according to claim 1 , The determination means is determining whether or not the dynamic object overlaps with the static object by comparing, in order, voxel group information representing the static object present on the movement path of the dynamic object with point group information representing the dynamic object for each coordinate axis; if it is determined that the dynamic object and the static object overlap, the order of the coordinate axes to be checked is changed, and for each coordinate axis, voxel group information representing the static object present on the movement path of the dynamic object is checked against point group information representing the dynamic object, thereby determining whether or not the dynamic object and the static object overlap. A collision determination device comprising:

5. The collision determination device according to claim 1 , Voxel group information for each state of the static object is generated; the determining means further determines whether or not the dynamic object overlaps with the static object by comparing voxel group information corresponding to the state of the static object with point group information representing the dynamic object. A collision determination device comprising:

6. The collision determination device according to claim 5, the static object includes a structure that moves back and forth between a first position and a second position; the state of the static object is a state in which the structure is at the first position, a state in which the structure is at the second position, or a state in which the structure is reciprocating between the first position and the second position; A collision determination device comprising:

7. The collision determination device according to claim 1 , the determining means further determines whether or not the dynamic object overlaps with the static object by comparing point cloud information representing the dynamic object according to the observation conditions by the charged particle beam device with voxel group information representing the static object. A collision determination device comprising:

8. The collision determination device according to claim 7, The observation conditions are conditions related to a voltage applied to a sample to be observed. A collision determination device comprising:

9. The collision determination device according to claim 1 , a user interface for receiving an input of a size of a sample to be mounted on the dynamic object; the determination means determines whether or not the dynamic object overlaps with the static object by comparing point cloud information representing a point cloud of the dynamic object reflecting the size accepted by the user interface with voxel group information representing the static object; A collision determination device comprising:

10. The collision determination device according to claim 1 , a calculation means for calculating a height of the sample based on a tilt operation and a movement operation in a height direction of the dynamic object on which the sample is mounted, the determining means determines whether or not the dynamic object overlaps with the static object by comparing point cloud information representing a point cloud of the dynamic object reflecting the height calculated by the calculating means with voxel group information representing the static object. A collision determination device comprising:

11. Computer, a determination means for determining whether or not the dynamic object overlaps with the static object by comparing point cloud information representing a three-dimensional shape of a dynamic object whose position changes within a sample chamber of a charged particle beam device as a point cloud, the point cloud information including position information of the dynamic object, with voxel group information representing a three-dimensional shape of a static object whose position does not change within the sample chamber as a voxel group, the voxel group information including position information of the static object and compiled into a database; an output means for outputting a determination result of the determination means; Functioning as a A coordinate system is defined that includes a three-dimensional orthogonal coordinate system that is composed of three mutually orthogonal coordinate axes, the determining means determines whether or not the dynamic object overlaps with the static object by comparing, for each coordinate axis, voxel group information representing the static object present on the movement path of the dynamic object with point group information representing the dynamic object; A program characterized by:

12. generating point cloud information representing the three-dimensional shape of a dynamic object whose position changes in a sample chamber of a charged particle beam device as a point cloud, the point cloud information including position information of the dynamic object; generating point cloud information representing the three-dimensional shape of a static object, the point cloud information including position information of the static object, based on a three-dimensional model representing the three-dimensional shape of the static object whose position does not change within the sample chamber; generating a database of voxel group information representing a three-dimensional shape of the static object using a voxel group based on the point cloud information of the static object; A coordinate system is defined that includes a three-dimensional orthogonal coordinate system that is composed of three mutually orthogonal coordinate axes, determining whether or not the dynamic object overlaps with the static object by comparing, for each coordinate axis, voxel group information representing the static object present on the movement path of the dynamic object with point group information representing the dynamic object; A collision determination method comprising:

Citation Information

Patent Citations

  • Collision avoidance system for object representation on basis of voxel

    JP1996241437A

  • Inter-object contact interaction simulating device

    JP2012094128A

  • Sample measurement method using charged particle beam system and scanning electron microscope

    JP2020126721A

  • Trajectory generation device

    JP2021024019A

  • Collision avoidance system for voxel-based object representation

    JP2726811B2