WO2019049260A1 - Image processing device - Google Patents

Abstract

In the present invention, a charged particle beam device uses an optical system with a lens, and because of this optical system, an image distortion occurs in a sample structure 53, an image of which is formed within an observation visual field 51. A first image (1), which is obtained by the charged particle beam device, and a predicted image (2), (3), or (4), produced respectively on the basis of a deformation function 60, 61, or 62, are created; the created predicted image is compared against a second image, which is obtained by the charged particle beam device after a sample is moved; on the basis of the result of the comparison, it is evaluated whether the deformation function is close to actual distortion characteristics; and a function parameter is adjusted so as to minimize the difference between the predicted image and an image that is actually obtained after the field of vision is moved, thereby obtaining a distortion function for the optical system of the device.

Description

Image processing device

The present invention relates to a measuring device, and more particularly to an image processing technique of a device using a charged particle beam.

Charged particle beam equipment is a generic name for devices that use charged particles, that is, devices that control the movement of particles using an external force applied from an electric field or a magnetic field to electrons or ions, and is typically a transmitted electron. A microscope, a scanning electron microscope, a focused ion beam device and the like can be mentioned.

In these, after accelerating electrons and ions by an electric field in a vacuum environment, the orbit is controlled by a lens using an electromagnetic field and a deflector using an electromagnetic field, and the sample to be observed is observed by irradiating the sample. I do. Although the polarity of the electric field to be used varies depending on the charge of the particles to be used, the common part among them is large, and in the following description, an example in a transmission electron microscope will be described as a representative example.

A transmission electron microscope (hereinafter referred to as TEM) irradiates an electron beam accelerated by a high voltage to a substance to be observed, forms an electron beam transmitted through the substance by an electromagnetic lens, and magnifies the fine structure of the substance. It is a device that performs observation and uses a lens electromagnetic field that acts on electrons. Then, in the optical system formed by this lens electromagnetic field, distortion is added to the obtained observation image in some cases due to the imperfection of the lens that forms an image, which will be described later with reference to the drawings. It is known to be observed in a form different from the sample structure.

The imperfection of the lens causing such distortion is difficult to solve as far as using an electromagnetic field as a lens, and in general, observation using only the central part of the visual field where distortion does not easily increase and conditions of the lens difficult to generate distortion Is addressed by making observations in However, such a countermeasure will limit the size of the field of view that can be observed and the conditions of the lenses that can be used. Therefore, as a separate countermeasure, the computer performs correction processing on the image recorded by the recording apparatus. Patent Document 1 proposes a method of obtaining an image in which the influence of distortion is reduced.

In the example of Patent Document 1, assuming that the electron beam diffraction pattern from a sample whose sample structure can be regarded as known has regularity, the irregularity of the pattern is used to measure the deformation given to the pattern by the optical system to correct the distortion. A technique is disclosed for determining the parameters to do so. In this case, strain can be measured by using a sample whose structure is known with high accuracy, for example, a crystalline sample or a sample prepared or processed with high accuracy, but on the other hand, under conditions in which the use of an appropriate sample is difficult. It is difficult to apply. In order to measure distortion in such a case, Patent Document 2 proposes a method of performing measurement using a sample whose structure is not known.

In the example of Patent Document 2, images of two fields of view in which a part of the area is common are acquired on the sample. A specific area in one image is cut out, and a portion including the same structure as the area is detected in the other image by matching processing. By performing such an evaluation on a plurality of specific areas in the image, information on distortion contained in the image is measured.

International Publication Number WO2009 / 123311 Patent No. 5596141

According to the method of Patent Document 2 described above, distortion can be measured without using a known sample, but since the matching processing of the local region used for the measurement is necessarily performed on the image including the distortion, the distortion is In the case where the direction and size of the image change in the image, the results of different distortions applied to the same structure are compared, and the accuracy of the position detected by the matching may be reduced due to the influence of the distortion. In addition, since the accuracy of matching is affected by the pattern used for evaluation, the distortion measurement accuracy may be reduced as a result in an area having a structure unsuitable for matching, such as an area with a small change in contrast in the sample. There is a possibility of getting different measurement results.

An object of the present invention is an image processing apparatus that performs measurement with little influence of distortion itself when measuring image distortion caused by an optical system, in order to solve the above problems, and realizes image distortion measurement using the result To provide.

In order to achieve the above object, according to the present invention, an image processing apparatus is provided which comprises a first image of a sample or a part thereof taken by a charged particle beam device and a first distortion function. An image processing apparatus configured to generate a predicted image and compare the generated first predicted image with a second image of a sample or a part of the image captured after movement of the field of view by the charged particle beam device I will provide a.

According to the present invention, it is possible to accurately measure distortion or aberration of the optical system of the charged particle beam device.

It is the figure which showed typically the optical system at the time of observation in a transmission electron microscope. FIG. 8 is a diagram showing an example of the result of distortion applied to an image in an optical system. FIG. 7 is a diagram showing an example of a ray diagram in the case where no distortion occurs and an example of a ray diagram in the case where distortion occurs in the optical system. FIG. 6 is a diagram showing an example of a distortion function applied to an image in an optical system. FIG. 7 is a diagram showing an example of means for changing the field of view of an image according to the first and second embodiments. FIG. 7 is a view showing an example of an image change when an observation field of view is changed in an optical system having distortion according to the first embodiment. FIG. 7 is a diagram showing an example of the relationship between an apparent deformation function and a distortion function generated by a change in visual field according to the first embodiment. FIG. 7 is a diagram showing an example in which an image is deformed by a plurality of different deformation functions according to the first embodiment. It is a figure which shows an example of the result of calculating | requiring the square of the difference with an acquired image with respect to the result of applying a different deformation | transformation function based on Example 1. FIG. FIG. 7 is a diagram showing an example of change in correlation value when comparison with acquired images is performed on a plurality of images deformed by a plurality of different deformation functions according to the first embodiment. FIG. 17 is a diagram showing a detailed flow of most likely distortion function evaluation based on a distortion function according to a third embodiment. FIG. 16 is a diagram showing a detailed flow of most probable distortion function evaluation based on a plurality of distortion functions according to a fourth embodiment. FIG. 18 is a diagram showing an extraction region when extracting only a part of the acquired image and evaluating distortion characteristics in one direction according to a fifth embodiment; FIG. 18 is a diagram showing an extraction region when extracting only a part of the acquired image and evaluating local distortion characteristics according to a sixth embodiment. As supplementary information at the time of creating a distortion function according to the seventh embodiment, a result of finding a distribution of square values of image differences and an example of representing a square value distribution of differences as a distribution with respect to a distance from an image center are shown. FIG. FIG. 21 is a diagram showing an example of a change in brightness distribution of an image as an influence caused by image distortion according to an eighth embodiment.

Hereinafter, embodiments of the present invention will be sequentially described with reference to the drawings. However, in order to deepen understanding of the contents of the present invention, the problems of the present invention described above will be described with reference to FIGS. In the drawings of the present specification, the same numbers indicate the same items.

FIG. 1 is a view schematically showing an optical system at the time of image observation of a TEM. The irradiation system lenses 2 and 3 disposed below the electron source 1 form lens electromagnetic fields 21 and 22 that act on electrons, and further below the apertures 7 and 10 that block part of the electron beam, electrons The deflectors 8 and 9 for deflecting the line, the objective lens 4 and the objective lens electromagnetic field 23 formed thereby, the sample 14 and the lens electromagnetic field 24 formed by the imaging system lenses 5 and 6 arranged thereunder And a recording device 11 for observing and recording an image formed by the electron beam.

The electron beam emitted from the electron source 1 passes through the irradiation system lenses 2 and 3 and then irradiates the sample 14. The electron beam transmitted through the sample 14 further passes through the imaging system lenses 5 and 6, and an image of the sample is formed on the imaging device 11. The intensity, size, and the like of the image formed at this time are changed by appropriately controlling the irradiation system lenses 2 and 3, the objective lens 4 and the imaging system lenses 5 and 6 from the control device 13. In the observation, the image formed on the recording device 11 is directly observed or recorded by the recording device controller 12 and observed. The observation image data recorded by the recording device control device 12 is subjected to image processing by a central processing unit (CPU) such as a personal computer (PC) connected to the recording device control device 12 or to the recording device control device 12. Ru. In the present specification, the recording device control device 12, the control device 13, and further the CPU are collectively referred to as a control unit.

With such a configuration, the transmission electron microscope image observes the structure contained in the sample 14 in an enlarged manner on the recording device 11 under the control of the control unit, and the structure on the sample has its two-dimensional position It is projected onto 11 with the relationship maintained approximately. The image formation is mainly performed by an optical system constituted by the objective lens 4 and the imaging system lenses 5 and 6. Then, in some cases, distortion is added to the observation image in some cases due to the imperfection of the lens that forms an image as described above, and the image is observed as a form different from the original sample structure.

FIG. 2 is a diagram showing an example of how image distortion occurs. (1) of the figure shows an example of the structure of the sample to be observed. An example of observing a sample having such a structure with a distorted optical system is shown in FIG. 2 (2). The image distortion caused by the optical system is often small at the central portion and large at the outer side. In the example of FIG. 2 (2), the structure appears to move further outward than the original position as the outside of the field of view, and the degree changes continuously from the center of the distortion to the outside, The image is an image stretched outward.

Such image distortion is mainly caused by the aberration of the lens, and the principle thereof is schematically shown in FIG. (1) of the same figure is the figure which showed the image formation by the ideal lens without aberration. Electron trajectories 36 emitted from a point separated by a distance r ₁ from the optical axis 30 on the sample reach a point separated by a distance R ₁ from the optical axis 30 on the image plane 15. The ratio of r ₁ to R ₁ at this time is the imaging magnification of the optical system. Further, the electron orbit 37 emitted from the point on the sample separated by r2 from the optical axis 30 reaches the point on the image plane 15 separated by R2 from the optical axis 30. At this time, the ratio of r2 to R2 is equal to the ratio of r1 to R1, and an image is formed on the image plane 15 at the same magnification.

On the other hand, in an actual electromagnetic field lens, an aberration, in particular, a spherical aberration exists as a large component, and an example of image formation at that time is shown in FIG. As in the case of (1), the electron orbit 36 incident on the lens 26 from the point decentered from the optical axis 30 by r ₁ passes the orbit as shown by 35 due to the influence of the aberration. Since this orbit receives stronger convergence than the orbit 33 which passes without aberration, it passes an orbit such as 34 and forms an image on the image plane 15 at a position separated by D1 outside the optical axis than the orbit 33 . Next, considering electron trajectories 37 coming from a point at a distance of R2 further on the sample than the optical axis, the trajectories pass a point further away from the optical axis 30 at the lens 26 than the electron trajectories 36 In that case, because it receives a stronger convergence than the trajectory 32 which passes without aberration, it passes a trajectory such as 34, and on the image plane 15, it is an image at a position separated by D2 outside the optical axis than the trajectory 32. Connect At this time, the ratio of R1 + D1 to r1 and the ratio of R2 + D2 to r2 are different values. This means that the electron trajectories from different points on the sample on the image plane 15 are connected at different magnifications. It will be an image. This appears as the contraction of the observation image on the image plane, which causes the image distortion as shown in FIG.

Since the image distortion caused by the aberration of such a lens changes according to the distance from the center of distortion, the characteristic can be expressed as a function of distortion according to the distance from the center. Figure 4 shows some examples of such distortion functions, where the horizontal axis is the distance from the center and the vertical axis shows the amount of displacement due to distortion from the position where it is imaged in the absence of distortion. If the vertical axis is a positive value, the trajectory corresponds to the outside, and if it is a negative value, the trajectory corresponds to the inward displacement. In the example of FIG. 4 (1), as the trajectory goes to the outside, it represents a distortion that is further shifted to the outside than the original position, and a distortion generally called pincushion distortion occurs. In the example of (2), as the orbit goes to the outside, it represents a distortion that is shifted inward from the original position, and a distortion generally called barrel distortion occurs. The example of (3) is an example of distortion that occurs in an optical system in which a plurality of lenses are combined, but the distortion takes both positive and negative values according to the distance from the center, and the image is drawn to the outside inside the image. There is a distortion that shrinks the image to the inside outside the image. In the example of (4), an example is shown in which an asymmetric distortion occurs due to the influence of the axis deviation between lenses or the like.

When the acquired image is expressed as an intensity distribution I (x, y) in a two-dimensional x, y coordinate space, such distortion function is defined as coordinates of the distortion center as x ₀ , y ₀ and relative coordinates from the distortion center _{X = x -x 0, Y =} y -y 0 to the the following examples function D (X, Y) can be expressed as.

By changing each of the coefficients A ₁ to A ₃ , B ₁ to B ₃ , C ₁ to C ₄ , E, and x ₀ , y ₀ in the equation ₁ , it is possible to represent distortion functions of various shapes. It is possible. The distortion function of Equation 1 is a polynomial function that includes a fourth-order term or less for variables X and Y. In many cases, since the distortion function often has a rotationally symmetric shape, among the coefficients in Equation 1, A ₁ to A ₃ , B ₁ to B ₃ , E, etc. can be regarded as zero.

Then, the lens imperfection causing such distortion is difficult to solve as long as an electromagnetic field is used as the lens, and the image recorded by the recording apparatus is corrected on the computer to reduce the influence of distortion. Although the method of obtaining has been proposed, there is a possibility that the measurement accuracy of the distortion may be lowered or the measurement result different from the actual distortion may be obtained.

Hereinafter, various embodiments of the present invention for solving such problems will be sequentially described according to the drawings.

Example 1 is an image processing apparatus, which creates and generates a first predicted image from a first image of a sample or a part thereof taken by a charged particle beam device and a first distortion function. This is an embodiment of an image processing apparatus configured to compare the first predicted image with a second image of a sample or a part of the image captured after movement of the field of view by the charged particle beam device. This image processing apparatus can be configured by program processing or the like in the control unit such as the recording apparatus controller 12 of the charged particle beam apparatus or a PC.

(1) of FIG. 5 is a diagram showing an example of the apparatus configuration according to the first embodiment. The electron orbit 37 emitted from the sample 14 passes through the objective lens electromagnetic field 23 disposed thereunder and the lens electromagnetic fields 24 and 25 of the imaging system to form an image on the recording device 11, and the image is a first image Are recorded by the recording device controller 12 as After that, when the sample 14 is moved by m ₁ by the movement mechanism of the sample table, the electron beam emitted from the same place on the sample passes a trajectory as shown by 36 and a trajectory like 36 on the recording device 11 As a result, the sample 14 is imaged at a position moved by M1 from the imaging position of the trajectory 37 before the sample 14 is moved. That is, the charged particle beam device of the configuration of the present embodiment includes a moving mechanism for moving the sample, and moves the sample by the moving mechanism while capturing the first image and the second image.

As a result, an image at a position moved by m ₁ on the sample is formed on the recording device 11, and the image is recorded by the recording device controller 12 as a second image. In this case, since the position of the sample with respect to the optical system constituted by the lenses 23, 24 and 25 causing distortion is changed, the distortion applied to the image does not move with respect to the movement of the field of view. An image with the same distortion applied to different fields of view is observed.

FIG. 6 shows an example of the change in the observation image due to the movement of the visual field. In the figure, (1) shows an example of a first image which is an observation image, and (2) shows an example of a second image which is an observation image after movement. That is, since distortion occurs regardless of the observation position on the sample, corresponding to the position in the observation field 51, when the observation area on the sample is changed by moving the sample, etc., an image is formed in the field of view The sample structure 52 at the same place on the sample to be measured is observed with different distortion depending on the position in the observation field 51 of observation. This is a distortion function that is a function of a variable based on coordinates in an image, as shown in Equation 1, for different positions, for example, by changing the field of view when considering the structural coordinates of the sample as a reference. It can be considered that observation is performed by applying D (X, Y).

Here, assuming that the information on the distortion function applied to the image is known, it is possible to either correct the distortion on the distorted image or add distortion to the undistorted image. Become. Furthermore, if distortion correction is performed on the distorted image, the entire image is further translated, and distortion is applied again, the result obtained is the result obtained before and after moving the field of view in the distorted optical system. It is equivalent to the relationship between the images. Therefore, a predicted image after movement of the visual field created by adding a temporary distortion function and visual field movement to the observation image obtained by the control unit such as the recording apparatus controller 12 of the charged particle beam apparatus or the PC. By comparing the similarity between two images obtained by actually moving the field of view, it can be evaluated whether the assumed distortion function is close to the actual distortion characteristics. That is, the parameters of the distortion function to be assumed are variously changed using the control unit such as the recording apparatus controller 12 or the PC, and the predicted image obtained by each distortion function and the image after the change of the field of view actually obtained By adjusting the parameters of the temporary distortion function so as to reduce the difference of, it is possible to indirectly evaluate the distortion function actually included in the image.

FIG. 7 is a view showing an example of the relationship between an apparent deformation function and a distortion function generated by the movement of the visual field. That is, the characteristics of the coordinate transformation which is finally added to the image by a series of processes of the correction of the image distortion, the movement of the visual field and the application of the image distortion in the control unit described above are represented as one function. The horizontal axis indicates the coordinate X before moving the field of view, and the vertical axis indicates the coordinate X 'at which the same region is arranged in the image after conversion. This function can be defined if the coefficient defining the shape of the distortion function, the shift amount of the image at the time of movement of the field of view, and the coordinates of the distortion center are determined.

FIG. 8A shows deformation functions 60, 61 and 62 determined for a plurality of distortion parameters, and FIG. 8B shows a sample structure 53 imaged in the observation field of view 51. An example of the first image in the observation field of view 53 obtained by observing a certain field of view on the sample is shown in (b) (1). An example of a predicted image as a result of applying the transformation function 60 of (a) to the image of (1) of (b), an example of applying the transformation function 61 to (2) and an example of applying the transformation function 61 to (3) An example to which 62 is applied is shown in (4). By changing the distortion parameters used when creating the deformation function, it is possible to obtain a plurality of predicted images under different distortion conditions as shown in FIG. 8 (b). The generation of the predicted image is a process of deforming an image such as the first image in the coordinate space, and the deformation process also includes movement.

FIG. 9 shows an example of the result of comparison of each of these predicted images with the second image which is an image obtained by actually moving the field of view in the charged particle beam device by the control unit. In addition, although the case where one sheet of image is used as a 2nd image is illustrated in FIG. 9, the some 2nd image which varied the movement amount of visual field movement is image | photographed, and it compares with an estimated image. Also good. In this case, the expected image is compared with a plurality of second images having different amounts of movement of the visual field by the charged particle beam device.

There are various methods for comparison, but here, for example, the difference between the intensities of two images is taken for each pixel, and the distribution of the result obtained by squaring is shown. In addition, as a method of performing comparison, any one or more of simple difference amount, sum, kan, quotient, correlation value of two images, histogram of two images, Fourier transformation result, and feature point of image are calculated. There is processing, and any of them can be realized by program execution of the control unit.

Fig. 9 (1) is an example of the result of comparison of images obtained using a distortion function that is significantly different from the actual distortion function, and (2) in the figure is more actual than the example of (1) Is an example of the result of comparison of images obtained using functions close to the distortion function of. When (1) and (2) are compared, since the difference between the two images compared is larger in (1), there are many white areas where the square of the difference has a value I understand. For the comparison results obtained in this way, it is possible to evaluate the similarity of the image, for example, by summing the intensities for all pixels. Such an evaluation index is known as a sum of squared difference (SSD) value.

The SSD value of the second image which is the observation result obtained by actually moving the field of view with respect to the plurality of results obtained by the plurality of deformation functions obtained by changing the parameters of the distortion function in FIG. An example showing the relation of When the value of the distortion parameter (Distortion parameter) defining the shape of the distortion function is changed from 20 to 60 as an example, it is understood that the SSD value is minimum under the condition that the distortion parameter is 40. This shows that, when the distortion parameter is 40, the difference between the predicted image obtained from the assumed distortion function and the second image obtained by actually moving the field of view is the smallest. Among the evaluated conditions, it can be said that it is the one closest to the actual distortion function. Further, it is possible to improve the accuracy of the distortion function evaluation by finely changing the distortion parameter near the condition where the SSD value is minimum and searching for the condition where the SSD value is further reduced.

When a deformation function is applied, the comparison result may not necessarily correspond to the validity of the distortion function used in the evaluation when the two images to be compared each include information on different regions in the sample. . In such a case, the comparison is performed only in the part corresponding to the common area expected from the movement amount of the visual field, the evaluation is performed in only a partial area such as the center of the visual field, and the weighting is changed according to the position in the image. Using a method such as comparison is also effective in enhancing the accuracy of distortion function evaluation. In the present specification, the common area means, for example, two images before and after movement of the visual field, that is, an area commonly included in the first image and the second image.

According to the configuration of the present embodiment, it is possible to accurately measure distortion or aberration of the optical system of the device using the observation image obtained by the charged particle beam device.

(2) of FIG. 5 is a view showing an example of an apparatus configuration different from the configuration in which the sample itself is moved by the moving mechanism described in the first embodiment with respect to the movement of the visual field performed in performing measurement. In the configuration of the present embodiment, instead of moving the sample 14 to change the field of view of the observation image, the trajectory of the electron beam is changed using the deflector 9 disposed below the sample 14 so that the sample The same effect as moving it is obtained. That is, the optical system of the charged particle beam apparatus of the present embodiment includes a deflector that changes the trajectory of the charged particle beam to be irradiated to the sample, and the deflector makes it possible to capture the first image and the second image. The trajectory of the charged particle beam is changed.

The shift m ₂ of the trajectory of the electron beam generated by the deflector 9 passes through the objective lens electromagnetic field 23 and the imaging system lens electromagnetic fields 24 and 25 disposed below, and finally as a second image on the recording device 11 Connect the image. At that time, the shift amount m ₂ of trajectory given by the deflector 9 is a shift quantity M ₂ on the recording apparatus, which can be considered to be equivalent to the case of moving the sample by m _2. Further, although the deflector 9 is disposed below the sample in the example of (2) of FIG. 5, in addition, it may be disposed between the objective lens 23 and the imaging system lens 24 or between the imaging system lenses 24 and 25 It may be arranged at different places. When a lens is present between the sample 14 and the deflector 9, the distortion can be measured only for the optical system between the deflector 9 and the recording device 11. However, in the case of an optical system that performs enlargement in general, the sample Since the influence of the lens close to the image distortion on the image is small, the configuration of the present embodiment makes it possible to evaluate the distortion component of the optical system substantially sufficiently included in the image.

This embodiment is an embodiment showing a detailed flow of the flow of distortion function evaluation of the image processing apparatus described in the first embodiment. FIG. 11 shows an example of a detailed flow showing the flow of distortion function evaluation of this embodiment. The main processing unit of this flow is the control unit 13 of the charged particle beam apparatus such as the TEM described above, the recording apparatus control apparatus 12, and a control unit such as a PC connected thereto. Under conditions for distortion function evaluation, an image A is acquired as a first image from a certain position on the sample (S1101), and then the observation field of view is changed (S1102), and an image B is acquired as a second image. (S1103). After that, the initial value F ₁ of the distortion function is created, but in that case the initial value may be set based on a typical distortion value, or for the obtained image A, image B or part thereof. by measuring the distribution of the correlation values and the difference value to acquire the supplementary information (S1104), it is also possible to create an initial value F ₁ reasonable distortion function by the basis of the results.

After the initial value F ₁ of distortion function created (S1105), creating a deformation function FD ₁ applied to the image upon the scrolling from the distortion function (S1106). By acting on the deformation function FD ₁ for the image A, image B _exp is created the expected image of the image B (S1107). Subsequently, the predicted image B _exp and the image B, which is the second image actually obtained, are compared to determine the similarity C between them (S 1108). Then, it is determined whether the similarity C is equal to or more than a reference value (S1109). If the similarity C is higher than the reference value (Yes), measurement is terminated with the distortion function at that time as a measurement result (S1111). If the similarity C is lower than the reference value (No), after creating a new distortion function F ₂ based on the evaluation history up to that point and the correlation value evaluation immediately before (S1110), repeat the above evaluation Repeatingly (S1106 to S1110), the parameters of the distortion function are changed so that the similarity C is high. When the similarity C becomes equal to or higher than the reference value in step S1109, the flow is ended using the distortion function at that time as the measured value (S1111).

In the generation of a new distortion function in S1110 based on the comparison result, the flow is finally ended by using the comparison result of the image A and the image B as the supplementary information as the comparison result of the image B and the image B _exp. It is possible to reduce the number of steps required to Further, the similarity C determined by comparing the image B and the image B _exp in S1108 corresponds to the fact that the higher the value is, the more similar the two images are in this example, but the evaluation value is determined depending on the comparison standard used. It is also conceivable that the smaller the is, the higher the similarity, in which case the evaluation of the similarity C is evaluated depending on whether it is lower than the reference value. In addition, it is also possible to determine the end of the flow using the magnitude and slope of the change with respect to the parameter to be changed without setting a clear reference value for the reference value C. When the method of the present embodiment described above is used, the distortion function can be appropriately measured without depending on the type of image used for evaluation.

This embodiment is an embodiment showing a detailed flow of distortion function evaluation using a plurality of distortion functions of the image processing apparatus described in the first embodiment. FIG. 12 shows an example of a flow in the case of evaluating a plurality of distortion functions simultaneously. That is, in the present embodiment, a plurality of first predicted images or second predicted images are created using a plurality of distortion functions. Similar to the flow chart of FIG. 11 of the third embodiment, the processing entity of this flow chart is the control unit 13 of the TEM, the recording apparatus control unit 12, and a control unit such as a PC connected thereto.

In the figure, under the condition to be evaluated, an image A which is a first image is acquired from a certain position on a sample (S1201), and then the observation visual field is changed (S1202), an image which is a second image B is acquired (S1203). After that, a plurality of distortion function candidates F ₁ to F _N determined by assuming a plurality of values for the parameters constituting the distortion function are created, in which case distortion values roughly estimated in the optical system are used as a basis. The initial value may be set in the image A, the distribution of the correlation value or the difference value is measured with respect to the obtained image A, the image B or a part of them, and supplemental information is acquired (SD 1204). It is also possible to create a reasonable initial value of the distortion function by. After distortion function candidates F ₁ to F _N are created (S 1205), a plurality of deformation function candidates FD ₁ to FD _N to be added to the image when the field of view is shifted from those distortion function candidates are created (S1206).

Then, by the action of deformation function candidate FD ₁ ~ FD _N for the image A, predicted images of a plurality of images B, the image _{B exp1} ~ B expN is obtained (S1207). The image B _exp1 to B _expN is compared with the image B actually obtained to determine the similarity to each of the predicted images (S1208), and the predicted image having the highest similarity among the images B _exp1 to B _{expN Is} determined as the expected image B _{exp_best} (S1209). After that, the distortion function F _best used when generating the predicted image B _{exp_best} may be determined as the final one as it is, and it is judged whether the similarity obtained for B _{exp_best} is a reference value or more. (S1210) If the similarity is equal to or higher than the reference value (Yes), the flow is ended with the distortion function F _best used when generating the predicted image B _{exp_best} as a measurement result (S1212). As a result of the determination in S1210, when the similarity is equal to or less than the reference value (No), after changing the range of parameters used in creating distortion function candidates (S1211), distortion function candidates are created again, The above-mentioned evaluation (S1205 to S1210) may be repeated. When the processing flow of the present embodiment described above is used, it is possible to appropriately measure the distortion function by using a plurality of distortion functions.

The fifth embodiment is an image processing apparatus, which creates a first predicted image from a part of a first image of a sample taken by a charged particle beam device and a first distortion function. It is an embodiment of an image processing apparatus configured to compare a first expected image with a part of a second image of a moved sample taken by a charged particle beam device.

FIG. 13 is a diagram showing an example in which a long region is cut out in one direction as a part of each acquired image and the distortion function is evaluated with respect to them. In the present embodiment, when (1) and (2) in the same figure are considered as an example of the acquired image, a strip or line long in one direction as a part of the image in each visual field 51 By cutting out the area 54, an image as shown in (3) and (4) of the same figure is obtained. Region 54 includes a portion of sample structure 53. It is possible to measure distortion as described in each of the above-described embodiments. In the case of the present embodiment, it is possible to evaluate the shape of the distortion function in one direction after reducing the number of variables used to represent the distortion function and simplifying them. That is, the distortion function for variables based on coordinates in the image is a function for one variable based on coordinates in the image. Since distortion often takes a shape symmetrical with respect to a specific direction, in the configuration of the present embodiment, for example, with the center of distortion as an origin, distortion in one dimension in several directions such as two directions orthogonal to each other It is possible to measure the functions and evaluate the two-dimensional distortion function contained in the image based on the results.

A sixth embodiment is an image processing apparatus, which creates and generates a first predicted image from a part of a first image of a sample taken by a charged particle beam device and a first distortion function. It is an Example of the image processing apparatus of the other structure which compares with a part of 2nd image of the sample after movement which image | photographed the 1st estimated image and the charged particle beam apparatus.

FIG. 14 is a diagram showing another embodiment in the case where a partial region is cut out from each of the acquired images and the distortion function is evaluated for them. In the figure, when (1) and (2) of the figure are considered as the specifically acquired image, by cutting out the area 54 corresponding to the part in each visual field 51 as a new local image, An image as shown in (3) and (4) of the same figure is obtained. Region 54 includes a portion of sample structure 53. It is possible to measure distortion as described in each of the above-mentioned embodiments. Since distortion applied to an image changes continuously, it can be regarded as simple deformation such as stretching in one direction mainly from a local view.

It is possible to measure a local deformation function represented by a combination of enlargement, reduction, or shear, rotation, translation, etc. with respect to each coordinate axis which composes an image, with respect to two local images cut out. The measurement of the affine transformation matrix using feature points included in the image of each area 54, the comparison after applying the coordinate transformation such as Log-Polar transformation and Hough transformation, the measurement of the correlation value of the image, etc. can be used Can be mentioned as an example of The measurement of such a local deformation function is performed on a plurality of places in the acquired image, and fitting and interpolation processing are performed on those results, so that the configuration of the present embodiment allows two-dimensional inclusion in the acquired image. It is possible to evaluate the information of dynamic distortion function.

In the seventh embodiment, the control unit of the above-described embodiment evaluates the amount of error with respect to the distance from the strain center with respect to the temporary distortion function, and generates a new distortion function based on the amount of error. It is an example of the structure which corrects with respect to an appropriate parameter.

(1) of FIG. 15 shows an example of the result of finding the square of the difference of the intensity value at each coordinate with respect to the images A and B having different amounts of distortion, and the portions shown in white are It corresponds to a portion with strength. Since the degree of monotonically changing image distortion often increases with distance from the distortion center, the difference between the distortion center and the distortion center is small when comparing two results in which different amounts of distortion are added to the same structure. The difference is likely to increase as you move away from. Therefore, also in the result shown in (1) representing the size of the difference, the intensity is low at the center of the image, and the intensity tends to occur as it goes to the outside of the image.

In (2) of the figure, the sum of intensity values possessed by pixels of equal distances is calculated with reference to the image center of (1), normalized with the area of the corresponding pixel, and calculated with respect to the distance from the center. It is shown as a distribution. In this example, although the fluctuation of the intensity caused by the pattern is also included, it is understood that the sum of the intensities increases as the distance from the center increases. When performing distortion measurement as shown in Example 1, the provisional distortion used for evaluation is obtained by performing such comparison between the predicted image obtained from the temporary distortion function and the image actually obtained. The amount of error over distance from the strain center can be evaluated for the function. When generating a new distortion function in the measurement, it is possible to improve the convergence efficiency of the distortion function by correcting the appropriate parameters in the distortion function based on the comparison result. In addition, although one-dimensional error value evaluation is performed with respect to the distance from the radius in the present embodiment, it is also possible to perform two-dimensional error value evaluation using the result of taking the difference as it is. In addition, since it is important to match the positional relationship between the two images whose differences are to be measured in such an evaluation, it is needless to say that the image shift should be corrected before the evaluation.

The eighth embodiment is an embodiment of a configuration for processing the intensity distribution in the observation image captured by the charged particle beam device using the distortion function obtained by the control unit of the above-described embodiment.

FIG. 16 is a diagram for explaining how the intensity distribution in the observation image changes due to the occurrence of image distortion. When obtaining a transmission image of a sample, it is general to irradiate the sample with a two-dimensional beam that can be regarded as a uniform intensity distribution and perform observation. Therefore, the reference value of the observation image intensity is one in the entire visual field. It can often be regarded as However, if distortion occurs in the image, the image to be imaged with a uniform intensity will be locally contracted in the field of view, and the density of intensity in the field of view will change according to the distortion. May cause a bias in intensity.

(1) in the same figure is an example showing a reference intensity distribution in the case where an image is formed by an optical system having no distortion in a state where the beam is irradiated to the sample with uniform intensity. (2) is an example showing a reference intensity distribution in the case of performing imaging with a distorted optical system in a state in which the beam is irradiated to the sample with uniform intensity, and the image is outside the field of view Furthermore, it is stretched to the outside, resulting in a decrease in density of the beam and a decrease in the reference intensity, so that the brightness looks dark. Since the deviation of brightness due to such distortion can be obtained from the distortion function, observation taken by a charged particle beam apparatus such as a TEM using the distortion function measured by the method described in the first embodiment etc. It is possible to correct for brightness non-uniformities in the image.

In this case, assuming that the distortion function is D (x, y) as an example, the following relationship holds between the brightness distribution I (x, y) in the field of view and the distortion function.

Alternatively, it is also possible to measure the bias of the reference intensity in the image and use that value to make an evaluation regarding the distortion function.

In each of the embodiments described above, the image distortion applied when observing the structure of the sample is described, but in addition, when the sample is irradiated with an electron beam, based on the periodicity of the sample structure. When observing the resulting electron beam diffraction pattern or observing the backscattered electron diffraction pattern formed by the irradiated electron beam, it is possible to apply the configuration of each embodiment for the purpose of measuring the distortion caused by the optical system. is there.

The tenth embodiment is an embodiment of a configuration for controlling the scanning of the charged particle beam of the charged particle beam device based on the obtained distortion function.

In each of the above-described embodiments, although an example of distortion generated for image observation by transmitted electrons in the charged particle beam apparatus has been described, in addition, charged particle beams such as electrons and ions converged on the sample are scanned, and In a scanning electron microscope or scanning ion microscope that obtains an image by detecting secondary electrons or backscattered electrons that are generated, the purpose of measuring distortion of a scanning pattern caused by an aberration of a lens disposed in the vicinity of a sample, such as an objective lens. But it is possible to apply. In this case, based on the measured distortion information, it is possible to form a scanning pattern which is substantially less affected by distortion by changing the signal for scanning the charged particle beam.

The present invention is not limited to the embodiments described above, but includes various modifications. For example, the embodiments described above have been described in detail for better understanding of the present invention, and are not necessarily limited to those having all the configurations of the description. In addition, part of the configuration of one embodiment can be replaced with the configuration of another embodiment, and the configuration of another embodiment can be added to the configuration of one embodiment. In addition, with respect to a part of the configuration of each embodiment, it is possible to add, delete, and replace other configurations.

Furthermore, although each configuration, function, control part, etc. mentioned above explained the example which creates a program of CPU which realizes a part or all of them, by designing a part or all of them with an integrated circuit etc. It goes without saying that it may be realized by hardware. That is, all or part of the functions of the control unit may be realized by an integrated circuit such as an application specific integrated circuit (ASIC) or a field programmable gate array (FPGA) instead of the program.

In the above-described specification, various inventions are disclosed in addition to the invention according to claims 1-15 as defined in the claims. Some of them are described below.

<List 1>
An image processing method by the control unit,
The control unit
Creating a first expected image from the first image of the sample or a portion thereof taken by the charged particle beam device and the first distortion function;
Comparing the generated first expected image with a second image of the sample or a portion of the image taken after movement of the field of view by the charged particle beam device;
An image processing method characterized in that.

<List 2>
It is an image processing method described in List 1, and
A second predicted image is created from a second distortion function obtained by correcting the first distortion function and the first image or a part thereof based on the result of the comparison;
Compare the second expected image with the second image or a portion thereof
An image processing method characterized in that.

<List 3>
It is an image processing method described in List 1, and
The creation of the first predicted image or the second predicted image is a process of deforming the first image or a part thereof in a coordinate space.
An image processing method characterized in that.

<List 4>
A charged particle beam device,
An optical system for irradiating a charged particle beam to a sample;
A recording device for recording a secondary charged particle beam obtained from the sample by irradiating the sample with the charged particle beam;
A control unit that controls the optical system and the recording device;
The control unit
Creating a first expected image from the first image of the sample or a portion thereof recorded by the recording device and the first distortion function;
Comparing the generated first expected image with a second image of the sample or a portion of the image taken after movement of the field of view;
Charged particle beam device characterized in that.

<List 5>
The charged particle beam device according to the fourth aspect, wherein
The control unit
A second predicted image is created from a second distortion function obtained by correcting the first distortion function and the first image or a part thereof based on the result of the comparison;
Compare the second expected image with the second image or a portion thereof
Charged particle beam device characterized in that.

<Ex rest 6>
The charged particle beam device according to the fourth aspect, wherein
The creation of the first predicted image or the second predicted image is a process of deforming the first image or a part thereof in a coordinate space.
Charged particle beam device characterized in that.

<List 7>
The charged particle beam device according to the fourth aspect, wherein
It further comprises a moving mechanism for moving the sample, or a deflector for changing the trajectory of the charged particle beam irradiated to the sample,
The control unit moves the sample by the moving mechanism while changing the trajectory of the charged particle beam by the deflector while taking the first image and the second image, and moves the field of view. Control to do,
Charged particle beam device characterized in that.

REFERENCE SIGNS LIST 1 electron source 2, 3 irradiation system lens 4 objective lens 5, 6 imaging system lens 7 aperture 8, 9 deflector 10 aperture 11 recording device 12 recording device control device 13 control device 14 sample 15 image plane 21, 22, 24, Reference Signs List 25 lens electromagnetic field 23 objective lens electromagnetic field 30 optical axis 31 electron trajectories 32, 33 electron trajectories 34, 35 without aberrations electron trajectories 36, 37, 38 electron trajectories 50, 51 viewed from above the sample 52, 53, 54 Sample structures 60, 61, 62 imaged in the field of view

Claims (15)

An image processing apparatus,
Creating a first expected image from the first image of the sample or a portion thereof taken by the charged particle beam device and the first distortion function;
Comparing the generated first expected image with a second image of the sample, or a portion thereof, taken after movement of the field of view by the charged particle beam device;
An image processing apparatus characterized by The image processing apparatus according to claim 1, wherein
A second predicted image is created from a second distortion function obtained by correcting the first distortion function and the first image or a part thereof based on the result of the comparison;
Compare the second expected image with the second image or a portion thereof
An image processing apparatus characterized by The image processing apparatus according to claim 1, wherein
The creation of the first predicted image is a process of deforming the first image or a part thereof in a coordinate space.
An image processing apparatus characterized by The image processing apparatus according to any one of claims 1 to 3, wherein
The charged particle beam apparatus includes a moving mechanism that moves the sample, or a deflector that changes a trajectory of a charged particle beam irradiated to the sample.
While taking the first image and the second image, the movement mechanism moves the sample or the deflector changes the trajectory of the charged particle beam to move the visual field.
An image processing apparatus characterized by The image processing apparatus according to any one of claims 1 to 4, wherein
The first image and the second image have a common area at least partially in common,
An image processing apparatus characterized by The image processing apparatus according to any one of claims 1 to 5, wherein
Generating a plurality of the first predicted images using a plurality of the first distortion functions;
An image processing apparatus characterized by The image processing apparatus according to any one of claims 1 to 6, wherein
Comparing the first predicted image with a plurality of the second images or their parts having different amounts of movement of the visual field by the charged particle beam device;
An image processing apparatus characterized by The image processing apparatus according to any one of claims 1 to 7, wherein
Generating a third distortion function based on the result of the comparison;
An image processing apparatus characterized by The image processing apparatus according to any one of claims 1 to 8, wherein
An image processing apparatus characterized by comparing the first image or a part thereof with the second image or a part thereof and generating a distortion function based on the result. The image processing apparatus according to any one of claims 1 to 9, wherein
The comparison is a process of obtaining one or more of a sum, a difference, a quotient, a correlation value, a Fourier transform result, a histogram, and a feature point of the image.
An image processing apparatus characterized by An image processing apparatus according to any one of claims 1 to 10, wherein
The first distortion function is a function for a variable based on coordinates in the image,
An image processing apparatus characterized by The image processing apparatus according to claim 11, wherein
The first distortion function is a function for one variable based on coordinates in the image,
An image processing apparatus characterized by The image processing apparatus according to any one of claims 1 to 12, wherein
Processing the observation image using the obtained distortion function,
An image processing apparatus characterized by The image processing apparatus according to any one of claims 1 to 13, wherein
Processing the intensity distribution in the observation image captured by the charged particle beam device using the obtained distortion function;
An image processing apparatus characterized by The image processing apparatus according to any one of claims 1 to 13, wherein
Controlling scanning of the charged particle beam of the charged particle beam device based on the obtained distortion function,
An image processing apparatus characterized by

Images